From Rudi at SoftDux.com Sun Mar 2 13:10:42 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Sun, 02 Mar 2008 15:10:42 +0200 Subject: How do I install a NIC driver image from the CD & a kickstart file? Message-ID: <47CAA752.6010503@SoftDux.com> Hi Does anyone know how todo this? According to the CentOS 5 kickstart notes, http://www.linuxtopia.org/online_books/centos5/centos5_installation_guide/centos5_s1-kickstart2-options.html I need to use the driverdisk option, but I can't get it to work The motherboard I use is a Gigabyte GA-32MX-S2, which has the following NIC: http://wiki.centos.org/HardwareList/CentOS5/RealTek/r1000 So, I saved the r1000-2.6.18_8.img file to the ISO image before I burned it to a CD Adding the line "driverdisk=r1000-2.6.18_8.img" to ks.cfg just gives me an error. Trying the option "expert" also gives me an error. So, how do I load a NIC driver disk via the kickstart file? -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From rafaelmorales at cmn-consulting.com Sun Mar 2 16:36:11 2008 From: rafaelmorales at cmn-consulting.com (Rafael Morales) Date: Sun, 2 Mar 2008 11:36:11 -0500 Subject: How do I install a NIC driver image from the CD & a kickstart file? In-Reply-To: <47CAA752.6010503@SoftDux.com> References: <47CAA752.6010503@SoftDux.com> Message-ID: Hi Rudi, Maybe you must add the --type FLAG and give the path of driver like this *driverdisk --source=cdrom://path/to/dd.img *as the path of ks.cfg file* * I hope this help you. On Sun, Mar 2, 2008 at 8:10 AM, Rudi Ahlers wrote: > Hi > > Does anyone know how todo this? > According to the CentOS 5 kickstart notes, > > http://www.linuxtopia.org/online_books/centos5/centos5_installation_guide/centos5_s1-kickstart2-options.html > I need to use the driverdisk option, but I can't get it to work The > motherboard I use is a Gigabyte GA-32MX-S2, which has the following NIC: > http://wiki.centos.org/HardwareList/CentOS5/RealTek/r1000 > So, I saved the r1000-2.6.18_8.img file to the ISO image before I burned > it to a CD > > Adding the line "driverdisk=r1000-2.6.18_8.img" to ks.cfg just gives me > an error. > > Trying the option "expert" also gives me an error. So, how do I load a > NIC driver disk via the kickstart file? > > -- > > Kind Regards > Rudi Ahlers > CEO, SoftDux > > Web: http://www.SoftDux.com > Check out my technical blog, http://blog.softdux.com for Linux or other > technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting > stugg > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > -- Atentamente, Rafael A. Morales C -------------- next part -------------- An HTML attachment was scrubbed... URL: From Rudi at SoftDux.com Mon Mar 3 11:03:50 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 03 Mar 2008 13:03:50 +0200 Subject: how do I detect the HDD's on the system? In-Reply-To: <1204271466.11610.47.camel@iranzo.users.redhat.com> References: <47C718DE.5080206@SoftDux.com> <4484c95febc084b4383b60d754ae4dcc@mail.half-asleep.com> <47C73829.8040509@SoftDux.com> <1204270097.11610.43.camel@iranzo.users.redhat.com> <47C7B85A.4020007@SoftDux.com> <1204271466.11610.47.camel@iranzo.users.redhat.com> Message-ID: <47CBDB16.3090405@SoftDux.com> Pablo Iranzo G?mez wrote: > Hi! > > %include /tmp/part-include > > %pre > set $(list-harddrives) > let numd=$#/2 # N??mero de discos > d1=$1 # 1st HDD drive > S1=$2 # 1st HDD size > > DISCO=$d1 > > echo "clearpart --drives=$DISCO --all --initlabel" >> /tmp/part-include > echo "part /boot --fstype ext3 --size=100 --ondisk=$DISCO" >> /tmp/part-include > echo "part pv.100000 --size=1 --grow --ondisk=$DISCO" >> /tmp/part-include > echo "volgroup VolGroup --pesize=32768 pv.100000" >> /tmp/part-include > echo "logvol swap --fstype swap --name=Swap --vgname=VolGroup --size=2047" >> /tmp/part-include > echo "logvol / --fstype ext3 --name=root --vgname=VolGroup --size=1024" >> /tmp/part-include > echo "logvol /home --fstype ext3 --size=1024 --name=home --vgname=VolGroup" >> /tmp/part-include > echo "logvol /tmp --fstype ext3 --size=1024 --name=tmp --vgname=VolGroup" >> /tmp/part-include > echo "logvol /usr --fstype ext3 --size=3000 --name=usr --vgname=VolGroup" >> /tmp/part-include > echo "logvol /opt --fstype ext3 --size=12000 --name=opt --vgname=VolGroup" >> /tmp/part-include > echo "part /opt/iso --fstype vfat --size=4410 --ondisk=$DISCO" >> /tmp/part-include > > > But with your intended partitioning layout and required "conditionals" > to automatically setup raid as expressed in previous posts on this > thread > > Regards > Pablo > > > > Thanx Pablo I've only had time to play around with this today, but with very little success. Here's my kickstart config for the partitioning: # Determine how many number/type/size of drives we have set $(list-harddrives) let numd=$#/2 # This will provide the total # of drives d1=$1 # This is the device of disk 1 d2=$3 # This is the device of disk 2, etc. d3=$5 d4=$7 d5=%9 S1=$2 # This is the size of disk 1 S2=$4 # This is the size of disk 2, etc. S3=$6 S4=$8 S5=$10 %include /tmp/partinfo # This would be a partition scheme for three or more drives if [ $numd -ge 3 ] ; then cat << EOF >> /tmp/partinfo part raid.11 --size 200 --asprimary --ondrive=$d1 part raid.12 --size 2048 --asprimary --ondrive=$d1 part raid.13 --size 1 --grow --ondrive=$d1 part raid.21 --size 200 --asprimary --ondrive=$d2 part raid.22 --size 2048 --asprimary --ondrive=$d2 part raid.23 --size 1 --grow --ondrive=$d2 part raid.31 --size 200 --asprimary --ondrive=$d3 part raid.32 --size 2048 --asprimary --ondrive=$d3 part raid.33 --size 1 --grow --ondrive=$d3 raid /boot --fstype ext3 --device md0 --level=RAID1 raid.11 raid.21 raid.31 --spare=1 raid swap --fstype swap --device md1 --level=RAID1 raid.12 raid.22 raid.32 --spare=1 raid pv.01 --fstype ext3 --device md2 --level=RAID1 raid.13 raid.23 raid.33 --spare=1 EOF elif [ $numd - ge 2 ]; then cat << EOF >> /tmp/partinfo part /boot --fstype ext3 --size=150 --ondisk=$d1 part swap --size=4092 --ondisk=$d1,$d2 part pv.01 --fstype ext3 --size=100 --grow --ondisk=$d1 part pv.02 --fstype ext3 --size=100 --grow --ondisk=$d2 EOF else cat << EOF >> /tmp/partinfo part /boot --fstype ext3 --size=150 --ondisk=$d1 part swap --size=4092 --ondisk=$d1 part pv.01 --fstype ext3 --size=100 --grow --ondisk=$d1 EOF fi # LVM configuration so that we can resize /, /var and /usr/ later logvol / --vgname=sysvg --size=3072 --name=root logvol /usr --vgname=sysvg --size=8192 --name=usr logvol /tmp --vgname=sysvg --size=1024 --name=tmp logvol /var --vgname=sysvg --size=8192 --name=var logvol /home --vgname=sysvg --size=20480 --name=home if [ pv.02 ]; then logvol /bck --vgname=pv.02 --size=1 --grow --name=bck fi Yet, when I run fdisk -l, there's no partitioning information, almost asif the installation script doesn't run the included /tmp/partinfo file Attached the the /tmp/partinfo & /tmp/anaconda.log files - maybe you could see something I missed? -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: anaconda2.log URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: partinfo URL: From Rudi at SoftDux.com Mon Mar 3 11:40:40 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 03 Mar 2008 13:40:40 +0200 Subject: How do I install a NIC driver image from the CD & a kickstart file? In-Reply-To: References: <47CAA752.6010503@SoftDux.com> Message-ID: <47CBE3B8.5000704@SoftDux.com> Rafael Morales wrote: > Hi Rudi, > > Maybe you must add the --type FLAG and give the path of driver like > this > > *driverdisk --source=cdrom://path/to/dd.img > > > *as the path of ks.cfg file* > > * > I hope this help you. > > On Sun, Mar 2, 2008 at 8:10 AM, Rudi Ahlers > wrote: > > Hi > > Does anyone know how todo this? > According to the CentOS 5 kickstart notes, > http://www.linuxtopia.org/online_books/centos5/centos5_installation_guide/centos5_s1-kickstart2-options.html > I need to use the driverdisk option, but I can't get it to work The > motherboard I use is a Gigabyte GA-32MX-S2, which has the > following NIC: > http://wiki.centos.org/HardwareList/CentOS5/RealTek/r1000 > So, I saved the r1000-2.6.18_8.img file to the ISO image before I > burned > it to a CD > > Adding the line "driverdisk=r1000-2.6.18_8.img" to ks.cfg just > gives me > an error. > > Trying the option "expert" also gives me an error. So, how do I load a > NIC driver disk via the kickstart file? > > -- > > Kind Regards > Rudi Ahlers > CEO, SoftDux > > Web: http://www.SoftDux.com > Check out my technical blog, http://blog.softdux.com for Linux or > other technical stuff, or visit http://www.WebHostingTalk.co.za > for Web Hosting stugg > > _______________________________________________U > Unfortunately that option didn't work. The installer then stops and prompts for a driver disk, which I don't want. -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From Pablo.Iranzo at redhat.com Mon Mar 3 11:50:25 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 03 Mar 2008 12:50:25 +0100 Subject: How do I install a NIC driver image from the CD & a kickstart file? In-Reply-To: <47CBE3B8.5000704@SoftDux.com> References: <47CAA752.6010503@SoftDux.com> <47CBE3B8.5000704@SoftDux.com> Message-ID: <1204545025.19007.28.camel@iranzo.users.redhat.com> It's the driver for the network card... so installer cannot reach network to download it as has no network support yet ;) Regards Pablo El lun, 03-03-2008 a las 13:40 +0200, Rudi Ahlers escribi?: > Rafael Morales wrote: > > Hi Rudi, > > > > Maybe you must add the --type FLAG and give the path of driver like > > this > > > > *driverdisk --source=cdrom://path/to/dd.img > > > > > > *as the path of ks.cfg file* > > > > * > > I hope this help you. > > > > On Sun, Mar 2, 2008 at 8:10 AM, Rudi Ahlers > > wrote: > > > > Hi > > > > Does anyone know how todo this? > > According to the CentOS 5 kickstart notes, > > http://www.linuxtopia.org/online_books/centos5/centos5_installation_guide/centos5_s1-kickstart2-options.html > > I need to use the driverdisk option, but I can't get it to work The > > motherboard I use is a Gigabyte GA-32MX-S2, which has the > > following NIC: > > http://wiki.centos.org/HardwareList/CentOS5/RealTek/r1000 > > So, I saved the r1000-2.6.18_8.img file to the ISO image before I > > burned > > it to a CD > > > > Adding the line "driverdisk=r1000-2.6.18_8.img" to ks.cfg just > > gives me > > an error. > > > > Trying the option "expert" also gives me an error. So, how do I load a > > NIC driver disk via the kickstart file? > > > > -- > > > > Kind Regards > > Rudi Ahlers > > CEO, SoftDux > > > > Web: http://www.SoftDux.com > > Check out my technical blog, http://blog.softdux.com for Linux or > > other technical stuff, or visit http://www.WebHostingTalk.co.za > > for Web Hosting stugg > > > > _______________________________________________U > > > > Unfortunately that option didn't work. The installer then stops and > prompts for a driver disk, which I don't want. > -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Rudi at SoftDux.com Mon Mar 3 11:53:25 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 03 Mar 2008 13:53:25 +0200 Subject: How do I install a NIC driver image from the CD & a kickstart file? In-Reply-To: <1204545025.19007.28.camel@iranzo.users.redhat.com> References: <47CAA752.6010503@SoftDux.com> <47CBE3B8.5000704@SoftDux.com> <1204545025.19007.28.camel@iranzo.users.redhat.com> Message-ID: <47CBE6B5.5090106@SoftDux.com> Pablo Iranzo G?mez wrote: > It's the driver for the network card... so installer cannot reach > network to download it as has no network support yet ;) > > Regards > Pablo > > > El lun, 03-03-2008 a las 13:40 +0200, Rudi Ahlers escribi??: > >> Rafael Morales wrote: >> >>> Hi Rudi, >>> >>> Maybe you must add the --type FLAG and give the path of driver like >>> this >>> >>> *driverdisk --source=cdrom://path/to/dd.img >>> >>> >>> *as the path of ks.cfg file* >>> >>> * >>> I hope this help you. >>> >>> On Sun, Mar 2, 2008 at 8:10 AM, Rudi Ahlers >> > wrote: >>> >>> Hi >>> >>> Does anyone know how todo this? >>> According to the CentOS 5 kickstart notes, >>> http://www.linuxtopia.org/online_books/centos5/centos5_installation_guide/centos5_s1-kickstart2-options.html >>> I need to use the driverdisk option, but I can't get it to work The >>> motherboard I use is a Gigabyte GA-32MX-S2, which has the >>> following NIC: >>> http://wiki.centos.org/HardwareList/CentOS5/RealTek/r1000 >>> So, I saved the r1000-2.6.18_8.img file to the ISO image before I >>> burned >>> it to a CD >>> >>> Adding the line "driverdisk=r1000-2.6.18_8.img" to ks.cfg just >>> gives me >>> an error. >>> >>> Trying the option "expert" also gives me an error. So, how do I load a >>> NIC driver disk via the kickstart file? >>> >>> -- >>> >>> heh, that's funny! You probably didn't read the whole thing, did you? -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From Pablo.Iranzo at redhat.com Mon Mar 3 11:58:50 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 03 Mar 2008 12:58:50 +0100 Subject: How do I install a NIC driver image from the CD & a kickstart file? In-Reply-To: <47CBE6B5.5090106@SoftDux.com> References: <47CAA752.6010503@SoftDux.com> <47CBE3B8.5000704@SoftDux.com> <1204545025.19007.28.camel@iranzo.users.redhat.com> <47CBE6B5.5090106@SoftDux.com> Message-ID: <1204545530.19007.31.camel@iranzo.users.redhat.com> my fault :) I was answering from what I read yesterday :) Try to put --source=/mnt/sysimage/path/to/dd.img Regards :) Pablo El lun, 03-03-2008 a las 13:53 +0200, Rudi Ahlers escribi?: > Pablo Iranzo G?mez wrote: > > It's the driver for the network card... so installer cannot reach > > network to download it as has no network support yet ;) > > > > Regards > > Pablo > > > > > > El lun, 03-03-2008 a las 13:40 +0200, Rudi Ahlers escribi??: > > > >> Rafael Morales wrote: > >> > >>> Hi Rudi, > >>> > >>> Maybe you must add the --type FLAG and give the path of driver like > >>> this > >>> > >>> *driverdisk --source=cdrom://path/to/dd.img > >>> > >>> > >>> *as the path of ks.cfg file* > >>> > >>> * > >>> I hope this help you. > >>> > >>> On Sun, Mar 2, 2008 at 8:10 AM, Rudi Ahlers >>> > wrote: > >>> > >>> Hi > >>> > >>> Does anyone know how todo this? > >>> According to the CentOS 5 kickstart notes, > >>> http://www.linuxtopia.org/online_books/centos5/centos5_installation_guide/centos5_s1-kickstart2-options.html > >>> I need to use the driverdisk option, but I can't get it to work The > >>> motherboard I use is a Gigabyte GA-32MX-S2, which has the > >>> following NIC: > >>> http://wiki.centos.org/HardwareList/CentOS5/RealTek/r1000 > >>> So, I saved the r1000-2.6.18_8.img file to the ISO image before I > >>> burned > >>> it to a CD > >>> > >>> Adding the line "driverdisk=r1000-2.6.18_8.img" to ks.cfg just > >>> gives me > >>> an error. > >>> > >>> Trying the option "expert" also gives me an error. So, how do I load a > >>> NIC driver disk via the kickstart file? > >>> > >>> -- > >>> > >>> > heh, that's funny! You probably didn't read the whole thing, did you? > -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Pablo.Iranzo at redhat.com Mon Mar 3 12:11:57 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 03 Mar 2008 13:11:57 +0100 Subject: how do I detect the HDD's on the system? In-Reply-To: <47CBDB16.3090405@SoftDux.com> References: <47C718DE.5080206@SoftDux.com> <4484c95febc084b4383b60d754ae4dcc@mail.half-asleep.com> <47C73829.8040509@SoftDux.com> <1204270097.11610.43.camel@iranzo.users.redhat.com> <47C7B85A.4020007@SoftDux.com> <1204271466.11610.47.camel@iranzo.users.redhat.com> <47CBDB16.3090405@SoftDux.com> Message-ID: <1204546317.19007.34.camel@iranzo.users.redhat.com> Rudi, El lun, 03-03-2008 a las 13:03 +0200, Rudi Ahlers escribi?: > # Determine how many number/type/size of drives we have > set $(list-harddrives) > let numd=$#/2 # This will provide the total # of drives > d1=$1 # This is the device of disk 1 > d2=$3 # This is the device of disk 2, etc. > d3=$5 > d4=$7 > d5=%9 Check this "%" > 12:20:49 INFO : Running kickstart %%pre script(s) > 12:21:00 ERROR : Error code 127 encountered running a kickstart %pre/%post script > 12:21:00 INFO : All kickstart %%pre script(s) have been run Anaconda complains about an error running script, so check above comment Regards Pablo -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Rudi at SoftDux.com Mon Mar 3 13:15:17 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 03 Mar 2008 15:15:17 +0200 Subject: How do I install a NIC driver image from the CD & a kickstart file? In-Reply-To: <1204545530.19007.31.camel@iranzo.users.redhat.com> References: <47CAA752.6010503@SoftDux.com> <47CBE3B8.5000704@SoftDux.com> <1204545025.19007.28.camel@iranzo.users.redhat.com> <47CBE6B5.5090106@SoftDux.com> <1204545530.19007.31.camel@iranzo.users.redhat.com> Message-ID: <47CBF9E5.60802@SoftDux.com> Pablo Iranzo G?mez wrote: > my fault :) > I was answering from what I read yesterday :) > > Try to put --source=/mnt/sysimage/path/to/dd.img > > Regards :) > Pablo > > El lun, 03-03-2008 a las 13:53 +0200, Rudi Ahlers escribi??: > >> Pablo Iranzo G??mez wrote: >> >>> It's the driver for the network card... so installer cannot reach >>> network to download it as has no network support yet ;) >>> >>> Regards >>> Pablo >>> That didn't work either, since the CD isn't mounted yet. I guess this won't be possible... -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From Pablo.Iranzo at redhat.com Mon Mar 3 14:00:03 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 03 Mar 2008 15:00:03 +0100 Subject: How do I install a NIC driver image from the CD & a kickstart file? In-Reply-To: <47CBF9E5.60802@SoftDux.com> References: <47CAA752.6010503@SoftDux.com> <47CBE3B8.5000704@SoftDux.com> <1204545025.19007.28.camel@iranzo.users.redhat.com> <47CBE6B5.5090106@SoftDux.com> <1204545530.19007.31.camel@iranzo.users.redhat.com> <47CBF9E5.60802@SoftDux.com> Message-ID: <1204552803.19007.41.camel@iranzo.users.redhat.com> You could try the trick: mkdir -p /mnt/sysimage mount /dev/cdrom /mnt/sysimage/ cp /mnt/sysimage/path/to/dd.img /tmp umount /mnt/sysimage and in your ks, put --source=/tmp/dd.img Regards Pablo El lun, 03-03-2008 a las 15:15 +0200, Rudi Ahlers escribi?: > Pablo Iranzo G?mez wrote: > > my fault :) > > I was answering from what I read yesterday :) > > > > Try to put --source=/mnt/sysimage/path/to/dd.img > > > > Regards :) > > Pablo > > > > El lun, 03-03-2008 a las 13:53 +0200, Rudi Ahlers escribi??: > > > >> Pablo Iranzo G??mez wrote: > >> > >>> It's the driver for the network card... so installer cannot reach > >>> network to download it as has no network support yet ;) > >>> > >>> Regards > >>> Pablo > >>> > That didn't work either, since the CD isn't mounted yet. I guess this > won't be possible... > -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Rudi at SoftDux.com Mon Mar 3 15:25:16 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 03 Mar 2008 17:25:16 +0200 Subject: how do I detect the HDD's on the system? In-Reply-To: <1204546317.19007.34.camel@iranzo.users.redhat.com> References: <47C718DE.5080206@SoftDux.com> <4484c95febc084b4383b60d754ae4dcc@mail.half-asleep.com> <47C73829.8040509@SoftDux.com> <1204270097.11610.43.camel@iranzo.users.redhat.com> <47C7B85A.4020007@SoftDux.com> <1204271466.11610.47.camel@iranzo.users.redhat.com> <47CBDB16.3090405@SoftDux.com> <1204546317.19007.34.camel@iranzo.users.redhat.com> Message-ID: <47CC185C.1080800@SoftDux.com> Pablo Iranzo G?mez wrote: > Rudi, > > El lun, 03-03-2008 a las 13:03 +0200, Rudi Ahlers escribi??: > >> # Determine how many number/type/size of drives we have >> set $(list-harddrives) >> let numd=$#/2 # This will provide the total # of drives >> d1=$1 # This is the device of disk 1 >> d2=$3 # This is the device of disk 2, etc. >> d3=$5 >> d4=$7 >> d5=%9 >> > > Check this "%" > > > >> 12:20:49 INFO : Running kickstart %%pre script(s) >> 12:21:00 ERROR : Error code 127 encountered running a kickstart %pre/%post script >> 12:21:00 INFO : All kickstart %%pre script(s) have been run >> > > Anaconda complains about an error running script, so check above comment > > > Regards > Pablo > > ------------------------------------------------------------------------ > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list oops :) I fixed it, but still no luck. Here's the current output files. -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: partinfo3 URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: annaconda3.log URL: From Pablo.Iranzo at redhat.com Mon Mar 3 15:31:12 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 03 Mar 2008 16:31:12 +0100 Subject: how do I detect the HDD's on the system? In-Reply-To: <47CC185C.1080800@SoftDux.com> References: <47C718DE.5080206@SoftDux.com> <4484c95febc084b4383b60d754ae4dcc@mail.half-asleep.com> <47C73829.8040509@SoftDux.com> <1204270097.11610.43.camel@iranzo.users.redhat.com> <47C7B85A.4020007@SoftDux.com> <1204271466.11610.47.camel@iranzo.users.redhat.com> <47CBDB16.3090405@SoftDux.com> <1204546317.19007.34.camel@iranzo.users.redhat.com> <47CC185C.1080800@SoftDux.com> Message-ID: <1204558272.19007.46.camel@iranzo.users.redhat.com> ?Can you send me your ks file? Regards Pablo El lun, 03-03-2008 a las 17:25 +0200, Rudi Ahlers escribi?: > tep installtype does not exist > 16:52:32 WARNING : step complete does not exist > 16:52:32 WARNING : step complete does not exist > 16:52:32 WARNING : step complete does not exist > 16:52:33 INFO : moving (1) to step regkey > 16:52:33 INFO : moving (1) to step partitionobjinit > 16:52:33 INFO : no initiator set -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Rudi at SoftDux.com Mon Mar 3 15:48:24 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 03 Mar 2008 17:48:24 +0200 Subject: how do I detect the HDD's on the system? In-Reply-To: <1204558272.19007.46.camel@iranzo.users.redhat.com> References: <47C718DE.5080206@SoftDux.com> <4484c95febc084b4383b60d754ae4dcc@mail.half-asleep.com> <47C73829.8040509@SoftDux.com> <1204270097.11610.43.camel@iranzo.users.redhat.com> <47C7B85A.4020007@SoftDux.com> <1204271466.11610.47.camel@iranzo.users.redhat.com> <47CBDB16.3090405@SoftDux.com> <1204546317.19007.34.camel@iranzo.users.redhat.com> <47CC185C.1080800@SoftDux.com> <1204558272.19007.46.camel@iranzo.users.redhat.com> Message-ID: <47CC1DC8.8090006@SoftDux.com> Pablo Iranzo G?mez wrote: > ??Can you send me your ks file? > > Regards > Pablo > > El lun, 03-03-2008 a las 17:25 +0200, Rudi Ahlers escribi??: > >> tep installtype does not exist >> 16:52:32 WARNING : step complete does not exist >> 16:52:32 WARNING : step complete does not exist >> 16:52:32 WARNING : step complete does not exist >> 16:52:33 INFO : moving (1) to step regkey >> 16:52:33 INFO : moving (1) to step partitionobjinit >> 16:52:33 INFO : no initiator set >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list I've modified it a bit, with a less complex layout (in fact, the sample layout from the PDF - Kickstart tricks) The ks_raid is the full RAID kickstart file, and the ks_xen is my original kickstart file, before I tried to incorporate the RAID setup -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ks_xen.cfg URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ks.cfg URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ks_raid.cfg URL: From michael.gale at pason.com Tue Mar 4 15:47:17 2008 From: michael.gale at pason.com (Michael Gale) Date: Tue, 04 Mar 2008 08:47:17 -0700 Subject: LVM and kickstarts ? Message-ID: <47CD6F05.2080200@pason.com> Hey, Can anyone tell me why option 1 works and option 2 fails ? It fails on trying to format my logical volume, because the mount point does not exist (/dev/volgroup/logvol) It seems that with option 2, the partitions are created and LVM is setup correctly. However the volgroup / logvolume was not made active, so my /dev/volgroup/logvol did not exist. Running `lvm lvchange -a -y pathname` from with in the shell after anaconda failed made the volgroup / logvol active. Which would allow the format command to complete. Option 1: zerombr yes clearpart --all --initlabel part /boot --fstype ext3 --size=100 part pv.os --size=10000 --grow --maxsize=10000 --asprimary volgroup os_volgroup01 pv.os logvol swap --fstype swap --name=os_swap_logvol --vgname=os_volgroup01 --recommended logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 --grow Option 2: zerombr yes clearpart --all --initlabel part /boot --fstype ext3 --size=100 part swap --size=1000 part pv.os --size=10000 --grow --maxsize=10000 --asprimary volgroup os_volgroup01 pv.os logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 --grow Thank you -- Michael Gale Red Hat Certified Engineer Network Administrator Pason Systems Corp. "It's not that I'm so smart, it's just that I stay with problems longer." - Albert Einstein From debian at herakles.homelinux.org Wed Mar 5 13:13:14 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Wed, 05 Mar 2008 22:13:14 +0900 Subject: LVM and kickstarts ? In-Reply-To: <47CD6F05.2080200@pason.com> References: <47CD6F05.2080200@pason.com> Message-ID: <47CE9C6A.5030005@herakles.homelinux.org> Michael Gale wrote: > Hey, > > Can anyone tell me why option 1 works and option 2 fails ? What are you installing? > > It fails on trying to format my logical volume, because the mount point > does not exist (/dev/volgroup/logvol) > > It seems that with option 2, the partitions are created and LVM is setup > correctly. However the volgroup / logvolume was not made active, so my > /dev/volgroup/logvol did not exist. > > Running `lvm lvchange -a -y pathname` from with in the shell after > anaconda failed made the volgroup / logvol active. Which would allow the > format command to complete. > > Option 1: > zerombr yes In CentOS5, Anaconda complains (but continues) with that, and so my question. > clearpart --all --initlabel > part /boot --fstype ext3 --size=100 > part pv.os --size=10000 --grow --maxsize=10000 --asprimary > volgroup os_volgroup01 pv.os > logvol swap --fstype swap --name=os_swap_logvol --vgname=os_volgroup01 > --recommended > logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 > --grow > > Option 2: > zerombr yes > clearpart --all --initlabel > part /boot --fstype ext3 --size=100 > part swap --size=1000 > part pv.os --size=10000 --grow --maxsize=10000 --asprimary > volgroup os_volgroup01 pv.os > logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 > --grow I'll let you compare, this works with CentOS5: zerombr part /boot --fstype "ext3" --size=100 --ondisk=hda part pv.6 --size=0 --grow --ondisk=hda volgroup VolGroup00 pv.6 logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 --grow logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=512 -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From gary at mlbassoc.com Wed Mar 5 13:24:00 2008 From: gary at mlbassoc.com (Gary Thomas) Date: Wed, 05 Mar 2008 06:24:00 -0700 Subject: LVM and kickstarts ? In-Reply-To: <47CE9C6A.5030005@herakles.homelinux.org> References: <47CD6F05.2080200@pason.com> <47CE9C6A.5030005@herakles.homelinux.org> Message-ID: <47CE9EF0.10103@mlbassoc.com> John Summerfield wrote: > Michael Gale wrote: >> Hey, >> >> Can anyone tell me why option 1 works and option 2 fails ? > > What are you installing? >> >> It fails on trying to format my logical volume, because the mount point >> does not exist (/dev/volgroup/logvol) >> >> It seems that with option 2, the partitions are created and LVM is setup >> correctly. However the volgroup / logvolume was not made active, so my >> /dev/volgroup/logvol did not exist. >> >> Running `lvm lvchange -a -y pathname` from with in the shell after >> anaconda failed made the volgroup / logvol active. Which would allow the >> format command to complete. >> >> Option 1: >> zerombr yes > In CentOS5, Anaconda complains (but continues) with that, and so my > question. > >> clearpart --all --initlabel >> part /boot --fstype ext3 --size=100 >> part pv.os --size=10000 --grow --maxsize=10000 --asprimary >> volgroup os_volgroup01 pv.os >> logvol swap --fstype swap --name=os_swap_logvol --vgname=os_volgroup01 >> --recommended >> logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 >> --grow >> >> Option 2: >> zerombr yes >> clearpart --all --initlabel >> part /boot --fstype ext3 --size=100 >> part swap --size=1000 >> part pv.os --size=10000 --grow --maxsize=10000 --asprimary >> volgroup os_volgroup01 pv.os >> logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 >> --grow > > I'll let you compare, this works with CentOS5: > zerombr > part /boot --fstype "ext3" --size=100 --ondisk=hda > part pv.6 --size=0 --grow --ondisk=hda > volgroup VolGroup00 pv.6 > logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 > --grow > logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=512 This setup works on Fedor9 (alpha) # The following is the partition information you requested # Note that any partitions you deleted are not expressed # here so unless you clear all partitions first, this is # not guaranteed to work clearpart --linux --drives=sda part /boot --fstype ext3 --size=200 --ondisk=sda part pv.3 --size=0 --grow --ondisk=sda volgroup VolGroup00 --pesize=32768 pv.3 logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 --grow logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=256 --grow --maxsize=512 From michael.gale at pason.com Wed Mar 5 15:02:12 2008 From: michael.gale at pason.com (Michael Gale) Date: Wed, 05 Mar 2008 08:02:12 -0700 Subject: LVM and kickstarts ? In-Reply-To: <47CE9C6A.5030005@herakles.homelinux.org> References: <47CD6F05.2080200@pason.com> <47CE9C6A.5030005@herakles.homelinux.org> Message-ID: <47CEB5F4.9080102@pason.com> Hey, I am currently using Centos5, the method you provided has the swap part of the LVM. This works with out an issue, the problem I am having is with the setup I show under option2. If the swap is not part of the LVM, it fails. Michael John Summerfield wrote: > Michael Gale wrote: >> Hey, >> >> Can anyone tell me why option 1 works and option 2 fails ? > > What are you installing? >> >> It fails on trying to format my logical volume, because the mount point >> does not exist (/dev/volgroup/logvol) >> >> It seems that with option 2, the partitions are created and LVM is setup >> correctly. However the volgroup / logvolume was not made active, so my >> /dev/volgroup/logvol did not exist. >> >> Running `lvm lvchange -a -y pathname` from with in the shell after >> anaconda failed made the volgroup / logvol active. Which would allow the >> format command to complete. >> >> Option 1: >> zerombr yes > In CentOS5, Anaconda complains (but continues) with that, and so my > question. > >> clearpart --all --initlabel >> part /boot --fstype ext3 --size=100 >> part pv.os --size=10000 --grow --maxsize=10000 --asprimary >> volgroup os_volgroup01 pv.os >> logvol swap --fstype swap --name=os_swap_logvol --vgname=os_volgroup01 >> --recommended >> logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 >> --grow >> >> Option 2: >> zerombr yes >> clearpart --all --initlabel >> part /boot --fstype ext3 --size=100 >> part swap --size=1000 >> part pv.os --size=10000 --grow --maxsize=10000 --asprimary >> volgroup os_volgroup01 pv.os >> logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 >> --grow > > I'll let you compare, this works with CentOS5: > zerombr > part /boot --fstype "ext3" --size=100 --ondisk=hda > part pv.6 --size=0 --grow --ondisk=hda > volgroup VolGroup00 pv.6 > logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 > --grow > logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=512 > > > -- Michael Gale Red Hat Certified Engineer Network Administrator Pason Systems Corp. "It's not that I'm so smart, it's just that I stay with problems longer." - Albert Einstein From michael.gale at pason.com Wed Mar 5 15:02:45 2008 From: michael.gale at pason.com (Michael Gale) Date: Wed, 05 Mar 2008 08:02:45 -0700 Subject: LVM and kickstarts ? In-Reply-To: <47CE9EF0.10103@mlbassoc.com> References: <47CD6F05.2080200@pason.com> <47CE9C6A.5030005@herakles.homelinux.org> <47CE9EF0.10103@mlbassoc.com> Message-ID: <47CEB615.2040203@pason.com> Hey, I am currently using Centos5, the method you provided has the swap part of the LVM. This works with out an issue, the problem I am having is with the setup I show under option2. If the swap is not part of the LVM, it fails. Michael Gary Thomas wrote: > John Summerfield wrote: >> Michael Gale wrote: >>> Hey, >>> >>> Can anyone tell me why option 1 works and option 2 fails ? >> >> What are you installing? >>> >>> It fails on trying to format my logical volume, because the mount point >>> does not exist (/dev/volgroup/logvol) >>> >>> It seems that with option 2, the partitions are created and LVM is setup >>> correctly. However the volgroup / logvolume was not made active, so my >>> /dev/volgroup/logvol did not exist. >>> >>> Running `lvm lvchange -a -y pathname` from with in the shell after >>> anaconda failed made the volgroup / logvol active. Which would allow the >>> format command to complete. >>> >>> Option 1: >>> zerombr yes >> In CentOS5, Anaconda complains (but continues) with that, and so my >> question. >> >>> clearpart --all --initlabel >>> part /boot --fstype ext3 --size=100 >>> part pv.os --size=10000 --grow --maxsize=10000 --asprimary >>> volgroup os_volgroup01 pv.os >>> logvol swap --fstype swap --name=os_swap_logvol --vgname=os_volgroup01 >>> --recommended >>> logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 >>> --grow >>> >>> Option 2: >>> zerombr yes >>> clearpart --all --initlabel >>> part /boot --fstype ext3 --size=100 >>> part swap --size=1000 >>> part pv.os --size=10000 --grow --maxsize=10000 --asprimary >>> volgroup os_volgroup01 pv.os >>> logvol / --fstype ext3 --name=os_root --vgname=os_volgroup01 --size=1024 >>> --grow >> >> I'll let you compare, this works with CentOS5: >> zerombr >> part /boot --fstype "ext3" --size=100 --ondisk=hda >> part pv.6 --size=0 --grow --ondisk=hda >> volgroup VolGroup00 pv.6 >> logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 >> --grow >> logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=512 > > This setup works on Fedor9 (alpha) > > # The following is the partition information you requested > # Note that any partitions you deleted are not expressed > # here so unless you clear all partitions first, this is > # not guaranteed to work > clearpart --linux --drives=sda > part /boot --fstype ext3 --size=200 --ondisk=sda > part pv.3 --size=0 --grow --ondisk=sda > volgroup VolGroup00 --pesize=32768 pv.3 > logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 > --grow > logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=256 > --grow --maxsize=512 > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Michael Gale Red Hat Certified Engineer Network Administrator Pason Systems Corp. "It's not that I'm so smart, it's just that I stay with problems longer." - Albert Einstein From rainer at ultra-secure.de Wed Mar 5 15:07:29 2008 From: rainer at ultra-secure.de (Rainer Duffner) Date: Wed, 05 Mar 2008 16:07:29 +0100 Subject: LVM and kickstarts ? In-Reply-To: <47CEB615.2040203@pason.com> References: <47CD6F05.2080200@pason.com> <47CE9C6A.5030005@herakles.homelinux.org> <47CE9EF0.10103@mlbassoc.com> <47CEB615.2040203@pason.com> Message-ID: <47CEB731.8020109@ultra-secure.de> Michael Gale schrieb: > Hey, > > I am currently using Centos5, the method you provided has the swap > part of the LVM. This works with out an issue, the problem I am having > is with the setup I show under option2. > > If the swap is not part of the LVM, it fails. I think I've also seen this. In the end, I went the path of least resistance... Fancy filing a bug? cheers, Rainer From michael.gale at pason.com Wed Mar 5 15:12:28 2008 From: michael.gale at pason.com (Michael Gale) Date: Wed, 05 Mar 2008 08:12:28 -0700 Subject: LVM and kickstarts ? In-Reply-To: <47CEB731.8020109@ultra-secure.de> References: <47CD6F05.2080200@pason.com> <47CE9C6A.5030005@herakles.homelinux.org> <47CE9EF0.10103@mlbassoc.com> <47CEB615.2040203@pason.com> <47CEB731.8020109@ultra-secure.de> Message-ID: <47CEB85C.6090100@pason.com> Hey, Sure, where does one file a bug report ? Michael Rainer Duffner wrote: > Michael Gale schrieb: >> Hey, >> >> I am currently using Centos5, the method you provided has the swap >> part of the LVM. This works with out an issue, the problem I am having >> is with the setup I show under option2. >> >> If the swap is not part of the LVM, it fails. > > > I think I've also seen this. > In the end, I went the path of least resistance... > > Fancy filing a bug? > > > cheers, > Rainer > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Michael Gale Red Hat Certified Engineer Network Administrator Pason Systems Corp. "It's not that I'm so smart, it's just that I stay with problems longer." - Albert Einstein From rainer at ultra-secure.de Wed Mar 5 15:14:47 2008 From: rainer at ultra-secure.de (Rainer Duffner) Date: Wed, 05 Mar 2008 16:14:47 +0100 Subject: LVM and kickstarts ? In-Reply-To: <47CEB85C.6090100@pason.com> References: <47CD6F05.2080200@pason.com> <47CE9C6A.5030005@herakles.homelinux.org> <47CE9EF0.10103@mlbassoc.com> <47CEB615.2040203@pason.com> <47CEB731.8020109@ultra-secure.de> <47CEB85C.6090100@pason.com> Message-ID: <47CEB8E7.7080706@ultra-secure.de> Michael Gale schrieb: > Hey, > > Sure, where does one file a bug report ? > I think on the RH Bugzilla. https://bugzilla.redhat.com Rainer From rbrown at BallistaSec.com Wed Mar 5 16:36:47 2008 From: rbrown at BallistaSec.com (Rodrick Brown) Date: Wed, 5 Mar 2008 11:36:47 -0500 Subject: CentOS 5.1 kickstart with Intel Matrix Raid Manager Message-ID: I'm trying to Kickstart a server that's already configured with raid 5 from the bios (Intel Raid Manager) but during the installation I get a segmentation fault Error activating raid isw_bbadbhacab_Array failed: raid5_la 1 256 3 -l /dev/sda 0 /dev/sdb 0 /dev/sdc 0 Error exception: device mapper: reload ioctl failed invalid arguments. Has anyone ran into this bug before? --- Rodrick R. Brown Ballista Securities LLC 120 Wall St. Suite 2400 C: 347 702 0012 | O: 646 307 4709 http://www.ballistasec.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From rainer at ultra-secure.de Wed Mar 5 16:40:26 2008 From: rainer at ultra-secure.de (Rainer Duffner) Date: Wed, 05 Mar 2008 17:40:26 +0100 Subject: CentOS 5.1 kickstart with Intel Matrix Raid Manager In-Reply-To: References: Message-ID: <47CECCFA.4010404@ultra-secure.de> Rodrick Brown schrieb: > > I?m trying to Kickstart a server that?s already configured with raid 5 > from the bios (Intel Raid Manager) but during the installation I get a > segmentation fault > > Error activating raid isw_bbadbhacab_Array failed: raid5_la 1 256 3 > ?l /dev/sda 0 /dev/sdb 0 /dev/sdc 0 > > Error exception: device mapper: reload ioctl failed invalid arguments. > > > > Has anyone ran into this bug before? > Does MatrixRAID actually work with Linux? With RHEL4, I resorted to SW-RAID. We're lobbying for not purchasing any DL320 servers anymore for exactly this reason. The controllers don't even work correctly with W2K3, I'm told. cheers, Rainer From rbrown at BallistaSec.com Wed Mar 5 17:23:19 2008 From: rbrown at BallistaSec.com (Rodrick Brown) Date: Wed, 5 Mar 2008 12:23:19 -0500 Subject: CentOS 5.1 kickstart with Intel Matrix Raid Manager In-Reply-To: <47CECCFA.4010404@ultra-secure.de> References: <47CECCFA.4010404@ultra-secure.de> Message-ID: Yeah doesn?t look like its working :( I guess I will just revert to software raid. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Rainer Duffner Sent: Wednesday, March 05, 2008 11:40 AM To: Discussion list about Kickstart Subject: Re: CentOS 5.1 kickstart with Intel Matrix Raid Manager Rodrick Brown schrieb: > > I?m trying to Kickstart a server that?s already configured with raid 5 > from the bios (Intel Raid Manager) but during the installation I get a > segmentation fault > > Error activating raid isw_bbadbhacab_Array failed: raid5_la 1 256 3 > ?l /dev/sda 0 /dev/sdb 0 /dev/sdc 0 > > Error exception: device mapper: reload ioctl failed invalid arguments. > > > > Has anyone ran into this bug before? > Does MatrixRAID actually work with Linux? With RHEL4, I resorted to SW-RAID. We're lobbying for not purchasing any DL320 servers anymore for exactly this reason. The controllers don't even work correctly with W2K3, I'm told. cheers, Rainer _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From bkearney at redhat.com Wed Mar 5 19:33:21 2008 From: bkearney at redhat.com (Bryan Kearney) Date: Wed, 05 Mar 2008 14:33:21 -0500 Subject: Resend: Anaconda modifying network files Message-ID: <47CEF581.8000208@redhat.com> I am using cobbler/koan to generate xen disk images. After a successful install, if I mount the image I see infomation in various networking files which denote the information about the network I built the image on. Specifically: /etc/hosts has the ip address and dhcp fqdn of the image when anaconda was running. /etc/sysconfig/network has the fgdn of the image when anaconda was running. /etc/sysconfig/network-scripts/ifcfg-eth0 has the MAC address given out by xen. Is there a way to configure anaconda to not mutate these files? Thanks! -- bk From gene.huft at wachovia.com Wed Mar 5 19:43:45 2008 From: gene.huft at wachovia.com (gene.huft at wachovia.com) Date: Wed, 5 Mar 2008 14:43:45 -0500 Subject: Kickstarting to VM via ISO image Message-ID: Kickstart Gurus, I have a fully contained ISO image of the original RHEL5.1 32-bit source tree, along with some additional RPMs & a custom ks.cfg file. The goal is to totally automate the build to a VMware VM via mapping the ISO file to a virtual CDROM. I have modified the default isolinux/isolinux.cfg file as follows: default xander prompt 1 timeout 600 display boot.msg F1 boot.msg F2 options.msg F3 general.msg F4 param.msg F5 rescue.msg label xander kernel vmlinuz append initrd=initrd.img ramdisk_size=8192 ks=cdrom:/ks.cfg label text kernel vmlinuz append initrd=initrd.img text label ks kernel vmlinuz append ks initrd=initrd.img label local localboot 1 label memtest86 kernel memtest append - In my kickstart file, I of cource specify "cdrom". The ISO image will boot the VM into RedHat's initial screen, but when hit return or let it time out it, it complains that it cannot find the install media on any cdrom. I also tried to put the kickstart file into the initrd.img by uncompressing image, copying ks.cfg into it, modifying isolinux.cfg to: append initrd=initrd.img ramdisk_size=8192 ks=file:/ks.cfg Then recompressed it, moved it into proper place & re-generated ISO image. Again, the ISO will boot the VM to initial RH screen, but soon there after get the "cd not found" error. Is this just a VMware thing? My coworker has no trouble with SuSE builds working....am I overlooking something simple? Thanks in advance. Gene -------------- next part -------------- An HTML attachment was scrubbed... URL: From mdehaan at redhat.com Wed Mar 5 19:46:12 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Mar 2008 14:46:12 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CEF581.8000208@redhat.com> References: <47CEF581.8000208@redhat.com> Message-ID: <47CEF884.7040700@redhat.com> Bryan Kearney wrote: > I am using cobbler/koan to generate xen disk images. After a > successful install, if I mount the image I see infomation in various > networking files which denote the information about the network I > built the image on. Specifically: > > > /etc/hosts has the ip address and dhcp fqdn of the image when anaconda > was running. > /etc/sysconfig/network has the fgdn of the image when anaconda was > running. > /etc/sysconfig/network-scripts/ifcfg-eth0 has the MAC address given > out by xen. > > Is there a way to configure anaconda to not mutate these files? > > Thanks! > > -- bk > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list Re: disk-image reuse As far as I'm aware the "appliancey" way to reuse disk images is to login in and fix all this once you fire the machines up. This is something you could possibly make firstboot do. The approach we took with Virt-Factory is to not use disk images, but start the install fresh every time, and use config management to push changes down. In order to make sure all the MACs are unique if you regenerate them, you're going to need some sort of central management chokepoint though... --Michael From Greg.Caetano at hp.com Wed Mar 5 19:47:59 2008 From: Greg.Caetano at hp.com (Caetano, Greg) Date: Wed, 5 Mar 2008 19:47:59 +0000 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CEF581.8000208@redhat.com> References: <47CEF581.8000208@redhat.com> Message-ID: Bryan: I take care of removing all those "specifics" as part of the %post of a kickstart Greg Caetano HP ISS Linux Virtualization Solutions Engineering Chicago, IL greg.caetano at hp.com Red Hat Certified Engineer RHCE#805007310328754 -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Bryan Kearney Sent: Wednesday, March 05, 2008 1:33 PM To: kickstart-list at redhat.com Subject: Resend: Anaconda modifying network files I am using cobbler/koan to generate xen disk images. After a successful install, if I mount the image I see infomation in various networking files which denote the information about the network I built the image on. Specifically: /etc/hosts has the ip address and dhcp fqdn of the image when anaconda was running. /etc/sysconfig/network has the fgdn of the image when anaconda was running. /etc/sysconfig/network-scripts/ifcfg-eth0 has the MAC address given out by xen. Is there a way to configure anaconda to not mutate these files? Thanks! -- bk _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From Chip.Shabazian at bankofamerica.com Wed Mar 5 19:57:16 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Wed, 05 Mar 2008 11:57:16 -0800 Subject: Kickstarting to VM via ISO image In-Reply-To: Message-ID: Have you looked at Revisor? You should be able to spin your RHEL 5 install quickly and easily using Revisor: http://revisor.fedoraunity.org/ ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of gene.huft at wachovia.com Sent: Wednesday, March 05, 2008 11:44 AM To: kickstart-list at redhat.com Subject: Kickstarting to VM via ISO image Kickstart Gurus, I have a fully contained ISO image of the original RHEL5.1 32-bit source tree, along with some additional RPMs & a custom ks.cfg file. The goal is to totally automate the build to a VMware VM via mapping the ISO file to a virtual CDROM. I have modified the default isolinux/isolinux.cfg file as follows: default xander prompt 1 timeout 600 display boot.msg F1 boot.msg F2 options.msg F3 general.msg F4 param.msg F5 rescue.msg label xander kernel vmlinuz append initrd=initrd.img ramdisk_size=8192 ks=cdrom:/ks.cfg label text kernel vmlinuz append initrd=initrd.img text label ks kernel vmlinuz append ks initrd=initrd.img label local localboot 1 label memtest86 kernel memtest append - In my kickstart file, I of cource specify "cdrom". The ISO image will boot the VM into RedHat's initial screen, but when hit return or let it time out it, it complains that it cannot find the install media on any cdrom. I also tried to put the kickstart file into the initrd.img by uncompressing image, copying ks.cfg into it, modifying isolinux.cfg to: append initrd=initrd.img ramdisk_size=8192 ks=file:/ks.cfg Then recompressed it, moved it into proper place & re-generated ISO image. Again, the ISO will boot the VM to initial RH screen, but soon there after get the "cd not found" error. Is this just a VMware thing? My coworker has no trouble with SuSE builds working....am I overlooking something simple? Thanks in advance. Gene -------------- next part -------------- An HTML attachment was scrubbed... URL: From mdehaan at redhat.com Wed Mar 5 20:04:34 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Mar 2008 15:04:34 -0500 Subject: Kickstarting to VM via ISO image In-Reply-To: References: Message-ID: <47CEFCD2.6020403@redhat.com> Shabazian, Chip wrote: > Have you looked at Revisor? You should be able to spin your RHEL 5 > install quickly and easily using Revisor: > http://revisor.fedoraunity.org/ > ------------------------------------------------------------------------ > *From:* kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] *On Behalf Of > *gene.huft at wachovia.com > *Sent:* Wednesday, March 05, 2008 11:44 AM > *To:* kickstart-list at redhat.com > *Subject:* Kickstarting to VM via ISO image > > > Kickstart Gurus, > > I have a fully contained ISO image of the original RHEL5.1 32-bit > source tree, along with some additional RPMs & a custom ks.cfg file. > The goal is to totally automate the build to a VMware VM via mapping > the ISO file to a virtual CDROM. > Also of potential interest, if you have Cobbler and PXE infrastructure (or can set it up): https://www.redhat.com/archives/et-mgmt-tools/2008-March/msg00004.html If you want to just do ISOs, the easier approach (IMO) to the above is to take the rescue CD and just add the "ks=" line to it with a network kickstart -- in most applications that will work. Of course that requires that the server location be accessible over the network -- which in itself is not too bad. From mdehaan at redhat.com Wed Mar 5 20:09:30 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Mar 2008 15:09:30 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: References: <47CEF581.8000208@redhat.com> Message-ID: <47CEFDFA.9000603@redhat.com> Caetano, Greg wrote: > Bryan: > > I take care of removing all those "specifics" as part of the %post of a kickstart > > > So in that case, you still have to run system-config-network at firstboot (or equivalent)? That seems reasonable -- you may be handing something to someone with a static IP setup, you don't know. Rethinking it, if you clear things out in %post and also feed it network details in the libvirt XML, maybe that's good enough... Still, I'm not sure of a way to do that fully automatically and get them all unique MACs without being rather clever... > Greg Caetano > HP ISS Linux Virtualization Solutions Engineering > Chicago, IL > greg.caetano at hp.com > Red Hat Certified Engineer > RHCE#805007310328754 > > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Bryan Kearney > Sent: Wednesday, March 05, 2008 1:33 PM > To: kickstart-list at redhat.com > Subject: Resend: Anaconda modifying network files > > I am using cobbler/koan to generate xen disk images. After a successful > install, if I mount the image I see infomation in various networking > files which denote the information about the network I built the image > on. Specifically: > > > /etc/hosts has the ip address and dhcp fqdn of the image when anaconda > was running. > /etc/sysconfig/network has the fgdn of the image when anaconda was running. > /etc/sysconfig/network-scripts/ifcfg-eth0 has the MAC address given out > by xen. > > Is there a way to configure anaconda to not mutate these files? > > Thanks! > > -- bk > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From bkearney at redhat.com Wed Mar 5 20:36:38 2008 From: bkearney at redhat.com (Bryan Kearney) Date: Wed, 05 Mar 2008 15:36:38 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CEF884.7040700@redhat.com> References: <47CEF581.8000208@redhat.com> <47CEF884.7040700@redhat.com> Message-ID: <47CF0456.2010600@redhat.com> Michael DeHaan wrote: > As far as I'm aware the "appliancey" way to reuse disk images is to > login in and fix all this once you fire the machines up. This is > something you could possibly > make firstboot do. So.. each boot would need to do this, since in theory the mac address could change between launches. > > The approach we took with Virt-Factory is to not use disk images, but > start the install fresh every time, and use config management to push > changes down. In a controlled environment, or once koan takes over the world, I would agree. However... this will not work in all cases (e.g. demo applliances, or operating systems with no public tree) From bkearney at redhat.com Wed Mar 5 20:42:09 2008 From: bkearney at redhat.com (Bryan Kearney) Date: Wed, 05 Mar 2008 15:42:09 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CEFDFA.9000603@redhat.com> References: <47CEF581.8000208@redhat.com> <47CEFDFA.9000603@redhat.com> Message-ID: <47CF05A1.9080208@redhat.com> Good thing I am clever :) -- bk Michael DeHaan wrote: > Caetano, Greg wrote: >> Bryan: >> >> I take care of removing all those "specifics" as part of the %post of >> a kickstart >> >> >> > So in that case, you still have to run system-config-network at > firstboot (or equivalent)? That seems reasonable -- you may be > handing something to someone with a static IP setup, you don't know. > Rethinking it, if you clear things out in %post and also feed it network > details in the libvirt XML, maybe that's good enough... > > Still, I'm not sure of a way to do that fully automatically and get them > all unique MACs without being rather clever... > > >> Greg Caetano >> HP ISS Linux Virtualization Solutions Engineering >> Chicago, IL >> greg.caetano at hp.com >> Red Hat Certified Engineer >> RHCE#805007310328754 >> >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Bryan Kearney >> Sent: Wednesday, March 05, 2008 1:33 PM >> To: kickstart-list at redhat.com >> Subject: Resend: Anaconda modifying network files >> >> I am using cobbler/koan to generate xen disk images. After a successful >> install, if I mount the image I see infomation in various networking >> files which denote the information about the network I built the image >> on. Specifically: >> >> >> /etc/hosts has the ip address and dhcp fqdn of the image when anaconda >> was running. >> /etc/sysconfig/network has the fgdn of the image when anaconda was >> running. >> /etc/sysconfig/network-scripts/ifcfg-eth0 has the MAC address given out >> by xen. >> >> Is there a way to configure anaconda to not mutate these files? >> >> Thanks! >> >> -- bk >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From mdehaan at redhat.com Wed Mar 5 20:45:43 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Mar 2008 15:45:43 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CF0456.2010600@redhat.com> References: <47CEF581.8000208@redhat.com> <47CEF884.7040700@redhat.com> <47CF0456.2010600@redhat.com> Message-ID: <47CF0677.4010201@redhat.com> Bryan Kearney wrote: > > > Michael DeHaan wrote: >> As far as I'm aware the "appliancey" way to reuse disk images is to >> login in and fix all this once you fire the machines up. This is >> something you could possibly >> make firstboot do. > > So.. each boot would need to do this, since in theory the mac address > could change between launches. I don't think so. It should persist with the VM (if you keep the libvirt config with it)... > >> >> The approach we took with Virt-Factory is to not use disk images, but >> start the install fresh every time, and use config management to push >> changes down. > > In a controlled environment, or once koan takes over the world, I > would agree. However... this will not work in all cases (e.g. demo > applliances, or operating systems with no public tree) No argument there. In demo-land, you don't care about MAC collision, and randomization should be sufficient. I think. No public tree? You can still distribute the install server. > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From mdehaan at redhat.com Wed Mar 5 20:48:23 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Mar 2008 15:48:23 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CF0456.2010600@redhat.com> References: <47CEF581.8000208@redhat.com> <47CEF884.7040700@redhat.com> <47CF0456.2010600@redhat.com> Message-ID: <47CF0717.6020608@redhat.com> Bryan Kearney wrote: > > > Michael DeHaan wrote: >> As far as I'm aware the "appliancey" way to reuse disk images is to >> login in and fix all this once you fire the machines up. This is >> something you could possibly >> make firstboot do. > > So.. each boot would need to do this, since in theory the mac address > could change between launches. (Hmm, thought I replied to this, if so, apologize for the double-post). No, the MAC address should travel with the VM config. > >> >> The approach we took with Virt-Factory is to not use disk images, but >> start the install fresh every time, and use config management to push >> changes down. > > In a controlled environment, or once koan takes over the world, I > would agree. However... this will not work in all cases (e.g. demo > applliances, or operating systems with no public tree) In demo-ware, you shouldn't care about the MAC being persistant and DHCP should be your friend. As for no public tree cases, well, we're talking images, aren't we? If so, it's not /really/ kickstart-list anymore... you have the opportunity to log in, fix them up right, before redistribution, no? > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From bkearney at redhat.com Wed Mar 5 20:49:25 2008 From: bkearney at redhat.com (Bryan Kearney) Date: Wed, 05 Mar 2008 15:49:25 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CF0717.6020608@redhat.com> References: <47CEF581.8000208@redhat.com> <47CEF884.7040700@redhat.com> <47CF0456.2010600@redhat.com> <47CF0717.6020608@redhat.com> Message-ID: <47CF0755.6040200@redhat.com> Michael DeHaan wrote: > Bryan Kearney wrote: > (Hmm, thought I replied to this, if so, apologize for the double-post). > > No, the MAC address should travel with the VM config. Agreee.. it _should_ but may not. >> > In demo-ware, you shouldn't care about the MAC being persistant and DHCP > should be your friend. > As for no public tree cases, well, we're talking images, aren't we? If > so, it's not /really/ kickstart-list anymore... you have the > opportunity to log in, fix them up right, before redistribution, no? Agree.. outside of the pervue of kickstart. I was hoping I could do this via kickstart. I will hack it on my own. Maybe "koan --no-network-data".... -- bk From mdehaan at redhat.com Wed Mar 5 22:33:40 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Mar 2008 17:33:40 -0500 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CF0755.6040200@redhat.com> References: <47CEF581.8000208@redhat.com> <47CEF884.7040700@redhat.com> <47CF0456.2010600@redhat.com> <47CF0717.6020608@redhat.com> <47CF0755.6040200@redhat.com> Message-ID: <47CF1FC4.3080903@redhat.com> > Agree.. outside of the pervue of kickstart. I was hoping I could do > this via kickstart. I will hack it on my own. Maybe "koan > --no-network-data".... > > -- bk That could be dealt with with removal of those files in post, plus some extra glue on top of koan. Probably doesn't belong in koan itself as I'm trying to limit the number of options included there to keep the emphasis on central management. Plus, it's basically meant for a datacenter net-install tool anyway. You deploy with that and use other tools for migrations. A build system it's not... that's sort of why there was some talk about making Anaconda be able to be run stand-alone and spit out VM images directly. No details on the status of that... --Michael From gene.huft at wachovia.com Thu Mar 6 02:43:07 2008 From: gene.huft at wachovia.com (gene.huft at wachovia.com) Date: Wed, 5 Mar 2008 21:43:07 -0500 Subject: Kickstarting to VM via ISO image Message-ID: I got this problem solved - I hadn't realized the .discinfo file didn't get copied off media during my rsync ..anyway, I restored it & it makes it through the install. Now, however, it will not execute the %post section - in which I call another script that I try to copy off from a directory on virtual CDROM (ISO) that I'm building from... my question is: during %post, how is the cdrom mounted, another words, do I have to mount the cdrom in the script something like mount -t iso9660 -o ro /dev/cdrom /media ? Or do I use a %post --chroot in some fashion? Gene L. Huft Distributed Systems Engineering ------------------------------------------------ 512-310-4104 512-673-0717 (Cell) Gene.Huft at Wachovia.com ----- Forwarded by Gene Huft/IT/WACH on 03/05/2008 09:21 PM ----- Gene Huft/IT/WACH 03/05/2008 02:43 PM To kickstart-list at redhat.com cc Subject Kickstarting to VM via ISO image Kickstart Gurus, I have a fully contained ISO image of the original RHEL5.1 32-bit source tree, along with some additional RPMs & a custom ks.cfg file. The goal is to totally automate the build to a VMware VM via mapping the ISO file to a virtual CDROM. I have modified the default isolinux/isolinux.cfg file as follows: default xander prompt 1 timeout 600 display boot.msg F1 boot.msg F2 options.msg F3 general.msg F4 param.msg F5 rescue.msg label xander kernel vmlinuz append initrd=initrd.img ramdisk_size=8192 ks=cdrom:/ks.cfg label text kernel vmlinuz append initrd=initrd.img text label ks kernel vmlinuz append ks initrd=initrd.img label local localboot 1 label memtest86 kernel memtest append - In my kickstart file, I of cource specify "cdrom". The ISO image will boot the VM into RedHat's initial screen, but when hit return or let it time out it, it complains that it cannot find the install media on any cdrom. I also tried to put the kickstart file into the initrd.img by uncompressing image, copying ks.cfg into it, modifying isolinux.cfg to: append initrd=initrd.img ramdisk_size=8192 ks=file:/ks.cfg Then recompressed it, moved it into proper place & re-generated ISO image. Again, the ISO will boot the VM to initial RH screen, but soon there after get the "cd not found" error. Is this just a VMware thing? My coworker has no trouble with SuSE builds working....am I overlooking something simple? Thanks in advance. Gene -------------- next part -------------- An HTML attachment was scrubbed... URL: From debian at herakles.homelinux.org Thu Mar 6 03:51:24 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Thu, 06 Mar 2008 12:51:24 +0900 Subject: Kickstarting to VM via ISO image In-Reply-To: References: Message-ID: <47CF6A3C.9050608@herakles.homelinux.org> gene.huft at wachovia.com wrote: > I got this problem solved - I hadn't realized the .discinfo file didn't > get copied off media during my rsync ..anyway, I restored it & it makes it > through the install. > > Now, however, it will not execute the %post section - in which I call > another script that I try to copy off from a directory on virtual CDROM > (ISO) that I'm building from... my question is: during %post, how is the > cdrom mounted, another words, do I have to mount the cdrom in the script > something like mount -t iso9660 -o ro /dev/cdrom /media ? Or do I use a > %post --chroot in some fashion? > > Gene L. Huft > Distributed Systems Engineering > ------------------------------------------------ > 512-310-4104 > 512-673-0717 (Cell) > Gene.Huft at Wachovia.com > ----- Forwarded by Gene Huft/IT/WACH on 03/05/2008 09:21 PM ----- > > Gene Huft/IT/WACH > 03/05/2008 02:43 PM > > To > kickstart-list at redhat.com > cc > > Subject > Kickstarting to VM via ISO image > > > > > > Kickstart Gurus, > > I have a fully contained ISO image of the original RHEL5.1 32-bit source > tree, along with some additional RPMs & a custom ks.cfg file. The goal is > to totally automate the build to a VMware VM via mapping the ISO file to a > virtual CDROM. > > I have modified the default isolinux/isolinux.cfg file as follows: > > default xander > prompt 1 > timeout 600 > display boot.msg > F1 boot.msg > F2 options.msg > F3 general.msg > F4 param.msg > F5 rescue.msg > label xander > kernel vmlinuz > append initrd=initrd.img ramdisk_size=8192 ks=cdrom:/ks.cfg > label text > kernel vmlinuz > append initrd=initrd.img text > label ks > kernel vmlinuz > append ks initrd=initrd.img > label local > localboot 1 > label memtest86 > kernel memtest > append - > > > In my kickstart file, I of cource specify "cdrom". The ISO image will > boot the VM into RedHat's initial screen, but when hit return or let it > time out it, it complains that it cannot find the install media on any > cdrom. > > I also tried to put the kickstart file into the initrd.img by > uncompressing image, copying ks.cfg into it, modifying isolinux.cfg to: > > append initrd=initrd.img ramdisk_size=8192 ks=file:/ks.cfg > > Then recompressed it, moved it into proper place & re-generated ISO image. > Again, the ISO will boot the VM to initial RH screen, but soon there after > get the "cd not found" error. > Is this just a VMware thing? My coworker has no trouble with SuSE builds > working....am I overlooking something simple? > > I rarely use CD/DVD for installs, esp ks installs, so add salt. I think the CD is directly accessible if your %post is not chroot. Have a look at vc1 during any install to see what's there. I think the CD is not directly accessible if your %post is chroot. It's possible to mount it, but doing so requires some effort, or basic assumptions relying on the way you work. Simplest, is to copy from a non-chroot %post and use from a chroot %post. It's legitimate to have more than one (and for them to be in different languages). -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From gene.huft at wachovia.com Thu Mar 6 17:11:28 2008 From: gene.huft at wachovia.com (gene.huft at wachovia.com) Date: Thu, 6 Mar 2008 12:11:28 -0500 Subject: %post problems via CDROM to VM Message-ID: My shell code in the %post section of an all-CDROM-contained kickstart does not work the same as it network build. I'm still troubleshooting, but it appears I have a severly restricted shell environment during this phase. (Why would it be different from network install? same code?) Is it possible the %post section really does not kick in until after it's first reboot?? If so, that may be the difference, because I must manually, unmount the virtual CDROM after the first boot, or it will simply boot from the CDROM again & repeat all over. However, I do see the message that says it's doing post installation, but it's definately not working as expected. My new RHEL 5.1 32-bit VM is indeed installed, but with NONE of the post script stuff. I set up some echo statements to help me troubleshoot...note that it does not create a mount point like I tell it, and certainly does not copy off script: ------------------------------------------------------------------------------------------------- %post --interpreter /bin/bash # echo "Entering into %POST section" >> /opt/POST.txt #export METHOD=NETWORK export METHOD=CDROM echo "METHOD=${METHOD}" >> /opt/POST.txt export VER=RHEL5_32 echo "VERSION=${VER}" >> /opt/POST.txt ##################### mkdir -p /cdrom /bin/mount -t iso96660 -o ro /dev/cdrom /cdrom # mount >> /opt/POST.txt echo "Copy and execute post configure script." >> /opt/POST.txt # cp /cdrom/third-party/rhel_post.bash /opt/rhel_post.bash ls -la /opt >> /opt/POST.txt [ -f /opt/rhel_post.bash ] && cd /opt;./rhel_post.bash||\ echo "Trying just copy from third-party...not /cdrom/third-party" >> /opt/POST.txt cp third-party/rhel_pst.bash /opt/rhel_post.bash ls -al /opt >> /opt/POST.txt ------------------------------------------------------------------------------------------------- What's weird, is that it produced an /opt/POST.txt? file and a /opt/POST.txt file: cat POST.txt? Entering into %POST section METHOD=CDROM VERSION=RHEL5_32 /dev/root on / type ext3 (ro) Copy and execute post configure script. total 24 drwxr-xr-x 2 root root 4096 Mar 6 09:19. drwxr-xr-x 24 root root 4096 Mar 6 09:19.. -rw-r--r-- 1 root root 121 Mar 6 09:19 POST.txt? total 32 drwxr-xr-x 2 root root 4096 Mar 6 09:19. drwxr-xr-x 24 root root 4096 Mar 6 09:19.. -rw-r--r-- 1 root root 61 Mar 6 09:19 POST.txt -rw-r--r-- 1 root root 271 Mar 6 09:19 POST.txt? cat POST.txt Tyring just copy from third-party..not /cdrom/third-party Any ideas? Gene -------------- next part -------------- An HTML attachment was scrubbed... URL: From Floydsmith at aol.com Fri Mar 7 13:42:03 2008 From: Floydsmith at aol.com (Floydsmith at aol.com) Date: Fri, 7 Mar 2008 08:42:03 EST Subject: %post problems via CDROM to VM Message-ID: You have one too many 6(s) in line: /bin/mount -t iso96660 -o ro /dev/cdrom /cdrom Floyd, ************** It's Tax Time! Get tips, forms, and advice on AOL Money & Finance. (http://money.aol.com/tax?NCID=aolprf00030000000001) -------------- next part -------------- An HTML attachment was scrubbed... URL: From Gerrard.Geldenhuis at datacash.com Fri Mar 7 14:09:54 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Fri, 7 Mar 2008 14:09:54 -0000 Subject: Assigning machinename to volumegroup Message-ID: Hi I am trying to give my volume group a more specific name in the following way: install text network --bootproto dhcp url --url http://satellite.example.com/ty/rAhbfOeG lang en_GB langsupport --default en_GB en_GB keyboard uk mouse none device scsi cciss zerombr yes clearpart --all bootloader --location mbr timezone --utc Europe/London auth --enablemd5 --enableshadow selinux --disabled reboot firewall --disabled skipx %packages --resolvedeps -telnet ntp -bluez-utils -anacron @ Core -bluez-gnome %pre HOSTNAME=$(grep HOSTNAME /tmp/netinfo|cut -d= -f2|cut -d. -f1) vgn=vg-$HOSTNAME cat << EOF > /tmp/partinfo partition /boot --fstype=ext3 --size=200 partition pv.01 --size=1000 --grow partition swap --size=1000 --maxsize=2000 volgroup $vgn pv.01 logvol / --vgname=$vgn --name=rootvol --size=5120 logvol /home --vgname=$vgn --name=homevol --size=1024 logvol /tmp --vgname=$vgn --name=tmpvol --size=1024 logvol /var --vgname=$vgn --name=varvol --size=1536 logvol /var/log --vgname=$vgn --name=logvol --size=30720 logvol /var/tmp --vgname=$vgn --name=vartmpvol --size=1024 EOF However, I am still getting prompted for the partitioning details. The script generates the file and the contents looks fine. I am not sure what I am doing wrong. I am using Satellite 5.0. I have removed the partitioning details from the satellite interface and added it in the pre-script. I assumed that anakonda would then as option B look into /tmp/partinfo, but that does not seem to be the case. Any pointers in the right direction would be appreciated. Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From Pablo.Iranzo at redhat.com Fri Mar 7 14:54:09 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Fri, 07 Mar 2008 15:54:09 +0100 Subject: Assigning machinename to volumegroup In-Reply-To: References: Message-ID: <1204901649.3964.7.camel@iranzo.users.redhat.com> Echo each line to some file in /tmp then, %include that file Regards Pablo El vie, 07-03-2008 a las 14:09 +0000, Gerrard Geldenhuis escribi?: > Hi > > I am trying to give my volume group a more specific name in the > following way: > > > > install > text > network --bootproto dhcp > url --url http://satellite.example.com/ty/rAhbfOeG > lang en_GB > langsupport --default en_GB en_GB > keyboard uk > mouse none > device scsi cciss > zerombr yes > clearpart --all > bootloader --location mbr > timezone --utc Europe/London > auth --enablemd5 --enableshadow > selinux --disabled > reboot > firewall --disabled > skipx > > %packages --resolvedeps > > -telnet > ntp > -bluez-utils > -anacron > @ Core > -bluez-gnome > > %pre > HOSTNAME=$(grep HOSTNAME /tmp/netinfo|cut -d= -f2|cut -d. -f1) > vgn=vg-$HOSTNAME > > cat << EOF > /tmp/partinfo > partition /boot --fstype=ext3 --size=200 > partition pv.01 --size=1000 --grow > partition swap --size=1000 --maxsize=2000 > > volgroup $vgn pv.01 > > logvol / --vgname=$vgn --name=rootvol --size=5120 > logvol /home --vgname=$vgn --name=homevol --size=1024 > logvol /tmp --vgname=$vgn --name=tmpvol --size=1024 > logvol /var --vgname=$vgn --name=varvol --size=1536 > logvol /var/log --vgname=$vgn --name=logvol --size=30720 > logvol /var/tmp --vgname=$vgn --name=vartmpvol --size=1024 > > EOF > > > However, I am still getting prompted for the partitioning details. The > script generates the file and the contents looks fine. > > > > I am not sure what I am doing wrong. > > > > I am using Satellite 5.0. I have removed the partitioning details from > the satellite interface and added it in the pre-script. I assumed that > anakonda would then as option B look into /tmp/partinfo, but that does > not seem to be the case. > > > > Any pointers in the right direction would be appreciated. > > > > Regards > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Gerrard.Geldenhuis at datacash.com Fri Mar 7 15:03:36 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Fri, 7 Mar 2008 15:03:36 -0000 Subject: Assigning machinename to volumegroup In-Reply-To: <1204901649.3964.7.camel@iranzo.users.redhat.com> References: <1204901649.3964.7.camel@iranzo.users.redhat.com> Message-ID: Ok, any idea where I do includes for kickstarts generate via Satellite. I tried it at the end of my post script but it did not do the trick... Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > Sent: 07 March 2008 14:54 > To: Discussion list about Kickstart > Subject: Re: Assigning machinename to volumegroup > > Echo each line to some file in /tmp then, %include that file > > Regards > Pablo > > El vie, 07-03-2008 a las 14:09 +0000, Gerrard Geldenhuis escribi?: > > Hi > > > > I am trying to give my volume group a more specific name in the > > following way: > > > > > > > > install > > text > > network --bootproto dhcp > > url --url http://satellite.example.com/ty/rAhbfOeG > > lang en_GB > > langsupport --default en_GB en_GB > > keyboard uk > > mouse none > > device scsi cciss > > zerombr yes > > clearpart --all > > bootloader --location mbr > > timezone --utc Europe/London > > auth --enablemd5 --enableshadow > > selinux --disabled > > reboot > > firewall --disabled > > skipx > > > > %packages --resolvedeps > > > > -telnet > > ntp > > -bluez-utils > > -anacron > > @ Core > > -bluez-gnome > > > > %pre > > HOSTNAME=$(grep HOSTNAME /tmp/netinfo|cut -d= -f2|cut -d. -f1) > > vgn=vg-$HOSTNAME > > > > cat << EOF > /tmp/partinfo > > partition /boot --fstype=ext3 --size=200 > > partition pv.01 --size=1000 --grow > > partition swap --size=1000 --maxsize=2000 > > > > volgroup $vgn pv.01 > > > > logvol / --vgname=$vgn --name=rootvol --size=5120 > > logvol /home --vgname=$vgn --name=homevol --size=1024 > > logvol /tmp --vgname=$vgn --name=tmpvol --size=1024 > > logvol /var --vgname=$vgn --name=varvol --size=1536 > > logvol /var/log --vgname=$vgn --name=logvol --size=30720 > > logvol /var/tmp --vgname=$vgn --name=vartmpvol --size=1024 > > > > EOF > > > > > > However, I am still getting prompted for the partitioning details. The > > script generates the file and the contents looks fine. > > > > > > > > I am not sure what I am doing wrong. > > > > > > > > I am using Satellite 5.0. I have removed the partitioning details from > > the satellite interface and added it in the pre-script. I assumed that > > anakonda would then as option B look into /tmp/partinfo, but that does > > not seem to be the case. > > > > > > > > Any pointers in the right direction would be appreciated. > > > > > > > > Regards > > > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > -- > > Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) > RHCE/Global Profesional Services Consultant Spain > Phone: +34 645 01 01 49 (CET/CEST) > GnuPG KeyID: 0xFAD3CF0D > --- > Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, > 28016 Madrid, Spain > Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain > Inscrita en el Reg. Mercantil de Madrid - C.I.F. B82657941 From Pablo.Iranzo at redhat.com Fri Mar 7 15:11:19 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Fri, 07 Mar 2008 16:11:19 +0100 Subject: Assigning machinename to volumegroup In-Reply-To: References: <1204901649.3964.7.camel@iranzo.users.redhat.com> Message-ID: <1204902679.3964.13.camel@iranzo.users.redhat.com> You need to put it near the place where right now is the partition section, just replace the whole for your %include, and after %packages, create a %pre taking care of this. Regards Pablo El vie, 07-03-2008 a las 15:03 +0000, Gerrard Geldenhuis escribi?: > Ok, any idea where I do includes for kickstarts generate via Satellite. > > I tried it at the end of my post script but it did not do the trick... > > Regards > > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > > Sent: 07 March 2008 14:54 > > To: Discussion list about Kickstart > > Subject: Re: Assigning machinename to volumegroup > > > > Echo each line to some file in /tmp then, %include that file > > > > Regards > > Pablo > > > > El vie, 07-03-2008 a las 14:09 +0000, Gerrard Geldenhuis escribi?: > > > Hi > > > > > > I am trying to give my volume group a more specific name in the > > > following way: > > > > > > > > > > > > install > > > text > > > network --bootproto dhcp > > > url --url http://satellite.example.com/ty/rAhbfOeG > > > lang en_GB > > > langsupport --default en_GB en_GB > > > keyboard uk > > > mouse none > > > device scsi cciss > > > zerombr yes > > > clearpart --all > > > bootloader --location mbr > > > timezone --utc Europe/London > > > auth --enablemd5 --enableshadow > > > selinux --disabled > > > reboot > > > firewall --disabled > > > skipx > > > > > > %packages --resolvedeps > > > > > > -telnet > > > ntp > > > -bluez-utils > > > -anacron > > > @ Core > > > -bluez-gnome > > > > > > %pre > > > HOSTNAME=$(grep HOSTNAME /tmp/netinfo|cut -d= -f2|cut -d. -f1) > > > vgn=vg-$HOSTNAME > > > > > > cat << EOF > /tmp/partinfo > > > partition /boot --fstype=ext3 --size=200 > > > partition pv.01 --size=1000 --grow > > > partition swap --size=1000 --maxsize=2000 > > > > > > volgroup $vgn pv.01 > > > > > > logvol / --vgname=$vgn --name=rootvol --size=5120 > > > logvol /home --vgname=$vgn --name=homevol --size=1024 > > > logvol /tmp --vgname=$vgn --name=tmpvol --size=1024 > > > logvol /var --vgname=$vgn --name=varvol --size=1536 > > > logvol /var/log --vgname=$vgn --name=logvol --size=30720 > > > logvol /var/tmp --vgname=$vgn --name=vartmpvol --size=1024 > > > > > > EOF > > > > > > > > > However, I am still getting prompted for the partitioning details. The > > > script generates the file and the contents looks fine. > > > > > > > > > > > > I am not sure what I am doing wrong. > > > > > > > > > > > > I am using Satellite 5.0. I have removed the partitioning details from > > > the satellite interface and added it in the pre-script. I assumed that > > > anakonda would then as option B look into /tmp/partinfo, but that does > > > not seem to be the case. > > > > > > > > > > > > Any pointers in the right direction would be appreciated. > > > > > > > > > > > > Regards > > > > > > > > > _______________________________________________ > > > Kickstart-list mailing list > > > Kickstart-list at redhat.com > > > https://www.redhat.com/mailman/listinfo/kickstart-list > > -- > > > > Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) > > RHCE/Global Profesional Services Consultant Spain > > Phone: +34 645 01 01 49 (CET/CEST) > > GnuPG KeyID: 0xFAD3CF0D > > --- > > Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, > > 28016 Madrid, Spain > > Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain > > Inscrita en el Reg. Mercantil de Madrid - C.I.F. B82657941 > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Gerrard.Geldenhuis at datacash.com Fri Mar 7 15:56:51 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Fri, 7 Mar 2008 15:56:51 -0000 Subject: Assigning machinename to volumegroup In-Reply-To: <1204902679.3964.13.camel@iranzo.users.redhat.com> References: <1204901649.3964.7.camel@iranzo.users.redhat.com> <1204902679.3964.13.camel@iranzo.users.redhat.com> Message-ID: Brilliant, obvious when you think about it. ;-) It looks like it does not like the "-" character... vg-hostname does not work but vgxx does work. I will test it a bit more on Monday... Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > Sent: 07 March 2008 15:11 > To: Discussion list about Kickstart > Subject: RE: Assigning machinename to volumegroup > > You need to put it near the place where right now is the partition > section, just replace the whole for your %include, and after %packages, > create a %pre taking care of this. > > Regards > Pablo > From Pablo.Iranzo at redhat.com Fri Mar 7 16:02:09 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Fri, 07 Mar 2008 17:02:09 +0100 Subject: Assigning machinename to volumegroup In-Reply-To: References: <1204901649.3964.7.camel@iranzo.users.redhat.com> <1204902679.3964.13.camel@iranzo.users.redhat.com> Message-ID: <1204905729.3964.24.camel@iranzo.users.redhat.com> You're creating your partinfo with << EOF Try with echo "logvol / --vgname=$vgn --name=rootvol --size=5120" >> /tmp/partinfo echo does replace $name for its value, not sure with << EOF. At my system I've "$ORG"-root, $ORG-Swap etc for LV's Regards Pablo El vie, 07-03-2008 a las 15:56 +0000, Gerrard Geldenhuis escribi?: > Brilliant, obvious when you think about it. ;-) > > It looks like it does not like the "-" character... > vg-hostname does not work but vgxx does work. I will test it a bit more on Monday... > > Regards > > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > > bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez > > Sent: 07 March 2008 15:11 > > To: Discussion list about Kickstart > > Subject: RE: Assigning machinename to volumegroup > > > > You need to put it near the place where right now is the partition > > section, just replace the whole for your %include, and after %packages, > > create a %pre taking care of this. > > > > Regards > > Pablo > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Pablo.Iranzo at redhat.com Fri Mar 7 21:18:59 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Fri, 7 Mar 2008 22:18:59 +0100 (CET) Subject: Best practices for partitioning hdds? Message-ID: Hi I'm trying to setup a ks profile which must work for both new and "used" hard drives. The idea is to have one layout that could be deployed on "non-compliant" disks, and to keep the one existing on "compliant" ones. My way to workarround this could be: Examine sfdisk -l /dev/firstharddrive If part order and types are "compliant", then keep them, and write a /tmp/partinfo with "--onpart". This second option could arise problems I've encountered (not tested on RHEL/CentOS 5.1) regarding having an already defined VG with LV's I don't want anymore as I'm reinstalling. (Which I could bypass zeroing the LVM partition) If not, write a new layout to /tmp/partinfo and include it. Have you dealt with this in the past? something you could provide about what to do and what to avoid? Regards Pablo -- Pablo Iranzo G?mez (http://Alufis35.uv.es/~iranzo/) (PGPKey Available on http://www.uv.es/~iranzop/PGPKey.pgp) -- Postulado de Boling sobre la Ley de Murphy: Si se encuentra bien, no se preocupe. Se le pasar? From mdehaan at redhat.com Fri Mar 7 21:33:55 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 07 Mar 2008 16:33:55 -0500 Subject: Best practices for partitioning hdds? In-Reply-To: References: Message-ID: <47D1B4C3.6060702@redhat.com> Pablo Iranzo G?mez wrote: > Hi > I'm trying to setup a ks profile which must work for both new and > "used" hard drives. > > The idea is to have one layout that could be deployed on > "non-compliant" disks, and to keep the one existing on "compliant" ones. > > My way to workarround this could be: > > Examine sfdisk -l /dev/firstharddrive > If part order and types are "compliant", then keep them, and write > a /tmp/partinfo with "--onpart". > > This second option could arise problems I've encountered (not > tested on RHEL/CentOS 5.1) regarding having an already defined VG with > LV's I don't want anymore as I'm reinstalling. (Which I could bypass > zeroing the LVM partition) > > If not, write a new layout to /tmp/partinfo and include it. > > Have you dealt with this in the past? something you could provide > about what to do and what to avoid? > > Regards > Pablo > > > > Can you clarify what "complaint" and "non-compliant" mean? From Pablo.Iranzo at redhat.com Fri Mar 7 21:45:05 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Fri, 7 Mar 2008 22:45:05 +0100 (CET) Subject: Best practices for partitioning hdds? In-Reply-To: <47D1B4C3.6060702@redhat.com> References: <47D1B4C3.6060702@redhat.com> Message-ID: On Fri, 7 Mar 2008, Michael DeHaan wrote: > Pablo Iranzo G?mez wrote: > > The idea is to have one layout that could be deployed on > > "non-compliant" disks, and to keep the one existing on "compliant" ones. > > Can you clarify what "complaint" and "non-compliant" mean? Sorry, I mean: I want one part for /boot, one for LVM, and one for vfat, having that schema would be considered "compliant" if part sizes are at least a min size, if not, and if drive is bigger, trash the disk and repartition with that schema. Thanks Pablo From dwight at supercomputer.org Sun Mar 9 00:00:15 2008 From: dwight at supercomputer.org (dwight at supercomputer.org) Date: Sat, 8 Mar 2008 16:00:15 -0800 Subject: Resend: Anaconda modifying network files In-Reply-To: <47CF05A1.9080208@redhat.com> References: <47CEF581.8000208@redhat.com> <47CEFDFA.9000603@redhat.com> <47CF05A1.9080208@redhat.com> Message-ID: <200803081600.15892.dwight@supercomputer.org> One thing to watch out for is kudzu. I ran into a similar situation a couple of weeks ago, with kudzu overwriting the ifcfg files and changing them away from what the %post section was doing, right after the %post section had completed, and before the system had rebooted. That is, kudzu will apparently do some final manipulations after the %post section has run. There doesn't seem to be anything one can do about this either. Note that kudzu is a fundamental part of anaconda/kickstart. You simply can't remove it. RedHat defines an absolute minimal installation as consisting of kudzu (and not coreutils + kernel, which one would expect). One can get around this limitation by creating the equivalent of a .firstboot in /etc/rc.d/rc.sysinit, which corrects things when one first boots up. Hope that helps. -dwight- On Wednesday 05 March 2008 12:42:09 pm Bryan Kearney wrote: > Good thing I am clever :) > > -- bk > > Michael DeHaan wrote: > > Caetano, Greg wrote: > >> Bryan: > >> > >> I take care of removing all those "specifics" as part of the > >> %post of a kickstart > > > > So in that case, you still have to run system-config-network at > > firstboot (or equivalent)? That seems reasonable -- you may > > be handing something to someone with a static IP setup, you > > don't know. Rethinking it, if you clear things out in %post and > > also feed it network details in the libvirt XML, maybe that's > > good enough... > > > > Still, I'm not sure of a way to do that fully automatically and > > get them all unique MACs without being rather clever... > > > >> Greg Caetano > >> HP ISS Linux Virtualization Solutions Engineering > >> Chicago, IL > >> greg.caetano at hp.com > >> Red Hat Certified Engineer > >> RHCE#805007310328754 > >> > >> -----Original Message----- > >> From: kickstart-list-bounces at redhat.com > >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Bryan > >> Kearney Sent: Wednesday, March 05, 2008 1:33 PM > >> To: kickstart-list at redhat.com > >> Subject: Resend: Anaconda modifying network files > >> > >> I am using cobbler/koan to generate xen disk images. After a > >> successful install, if I mount the image I see infomation in > >> various networking files which denote the information about the > >> network I built the image on. Specifically: > >> > >> > >> /etc/hosts has the ip address and dhcp fqdn of the image when > >> anaconda was running. > >> /etc/sysconfig/network has the fgdn of the image when anaconda > >> was running. > >> /etc/sysconfig/network-scripts/ifcfg-eth0 has the MAC address > >> given out by xen. > >> > >> Is there a way to configure anaconda to not mutate these files? > >> > >> Thanks! > >> > >> -- bk > >> > >> _______________________________________________ > >> Kickstart-list mailing list > >> Kickstart-list at redhat.com > >> https://www.redhat.com/mailman/listinfo/kickstart-list > >> > >> _______________________________________________ > >> Kickstart-list mailing list > >> Kickstart-list at redhat.com > >> https://www.redhat.com/mailman/listinfo/kickstart-list > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From debian at herakles.homelinux.org Mon Mar 10 08:08:15 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Mon, 10 Mar 2008 17:08:15 +0900 Subject: Best practices for partitioning hdds? In-Reply-To: References: <47D1B4C3.6060702@redhat.com> Message-ID: <47D4EC6F.7020101@herakles.homelinux.org> Pablo Iranzo G?mez wrote: > On Fri, 7 Mar 2008, Michael DeHaan wrote: >> Pablo Iranzo G?mez wrote: >>> The idea is to have one layout that could be deployed on >>> "non-compliant" disks, and to keep the one existing on "compliant" ones. >> Can you clarify what "complaint" and "non-compliant" mean? > > Sorry, I mean: > > I want one part for /boot, one for LVM, and one for vfat, having > that schema would be considered "compliant" if part sizes are at least a > min size, if not, and if drive is bigger, trash the disk and repartition > with that schema. "best Practice" doesn't mean a lot in this context. For a desktop I tend to go with /boot and /therest. If you want some space to share with windows, then decide how big it should be, and give the rest to Linux. How much to give to Windows depends on your requirements. If you want to share up to one DVD image, then 5 Gbytes. If you want everyone to have the same layout, nothing says you have to use the entire disk. Leaving some free allows for the occasional user who might want more space for this or that, maybe to share with Windows. If you're setting up servers, that might be different, but it still depends on your requirements. _I_ am not keen on lots of partitions on one drive, but if I were settingg up EL on a zSeries I might do things differently from an entry-level xServer. -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From Pablo.Iranzo at redhat.com Mon Mar 10 08:21:08 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 10 Mar 2008 09:21:08 +0100 Subject: Best practices for partitioning hdds? In-Reply-To: <47D4EC6F.7020101@herakles.homelinux.org> References: <47D1B4C3.6060702@redhat.com> <47D4EC6F.7020101@herakles.homelinux.org> Message-ID: <1205137268.3964.48.camel@iranzo.users.redhat.com> Sorry, I meant for checking partition layout: number or partitions, sizes, etc (something like the $(list-harddrives) buf for partitions. Thanks John Pablo El lun, 10-03-2008 a las 17:08 +0900, John Summerfield escribi?: > Pablo Iranzo G?mez wrote: > > On Fri, 7 Mar 2008, Michael DeHaan wrote: > >> Pablo Iranzo G?mez wrote: > >>> The idea is to have one layout that could be deployed on > >>> "non-compliant" disks, and to keep the one existing on "compliant" ones. > >> Can you clarify what "complaint" and "non-compliant" mean? > > > > Sorry, I mean: > > > > I want one part for /boot, one for LVM, and one for vfat, having > > that schema would be considered "compliant" if part sizes are at least a > > min size, if not, and if drive is bigger, trash the disk and repartition > > with that schema. > > "best Practice" doesn't mean a lot in this context. For a desktop I tend > to go with /boot and /therest. > If you want some space to share with windows, then decide how big it > should be, and give the rest to Linux. > > How much to give to Windows depends on your requirements. If you want to > share up to one DVD image, then 5 Gbytes. > > If you want everyone to have the same layout, nothing says you have to > use the entire disk. Leaving some free allows for the occasional user > who might want more space for this or that, maybe to share with Windows. > > If you're setting up servers, that might be different, but it still > depends on your requirements. _I_ am not keen on lots of partitions on > one drive, but if I were settingg up EL on a zSeries I might do things > differently from an entry-level xServer. > > > -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From Peter.Seidel at Micronas.com Mon Mar 10 09:01:48 2008 From: Peter.Seidel at Micronas.com (Seidel Peter) Date: Mon, 10 Mar 2008 10:01:48 +0100 Subject: 3rd Party Package Dependencies Message-ID: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> Hello, I'm currently trying to set up a RHEL5 kickstart installation for our company. Everything work fine so far, but I have a Problem with the dependencies of 3rd Party Packages(.rpm/closed source) like VMware Player or Vnc-E. The dependencies aren't set probably by the Distributors. Result is, that the Packages, which are running some shell scripts, are getting installed even before Coreutils and that goes terribly wrong. So, is there any possibility to change the Package installation order, or any other suggestions? I already tried to make a meta Package, which is an empty package that just have dependencies with coreutils and VMware, than take the Vmware Package out of the kickstart(.ks) file. But it didn't work the VMware still installed too early. Thank you all in advance Peter From jgranado at redhat.com Mon Mar 10 09:40:28 2008 From: jgranado at redhat.com (Joel Andres Granados) Date: Mon, 10 Mar 2008 10:40:28 +0100 Subject: 3rd Party Package Dependencies In-Reply-To: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> References: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> Message-ID: <47D5020C.8080802@redhat.com> Seidel Peter wrote: > Hello, > > I'm currently trying to set up a RHEL5 kickstart installation for our company. > > Everything work fine so far, but I have a Problem with the dependencies of 3rd Party Packages(.rpm/closed source) like VMware Player or Vnc-E. The dependencies aren't set probably by the Distributors. Result is, that the Packages, which are running some shell scripts, are getting installed even before Coreutils and that goes terribly wrong. > > So, is there any possibility to change the Package installation order, or any other suggestions? > > I already tried to make a meta Package, which is an empty package that just have dependencies with coreutils and VMware, than take the Vmware Package out of the kickstart(.ks) file. But it didn't work the VMware still installed too early. > > Thank you all in advance > > Peter > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list You can use kickstarts %post section: %post wget http://path/to/vmware/package.rpm wget http://path/to/vnc/package.rpm rpm -ihv *.rpm Just an idea :) -- Joel Andres Granados Red Hat / Brno, Czech Republic From Pablo.Iranzo at redhat.com Mon Mar 10 09:51:13 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 10 Mar 2008 10:51:13 +0100 Subject: 3rd Party Package Dependencies In-Reply-To: <47D5020C.8080802@redhat.com> References: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> <47D5020C.8080802@redhat.com> Message-ID: <1205142673.3964.53.camel@iranzo.users.redhat.com> Wouldn't that keep the RPM database locked by the package being installed and running %post scripts? Regards Pablo El lun, 10-03-2008 a las 10:40 +0100, Joel Andres Granados escribi?: > Seidel Peter wrote: > > Hello, > > > > I'm currently trying to set up a RHEL5 kickstart installation for our company. > > > > Everything work fine so far, but I have a Problem with the dependencies of 3rd Party Packages(.rpm/closed source) like VMware Player or Vnc-E. The dependencies aren't set probably by the Distributors. Result is, that the Packages, which are running some shell scripts, are getting installed even before Coreutils and that goes terribly wrong. > > > > So, is there any possibility to change the Package installation order, or any other suggestions? > > > > I already tried to make a meta Package, which is an empty package that just have dependencies with coreutils and VMware, than take the Vmware Package out of the kickstart(.ks) file. But it didn't work the VMware still installed too early. > > > > Thank you all in advance > > > > Peter > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > You can use kickstarts %post section: > > %post > > wget http://path/to/vmware/package.rpm > wget http://path/to/vnc/package.rpm > > rpm -ihv *.rpm > > Just an idea :) -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From jgranado at redhat.com Mon Mar 10 09:44:42 2008 From: jgranado at redhat.com (Joel Andres Granados) Date: Mon, 10 Mar 2008 10:44:42 +0100 Subject: Assigning machinename to volumegroup In-Reply-To: References: <1204901649.3964.7.camel@iranzo.users.redhat.com> <1204902679.3964.13.camel@iranzo.users.redhat.com> Message-ID: <47D5030A.1080906@redhat.com> Gerrard Geldenhuis wrote: > Brilliant, obvious when you think about it. ;-) > > It looks like it does not like the "-" character... > vg-hostname does not work but vgxx does work. I will test it a bit more on Monday... > > Regards > >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- >> bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez >> Sent: 07 March 2008 15:11 >> To: Discussion list about Kickstart >> Subject: RE: Assigning machinename to volumegroup >> >> You need to put it near the place where right now is the partition >> section, just replace the whole for your %include, and after %packages, >> create a %pre taking care of this. >> >> Regards >> Pablo >> > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list I would really not recommend the use of "-" for lvm names. LVM seems to use the "-" internally somehow and has to treat the names containing "-" in a special way. -- Joel Andres Granados Red Hat / Brno, Czech Republic From Peter.Seidel at Micronas.com Mon Mar 10 10:12:49 2008 From: Peter.Seidel at Micronas.com (Seidel Peter) Date: Mon, 10 Mar 2008 11:12:49 +0100 Subject: 3rd Party Package Dependencies References: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com><47D5020C.8080802@redhat.com> <1205142673.3964.53.camel@iranzo.users.redhat.com> Message-ID: <364C0DD5DC3BFB479C71C270784929F10540E5F5@EXCHANGE3.Micronas.com> He, to install th packages in the %post script is what we want to avoid. We had that in an early RHEL4 installation and it was "working". But now we are quite happy to got rid of it. If there won't be any other possibility, we have do it like this again. Anyway, thank you for your suggestions Peter -----Original Message----- From: kickstart-list-bounces at redhat.com on behalf of Pablo Iranzo G?mez Sent: Mon 3/10/2008 10:51 AM To: Discussion list about Kickstart Subject: Re: 3rd Party Package Dependencies Wouldn't that keep the RPM database locked by the package being installed and running %post scripts? Regards Pablo El lun, 10-03-2008 a las 10:40 +0100, Joel Andres Granados escribi?: > Seidel Peter wrote: > > Hello, > > > > I'm currently trying to set up a RHEL5 kickstart installation for our company. > > > > Everything work fine so far, but I have a Problem with the dependencies of 3rd Party Packages(.rpm/closed source) like VMware Player or Vnc-E. The dependencies aren't set probably by the Distributors. Result is, that the Packages, which are running some shell scripts, are getting installed even before Coreutils and that goes terribly wrong. > > > > So, is there any possibility to change the Package installation order, or any other suggestions? > > > > I already tried to make a meta Package, which is an empty package that just have dependencies with coreutils and VMware, than take the Vmware Package out of the kickstart(.ks) file. But it didn't work the VMware still installed too early. > > > > Thank you all in advance > > > > Peter > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > You can use kickstarts %post section: > > %post > > wget http://path/to/vmware/package.rpm > wget http://path/to/vnc/package.rpm > > rpm -ihv *.rpm > > Just an idea :) -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid - C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: winmail.dat Type: application/ms-tnef Size: 3999 bytes Desc: not available URL: From jgranado at redhat.com Mon Mar 10 10:38:44 2008 From: jgranado at redhat.com (Joel Andres Granados) Date: Mon, 10 Mar 2008 11:38:44 +0100 Subject: 3rd Party Package Dependencies In-Reply-To: <1205142673.3964.53.camel@iranzo.users.redhat.com> References: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> <47D5020C.8080802@redhat.com> <1205142673.3964.53.camel@iranzo.users.redhat.com> Message-ID: <47D50FB4.1000704@redhat.com> Pablo Iranzo G?mez wrote: > Wouldn't that keep the RPM database locked by the package being > installed and running %post scripts? AFAK the only thing you need to make sure for this approach to work is for all the dependencies to be in the system before %post. Don't know much about the internals of rpm database, but I would assume, for consistency purposes, that at install time the data base locks. So when you do the `rpm -ihv ...` the database is locked. however, This does not disallow you to install with the rpm command in the %post section. > > Regards > Pablo > > El lun, 10-03-2008 a las 10:40 +0100, Joel Andres Granados escribi?: >> Seidel Peter wrote: >>> Hello, >>> >>> I'm currently trying to set up a RHEL5 kickstart installation for our company. >>> >>> Everything work fine so far, but I have a Problem with the dependencies of 3rd Party Packages(.rpm/closed source) like VMware Player or Vnc-E. The dependencies aren't set probably by the Distributors. Result is, that the Packages, which are running some shell scripts, are getting installed even before Coreutils and that goes terribly wrong. >>> >>> So, is there any possibility to change the Package installation order, or any other suggestions? >>> >>> I already tried to make a meta Package, which is an empty package that just have dependencies with coreutils and VMware, than take the Vmware Package out of the kickstart(.ks) file. But it didn't work the VMware still installed too early. >>> >>> Thank you all in advance >>> >>> Peter >>> >>> _______________________________________________ >>> Kickstart-list mailing list >>> Kickstart-list at redhat.com >>> https://www.redhat.com/mailman/listinfo/kickstart-list >> You can use kickstarts %post section: >> >> %post >> >> wget http://path/to/vmware/package.rpm >> wget http://path/to/vnc/package.rpm >> >> rpm -ihv *.rpm >> >> Just an idea :) >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list -- Joel Andres Granados Red Hat / Brno, Czech Republic From Pablo.Iranzo at redhat.com Mon Mar 10 10:57:47 2008 From: Pablo.Iranzo at redhat.com (Pablo Iranzo =?ISO-8859-1?Q?G=F3mez?=) Date: Mon, 10 Mar 2008 11:57:47 +0100 Subject: 3rd Party Package Dependencies In-Reply-To: <47D50FB4.1000704@redhat.com> References: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> <47D5020C.8080802@redhat.com> <1205142673.3964.53.camel@iranzo.users.redhat.com> <47D50FB4.1000704@redhat.com> Message-ID: <1205146667.3964.55.camel@iranzo.users.redhat.com> Sorry, but as we were talking about virtual RPM, I asumed that you were referring to the %post in RPM, not the %post in Kickstart ;) Regards Pablo El lun, 10-03-2008 a las 11:38 +0100, Joel Andres Granados escribi?: > Pablo Iranzo G?mez wrote: > > Wouldn't that keep the RPM database locked by the package being > > installed and running %post scripts? > > AFAK the only thing you need to make sure for this approach to work is for all the dependencies to be in the system before %post. Don't know much about the internals of rpm database, but I would assume, for consistency purposes, that at install time the data base locks. So when you do the `rpm -ihv ...` the database is locked. however, This does not disallow you to install with the rpm command in the %post section. > > > > > Regards > > Pablo > > > > El lun, 10-03-2008 a las 10:40 +0100, Joel Andres Granados escribi?: > >> Seidel Peter wrote: > >>> Hello, > >>> > >>> I'm currently trying to set up a RHEL5 kickstart installation for our company. > >>> > >>> Everything work fine so far, but I have a Problem with the dependencies of 3rd Party Packages(.rpm/closed source) like VMware Player or Vnc-E. The dependencies aren't set probably by the Distributors. Result is, that the Packages, which are running some shell scripts, are getting installed even before Coreutils and that goes terribly wrong. > >>> > >>> So, is there any possibility to change the Package installation order, or any other suggestions? > >>> > >>> I already tried to make a meta Package, which is an empty package that just have dependencies with coreutils and VMware, than take the Vmware Package out of the kickstart(.ks) file. But it didn't work the VMware still installed too early. > >>> > >>> Thank you all in advance > >>> > >>> Peter > >>> > >>> _______________________________________________ > >>> Kickstart-list mailing list > >>> Kickstart-list at redhat.com > >>> https://www.redhat.com/mailman/listinfo/kickstart-list > >> You can use kickstarts %post section: > >> > >> %post > >> > >> wget http://path/to/vmware/package.rpm > >> wget http://path/to/vnc/package.rpm > >> > >> rpm -ihv *.rpm > >> > >> Just an idea :) > >> > >> ------------------------------------------------------------------------ > >> > >> _______________________________________________ > >> Kickstart-list mailing list > >> Kickstart-list at redhat.com > >> https://www.redhat.com/mailman/listinfo/kickstart-list > -- Pablo Iranzo G?mez (Pablo.Iranzo at redhat.com) RHCE/Global Profesional Services Consultant Spain Phone: +34 645 01 01 49 (CET/CEST) GnuPG KeyID: 0xFAD3CF0D --- Direcci?n Comercial: C/Jose Bardasano Baos, 9, Edif. Gorbea 3, planta 3?D, 28016 Madrid, Spain Direcci?n Registrada: Red Hat S.L., C/ Velazquez 63, Madrid 28001, Spain Inscrita en el Reg. Mercantil de Madrid ? C.I.F. B82657941 -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Esta parte del mensaje est? firmada digitalmente URL: From bkearney at redhat.com Mon Mar 10 12:13:20 2008 From: bkearney at redhat.com (Bryan Kearney) Date: Mon, 10 Mar 2008 08:13:20 -0400 Subject: Resend: Anaconda modifying network files In-Reply-To: <200803081600.15892.dwight@supercomputer.org> References: <47CEF581.8000208@redhat.com> <47CEFDFA.9000603@redhat.com> <47CF05A1.9080208@redhat.com> <200803081600.15892.dwight@supercomputer.org> Message-ID: <47D525E0.2060401@redhat.com> I finally went with a combination of a %post and firstboot (actually, every boot). Together I was able to use these to "clean" the network info. -- bk dwight at supercomputer.org wrote: > One thing to watch out for is kudzu. I ran into a similar situation a > couple of weeks ago, with kudzu overwriting the ifcfg files and > changing them away from what the %post section was doing, right > after the %post section had completed, and before the system had > rebooted. > > That is, kudzu will apparently do some final manipulations after > the %post section has run. There doesn't seem to be anything one can > do about this either. > > Note that kudzu is a fundamental part of anaconda/kickstart. You > simply can't remove it. RedHat defines an absolute minimal > installation as consisting of kudzu (and not coreutils + kernel, > which one would expect). > > One can get around this limitation by creating the equivalent of > a .firstboot in /etc/rc.d/rc.sysinit, which corrects things when one > first boots up. > > Hope that helps. > > -dwight- > > > On Wednesday 05 March 2008 12:42:09 pm Bryan Kearney wrote: >> Good thing I am clever :) >> >> -- bk >> >> Michael DeHaan wrote: >>> Caetano, Greg wrote: >>>> Bryan: >>>> >>>> I take care of removing all those "specifics" as part of the >>>> %post of a kickstart >>> So in that case, you still have to run system-config-network at >>> firstboot (or equivalent)? That seems reasonable -- you may >>> be handing something to someone with a static IP setup, you >>> don't know. Rethinking it, if you clear things out in %post and >>> also feed it network details in the libvirt XML, maybe that's >>> good enough... >>> >>> Still, I'm not sure of a way to do that fully automatically and >>> get them all unique MACs without being rather clever... >>> >>>> Greg Caetano >>>> HP ISS Linux Virtualization Solutions Engineering >>>> Chicago, IL >>>> greg.caetano at hp.com >>>> Red Hat Certified Engineer >>>> RHCE#805007310328754 >>>> >>>> -----Original Message----- >>>> From: kickstart-list-bounces at redhat.com >>>> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Bryan >>>> Kearney Sent: Wednesday, March 05, 2008 1:33 PM >>>> To: kickstart-list at redhat.com >>>> Subject: Resend: Anaconda modifying network files >>>> >>>> I am using cobbler/koan to generate xen disk images. After a >>>> successful install, if I mount the image I see infomation in >>>> various networking files which denote the information about the >>>> network I built the image on. Specifically: >>>> >>>> >>>> /etc/hosts has the ip address and dhcp fqdn of the image when >>>> anaconda was running. >>>> /etc/sysconfig/network has the fgdn of the image when anaconda >>>> was running. >>>> /etc/sysconfig/network-scripts/ifcfg-eth0 has the MAC address >>>> given out by xen. >>>> >>>> Is there a way to configure anaconda to not mutate these files? >>>> >>>> Thanks! >>>> >>>> -- bk >>>> >>>> _______________________________________________ >>>> Kickstart-list mailing list >>>> Kickstart-list at redhat.com >>>> https://www.redhat.com/mailman/listinfo/kickstart-list >>>> >>>> _______________________________________________ >>>> Kickstart-list mailing list >>>> Kickstart-list at redhat.com >>>> https://www.redhat.com/mailman/listinfo/kickstart-list >>> _______________________________________________ >>> Kickstart-list mailing list >>> Kickstart-list at redhat.com >>> https://www.redhat.com/mailman/listinfo/kickstart-list >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list > > From brian.hollifield1 at wachovia.com Mon Mar 10 12:17:25 2008 From: brian.hollifield1 at wachovia.com (brian.hollifield1 at wachovia.com) Date: Mon, 10 Mar 2008 07:17:25 -0500 Subject: Brian Hollifield is out of the office. Message-ID: I will be out of the office starting 03/07/2008 and will not return until 03/17/2008. I will be out of the office on PTO until Monday3/17. I will respond to e-mail when I return. If you need immediate assistance please contact: Mark Flom 704-715-7820 Thank You, Brian Hollifield CIBT - Platform Engineering 704-715-5674 From klandreth at theplanet.com Mon Mar 10 17:51:46 2008 From: klandreth at theplanet.com (Landreth, Kevin) Date: Mon, 10 Mar 2008 12:51:46 -0500 Subject: Assigning machinename to volumegroup In-Reply-To: <47D5030A.1080906@redhat.com> References: <1204901649.3964.7.camel@iranzo.users.redhat.com> <1204902679.3964.13.camel@iranzo.users.redhat.com> <47D5030A.1080906@redhat.com> Message-ID: <69F7A3B9C49D9D409AF0F06F8AC5076705233E46@HOUEXCH01.PLANET.LOCAL> I found good luck with using a dot (.) because it saves whitespace and is not used by device-mapper (LVM). -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Joel Andres Granados Sent: Monday, March 10, 2008 4:45 AM To: Discussion list about Kickstart Subject: Re: Assigning machinename to volumegroup Gerrard Geldenhuis wrote: > Brilliant, obvious when you think about it. ;-) > > It looks like it does not like the "-" character... > vg-hostname does not work but vgxx does work. I will test it a bit more on Monday... > > Regards > >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- >> bounces at redhat.com] On Behalf Of Pablo Iranzo G?mez >> Sent: 07 March 2008 15:11 >> To: Discussion list about Kickstart >> Subject: RE: Assigning machinename to volumegroup >> >> You need to put it near the place where right now is the partition >> section, just replace the whole for your %include, and after %packages, >> create a %pre taking care of this. >> >> Regards >> Pablo >> > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list I would really not recommend the use of "-" for lvm names. LVM seems to use the "-" internally somehow and has to treat the names containing "-" in a special way. -- Joel Andres Granados Red Hat / Brno, Czech Republic _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From rbrown at BallistaSec.com Mon Mar 10 18:06:31 2008 From: rbrown at BallistaSec.com (Rodrick Brown) Date: Mon, 10 Mar 2008 14:06:31 -0400 Subject: Unable to use nfs in %post Message-ID: I'm unable to get the following lines to work in my Kickstart profile I'm not sure what I could be doing Here is my configuration: %post # run custom scripts mkdir /a /bin/mount -t nfs 192.168.149.211:/mount/unix /a sleep 3 [[ -f /a/scripts/motd.pl ]] && /usr/bin/perl /a/scripts/motd.pl > /etc/motd # needed to connect to AD domain (cd /etc; tar xvf /a/packages/ldap_krbsettings.tar) --- Rodrick R. Brown Ballista Securities LLC 120 Wall St. Suite 2400 C: 347 702 0012 | O: 646 307 4709 http://www.ballistasec.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From kajtzu at basen.net Mon Mar 10 18:46:57 2008 From: kajtzu at basen.net (Kaj Niemi) Date: Mon, 10 Mar 2008 20:46:57 +0200 Subject: Unable to use nfs in %post In-Reply-To: References: Message-ID: <35F1582E-8516-465D-9A59-EF466AB22E1E@basen.net> Hi, You might want to put a very long sleep after the commands and either look at what happened from the command line (vty 2) or in your %post just put a practically infinite sleep, jump to vty 2, chroot /mnt/ sysimage and run the commands by hand to see what it bombs on. Good luck :) On Mar 10, 2008, at 20:06, Rodrick Brown wrote: > I?m unable to get the following lines to work in my Kickstart > profile I?m not sure what I could be doing > Here is my configuration: > > %post > # run custom scripts > mkdir /a > /bin/mount -t nfs 192.168.149.211:/mount/unix /a > sleep 3 > [[ -f /a/scripts/motd.pl ]] && /usr/bin/perl /a/scripts/motd.pl > / > etc/motd > > # needed to connect to AD domain > (cd /etc; tar xvf /a/packages/ldap_krbsettings.tar) > > --- > Rodrick R. Brown > Ballista Securities LLC > 120 Wall St. Suite 2400 > C: 347 702 0012 | O: 646 307 4709 > http://www.ballistasec.com > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list HTH Kaj -- Kaj J. Niemi +358 45 63 12000 -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 3811 bytes Desc: not available URL: From redhat-list at xdroop.com Mon Mar 10 18:59:41 2008 From: redhat-list at xdroop.com (David Mackintosh) Date: Mon, 10 Mar 2008 14:59:41 -0400 Subject: Unable to use nfs in %post In-Reply-To: <35F1582E-8516-465D-9A59-EF466AB22E1E@basen.net> References: <35F1582E-8516-465D-9A59-EF466AB22E1E@basen.net> Message-ID: <20080310185941.GA7961@xdroop.com> > On Mar 10, 2008, at 20:06, Rodrick Brown wrote: > > >I?m unable to get the following lines to work in my Kickstart > >profile I?m not sure what I could be doing > >Here is my configuration: [...] > >/bin/mount -t nfs 192.168.149.211:/mount/unix /a I've always had to do this: mount -o nolock,udp 172.20.0.1:/home /mnt/temp Without the nolock, the mount was either excessively slow or failed completely. -- /\oo/\ / /()\ \ David Mackintosh | dave at xdroop.com | http://www.xdroop.com -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From rbrown at BallistaSec.com Mon Mar 10 20:21:32 2008 From: rbrown at BallistaSec.com (Rodrick Brown) Date: Mon, 10 Mar 2008 16:21:32 -0400 Subject: Unable to use nfs in %post In-Reply-To: <35F1582E-8516-465D-9A59-EF466AB22E1E@basen.net> References: <35F1582E-8516-465D-9A59-EF466AB22E1E@basen.net> Message-ID: Everything works like a charm when using vty 2 and issuing the command manually however I'm still not able to get it working from my Kickstart profile Here is the %post section from my profile.ks %post /usr/sbin/ntpdate -su pool.ntp.org echo "Turning off some packages" /sbin/chkconfig --level 0123456 apmd off /sbin/chkconfig --level 0123456 atd off /sbin/chkconfig --level 0123456 sendmail off /sbin/chkconfig --level 0123456 lpd off /sbin/chkconfig --level 0123456 xfs off /sbin/chkconfig --level 0123456 gpm off /sbin/chkconfig --level 0123456 wine off /sbin/chkconfig --level 0123456 cups off /sbin/chkconfig --level 0123456 ntpd off /sbin/chkconfig --level 0123456 nscd off /sbin/chkconfig --level 0123456 sendmail off /sbin/chkconfig --level 0123456 iptables off /sbin/chkconfig --level 0123456 ip6tables off perl -pi.bak -e 's/#(PermitRootLogin\s+)yes/\1 no/i' /etc/ssh/sshd_config cat << EOF > /etc/sysctl.conf kernel.core_uses_pid = 1 kernel.msgmax = 65536 kernel.msgmnb = 65536 kernel.shmall = 4294967296 kernel.shmmax = 68719476736 net.core.netdev_max_backlog = 3000 net.core.rmem_default = 65536 net.core.rmem_max = 8388608 net.core.wmem_default = 65536 net.core.wmem_max = 8388608 net.ipv4.conf.default.rp_filter = 1 net.ipv4.ip_forward = 0 net.ipv4.route.flush = 1 net.ipv4.tcp_mem = 8388608 8388608 8388608 net.ipv4.tcp_moderate_rcvbuf = 1 net.ipv4.tcp_no_metrics_save =1 net.ipv4.tcp_syncookies = 1 net.ipv4.tcp_wmem = 4096 65536 8388608 EOF $yum_config_stanza $kickstart_done cat << EOF > /etc/auto.master /misc /etc/auto.misc /home /etc/auto.home /net -hosts +auto.master EOF ################################ This refuses to work what could be wrong? echo "chrooting /mnt/sysimage" <## I never get this message chroot /mnt/sysimage echo "mounting nfsfiler:/mount/unix to /a" mkdir /a /bin/mount -t nfs -o nolock,udp 192.168.149.211:/mount/unix /a sleep 15 /usr/bin/perl /a/scripts/motd.pl > /etc/motd (cd /etc; tar xvf /a/packages/ldap_krb.tar) yum update -y From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Kaj Niemi Sent: Monday, March 10, 2008 2:47 PM To: Discussion list about Kickstart Subject: Re: Unable to use nfs in %post Hi, You might want to put a very long sleep after the commands and either look at what happened from the command line (vty 2) or in your %post just put a practically infinite sleep, jump to vty 2, chroot /mnt/sysimage and run the commands by hand to see what it bombs on. Good luck :) On Mar 10, 2008, at 20:06, Rodrick Brown wrote: I'm unable to get the following lines to work in my Kickstart profile I'm not sure what I could be doing Here is my configuration: %post # run custom scripts mkdir /a /bin/mount -t nfs 192.168.149.211:/mount/unix /a sleep 3 [[ -f /a/scripts/motd.pl ]] && /usr/bin/perl /a/scripts/motd.pl > /etc/motd # needed to connect to AD domain (cd /etc; tar xvf /a/packages/ldap_krbsettings.tar) --- Rodrick R. Brown Ballista Securities LLC 120 Wall St. Suite 2400 C: 347 702 0012 | O: 646 307 4709 http://www.ballistasec.com _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list HTH Kaj -- Kaj J. Niemi +358 45 63 12000 -------------- next part -------------- An HTML attachment was scrubbed... URL: From andreas.fuchs at rtc.ch Tue Mar 11 08:19:27 2008 From: andreas.fuchs at rtc.ch (Fuchs Andreas) Date: Tue, 11 Mar 2008 09:19:27 +0100 Subject: AW: Assigning machinename to volumegroup In-Reply-To: References: Message-ID: Put the include statement in ->SystemDetails->Partitioning that way it should work Andreas ________________________________ Von: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] Im Auftrag von Gerrard Geldenhuis Gesendet: Freitag, 7. M?rz 2008 15:10 An: Discussion list about Kickstart Betreff: Assigning machinename to volumegroup Hi I am trying to give my volume group a more specific name in the following way: install text network --bootproto dhcp url --url http://satellite.example.com/ty/rAhbfOeG lang en_GB langsupport --default en_GB en_GB keyboard uk mouse none device scsi cciss zerombr yes clearpart --all bootloader --location mbr timezone --utc Europe/London auth --enablemd5 --enableshadow selinux --disabled reboot firewall --disabled skipx %packages --resolvedeps -telnet ntp -bluez-utils -anacron @ Core -bluez-gnome %pre HOSTNAME=$(grep HOSTNAME /tmp/netinfo|cut -d= -f2|cut -d. -f1) vgn=vg-$HOSTNAME cat << EOF > /tmp/partinfo partition /boot --fstype=ext3 --size=200 partition pv.01 --size=1000 --grow partition swap --size=1000 --maxsize=2000 volgroup $vgn pv.01 logvol / --vgname=$vgn --name=rootvol --size=5120 logvol /home --vgname=$vgn --name=homevol --size=1024 logvol /tmp --vgname=$vgn --name=tmpvol --size=1024 logvol /var --vgname=$vgn --name=varvol --size=1536 logvol /var/log --vgname=$vgn --name=logvol --size=30720 logvol /var/tmp --vgname=$vgn --name=vartmpvol --size=1024 EOF However, I am still getting prompted for the partitioning details. The script generates the file and the contents looks fine. I am not sure what I am doing wrong. I am using Satellite 5.0. I have removed the partitioning details from the satellite interface and added it in the pre-script. I assumed that anakonda would then as option B look into /tmp/partinfo, but that does not seem to be the case. Any pointers in the right direction would be appreciated. Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From andreas.fuchs at rtc.ch Tue Mar 11 08:23:48 2008 From: andreas.fuchs at rtc.ch (Fuchs Andreas) Date: Tue, 11 Mar 2008 09:23:48 +0100 Subject: AW: 3rd Party Package Dependencies In-Reply-To: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> References: <364C0DD5DC3BFB479C71C270784929F10540E5EF@EXCHANGE3.Micronas.com> Message-ID: install the 3rd party stuff in %post section and not in the main section Andreas -----Urspr?ngliche Nachricht----- Von: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] Im Auftrag von Seidel Peter Gesendet: Montag, 10. M?rz 2008 10:02 An: kickstart-list at redhat.com Betreff: 3rd Party Package Dependencies Hello, I'm currently trying to set up a RHEL5 kickstart installation for our company. Everything work fine so far, but I have a Problem with the dependencies of 3rd Party Packages(.rpm/closed source) like VMware Player or Vnc-E. The dependencies aren't set probably by the Distributors. Result is, that the Packages, which are running some shell scripts, are getting installed even before Coreutils and that goes terribly wrong. So, is there any possibility to change the Package installation order, or any other suggestions? I already tried to make a meta Package, which is an empty package that just have dependencies with coreutils and VMware, than take the Vmware Package out of the kickstart(.ks) file. But it didn't work the VMware still installed too early. Thank you all in advance Peter _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From i.jochim at querysoft.de Wed Mar 12 08:38:42 2008 From: i.jochim at querysoft.de (Ingo Jochim) Date: Wed, 12 Mar 2008 09:38:42 +0100 Subject: kickstart installation failure - volgroup Message-ID: <47D79692.6000400@querysoft.de> I try to install Fedora 8 and 9 using a kickstart script. It works well on FC4 and F7. But now I get the following error on the installation: AnacondaKSHandler: instance has no attribute VolGroup The line starting with volgroup causes the error. If I comment this the script works fine. ........ bootloader --location=mbr --append="rhgb quiet" clearpart --all --initlabel part /boot --fstype ext3 --size=100 part pv.11 --size=100 --grow volgroup VolGroup00 --pesize=32768 pv.11 ....... -- Checked by AVG. Version: 7.5.518 / Virus Database: 269.21.7/1324 - Release Date: 10.03.2008 19:27 From gary at mlbassoc.com Wed Mar 12 10:44:31 2008 From: gary at mlbassoc.com (Gary Thomas) Date: Wed, 12 Mar 2008 04:44:31 -0600 Subject: kickstart installation failure - volgroup In-Reply-To: <47D79692.6000400@querysoft.de> References: <47D79692.6000400@querysoft.de> Message-ID: <47D7B40F.3070705@mlbassoc.com> Ingo Jochim wrote: > I try to install Fedora 8 and 9 using a kickstart script. > > It works well on FC4 and F7. > > But now I get the following error on the installation: > > AnacondaKSHandler: instance has no attribute VolGroup > > The line starting with volgroup causes the error. If I comment this the > script works fine. Exactly which version of anaconda are you using? This was a problem a while ago, but was fixed by about 11.4.0.37 (IIRC - it most certainly works for me in 11.4.0.40) -- ------------------------------------------------------------ Gary Thomas | Consulting for the MLB Associates | Embedded world ------------------------------------------------------------ From i.jochim at querysoft.de Wed Mar 12 14:34:22 2008 From: i.jochim at querysoft.de (Ingo Jochim) Date: Wed, 12 Mar 2008 15:34:22 +0100 Subject: kickstart installation failure - volgroup In-Reply-To: <47D7B40F.3070705@mlbassoc.com> References: <47D79692.6000400@querysoft.de> <47D7B40F.3070705@mlbassoc.com> Message-ID: <47D7E9EE.9050407@querysoft.de> Gary Thomas schrieb: > Ingo Jochim wrote: >> I try to install Fedora 8 and 9 using a kickstart script. >> >> It works well on FC4 and F7. >> >> But now I get the following error on the installation: >> >> AnacondaKSHandler: instance has no attribute VolGroup >> >> The line starting with volgroup causes the error. If I comment this >> the script works fine. > > Exactly which version of anaconda are you using? This was a problem > a while ago, but was fixed by about 11.4.0.37 (IIRC - it most certainly > works for me in 11.4.0.40) > The Fedora 9 Alpha version comes with Anaconda 11.4.0.28. How can I modify the iso-image with a newer version of Anaconda? Ingo -- Checked by AVG. Version: 7.5.518 / Virus Database: 269.21.7/1324 - Release Date: 10.03.2008 19:27 From steven.hajducko at digitalinsight.com Thu Mar 13 00:35:17 2008 From: steven.hajducko at digitalinsight.com (Hajducko, Steven) Date: Wed, 12 Mar 2008 17:35:17 -0700 Subject: Wrong packages being installed. Message-ID: <21F8C5D33731F44BB9DCC396A12F9F2109067814@WLVEXM01.corp.ad.diginsite.com> So we're trying to build a very minimal build that we then build upon to form our systems. I'm trying to get rid of alot of unecessary packages that we don't require for our systems. I'm running into a problem where the kickstart is still installing some packages, even though I'm specifically telling it not too and the package not being a required dependancy of any other package I'm installing. My kickstart is very, very basic for this testing: install cdrom lang en_US langsupport --default=en_US en_US keyboard us skipx key #### text network --device eth0 --bootproto dhcp --hostname host.example.com bootloader --location=mbr --driveorder=hda --append="rhgb quiet" rootpw --iscrypted ########## authconfig --enableshadow --enablemd5 firewall --disabled timezone America/Los_Angeles selinux --disabled clearpart --all --drives=hda part /boot --fstype ext3 --size=128 --asprimary --ondrive=hda part swap --size=12288 --asprimary part pv.01 --size=100 --grow volgroup rootvg pv.01 logvol / --fstype ext3 --vgname=rootvg --size=20480 --name=rootlv reboot %packages --nobase @Core -Deployment_Guide_en-US -system-config-securitylevel-tui The problem is, even with --nobase, I still keep getting the Deployment_Guide and system-config-securitylevel-tui. The system-config package is part of the Base group, so it shouldn't be installed anyways. The Deployment Guide is part of Core, but I'm specifically telling it not to install. What am I missing here? Also, system-config-securitylevel-tui isn't being brought in because of a requirement, it's just showing up. You can remove the package as soon as the system comes up. It also doesn't show in the anaconda.log as being added to fulfill a requirement. It just suddenly shows up in the log, because slang and newt do get added due to a requirement that system-config-securitylevel has. I do see an error in the log that 'no such package system-config-securitylevel-tui', right after it tells me 'not adding Base group'. Any input to this would be greatly appreciated it. I'd rather not muck with post just to get some RPM's removed that should be taken care of via the %packages section. ..................................................................... Steven Hajducko Systems Engineer, Systems Integration w: 818.597.6443 m: 805.377.9074 This email may contain confidential and privileged material for the sole use of the intended recipient. Any review or distribution by others is strictly prohibited. If you are not the intended recipient please contact the sender and delete all copies. From gary at mlbassoc.com Thu Mar 13 11:10:03 2008 From: gary at mlbassoc.com (Gary Thomas) Date: Thu, 13 Mar 2008 05:10:03 -0600 Subject: Wrong packages being installed. In-Reply-To: <21F8C5D33731F44BB9DCC396A12F9F2109067814@WLVEXM01.corp.ad.diginsite.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067814@WLVEXM01.corp.ad.diginsite.com> Message-ID: <47D90B8B.2060207@mlbassoc.com> Hajducko, Steven wrote: > So we're trying to build a very minimal build that we then build upon to > form our systems. I'm trying to get rid of alot of unecessary packages > that we don't require for our systems. > > I'm running into a problem where the kickstart is still installing some > packages, even though I'm specifically telling it not too and the > package not being a required dependancy of any other package I'm > installing. > > My kickstart is very, very basic for this testing: > > install > cdrom > lang en_US > langsupport --default=en_US en_US > keyboard us > skipx > key #### > text > network --device eth0 --bootproto dhcp --hostname host.example.com > bootloader --location=mbr --driveorder=hda --append="rhgb quiet" > rootpw --iscrypted ########## > authconfig --enableshadow --enablemd5 > firewall --disabled > timezone America/Los_Angeles > selinux --disabled > clearpart --all --drives=hda > part /boot --fstype ext3 --size=128 --asprimary --ondrive=hda > part swap --size=12288 --asprimary > part pv.01 --size=100 --grow > volgroup rootvg pv.01 > logvol / --fstype ext3 --vgname=rootvg --size=20480 --name=rootlv > reboot > > %packages --nobase > @Core > -Deployment_Guide_en-US > -system-config-securitylevel-tui > > > The problem is, even with --nobase, I still keep getting the > Deployment_Guide and system-config-securitylevel-tui. > > The system-config package is part of the Base group, so it shouldn't be > installed anyways. The Deployment Guide is part of Core, but I'm > specifically telling it not to install. > > What am I missing here? Also, system-config-securitylevel-tui isn't > being brought in because of a requirement, it's just showing up. You > can remove the package as soon as the system comes up. It also doesn't > show in the anaconda.log as being added to fulfill a requirement. It > just suddenly shows up in the log, because slang and newt do get added > due to a requirement that system-config-securitylevel has. > > I do see an error in the log that 'no such package > system-config-securitylevel-tui', right after it tells me 'not adding > Base group'. > > Any input to this would be greatly appreciated it. I'd rather not muck > with post just to get some RPM's removed that should be taken care of > via the %packages section. What's your base system? I don't see these requirements in Fedora 8. Once your system is installed, you can see what led to this, e.g. [root at localhost tmp]# rpm -q --whatrequires system-config-firewall anaconda-11.4.0.40-1.1_AM -- ------------------------------------------------------------ Gary Thomas | Consulting for the MLB Associates | Embedded world ------------------------------------------------------------ From clumens at redhat.com Thu Mar 13 15:12:29 2008 From: clumens at redhat.com (Chris Lumens) Date: Thu, 13 Mar 2008 11:12:29 -0400 Subject: Wrong packages being installed. In-Reply-To: <21F8C5D33731F44BB9DCC396A12F9F2109067814@WLVEXM01.corp.ad.diginsite.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067814@WLVEXM01.corp.ad.diginsite.com> Message-ID: <20080313151229.GA4313@exeter.boston.redhat.com> > What am I missing here? Also, system-config-securitylevel-tui isn't > being brought in because of a requirement, it's just showing up. You > can remove the package as soon as the system comes up. It also doesn't > show in the anaconda.log as being added to fulfill a requirement. It > just suddenly shows up in the log, because slang and newt do get added > due to a requirement that system-config-securitylevel has. system-config-securitylevel-tui is installed because anaconda itself needs it to be installed for setting the root password, configuring the firewall, etc. - Chris From tom at ng23.net Thu Mar 13 16:19:06 2008 From: tom at ng23.net (Tom Brown) Date: Thu, 13 Mar 2008 16:19:06 +0000 Subject: kickstart LANG issue Message-ID: <47D953FA.50005@ng23.net> Hi In my kix file i have LANG set to the following lang en_GB langsupport --default=en_GB en_GB however on install the box is # env | grep LANG LANG=en_GB.UTF-8 # cat /etc/sysconfig/i18n LANG="en_GB.UTF-8" SUPPORTED="en_GB.UTF-8:en_GB:en" How can i make it so the default LANG is en_GB and not UTF-8 ? thanks From steven.hajducko at digitalinsight.com Thu Mar 13 17:18:50 2008 From: steven.hajducko at digitalinsight.com (Hajducko, Steven) Date: Thu, 13 Mar 2008 10:18:50 -0700 Subject: Wrong packages being installed. In-Reply-To: <20080313151229.GA4313@exeter.boston.redhat.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067814@WLVEXM01.corp.ad.diginsite.com> <20080313151229.GA4313@exeter.boston.redhat.com> Message-ID: <21F8C5D33731F44BB9DCC396A12F9F21090678AD@WLVEXM01.corp.ad.diginsite.com> Blech, so basically no way around getting rid of it. Any idea why the deployment guide refuses to not install itself? Like I said, I checked after I installed the system for anything that would have been bringing it in, here's what you get: [root at titan ~]# rpm -q --whatrequires system-config-securitylevel-tui no package requires system-config-securitylevel-tui [root at titan ~]# rpm -q --whatrequires Deployment_Guide-en-US no package requires Deployment_Guide-en-US I understand why the system-config-securitylevel-tui gets installed now, but I don't see what's pulling in the Deployment guide. -- sh > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Chris Lumens > Sent: Thursday, March 13, 2008 8:12 AM > To: kickstart-list at redhat.com > Subject: Re: Wrong packages being installed. > > > What am I missing here? Also, > system-config-securitylevel-tui isn't > > being brought in because of a requirement, it's just > showing up. You > > can remove the package as soon as the system comes up. It also > > doesn't show in the anaconda.log as being added to fulfill a > > requirement. It just suddenly shows up in the log, because > slang and > > newt do get added due to a requirement that > system-config-securitylevel has. > > system-config-securitylevel-tui is installed because anaconda > itself needs it to be installed for setting the root > password, configuring the firewall, etc. > > - Chris > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From steven.hajducko at digitalinsight.com Thu Mar 13 17:44:16 2008 From: steven.hajducko at digitalinsight.com (Hajducko, Steven) Date: Thu, 13 Mar 2008 10:44:16 -0700 Subject: Wrong packages being installed. In-Reply-To: <21F8C5D33731F44BB9DCC396A12F9F21090678AD@WLVEXM01.corp.ad.diginsite.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067814@WLVEXM01.corp.ad.diginsite.com><20080313151229.GA4313@exeter.boston.redhat.com> <21F8C5D33731F44BB9DCC396A12F9F21090678AD@WLVEXM01.corp.ad.diginsite.com> Message-ID: <21F8C5D33731F44BB9DCC396A12F9F21090678C5@WLVEXM01.corp.ad.diginsite.com> Nevermind, ignore the Deployment guide. Was due to my lack of sight and typing skills. Deployment_Guide-en-US != Deployment_Guide-en_US Thanks for the explanation of why the other package is included though. > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of > Hajducko, Steven > Sent: Thursday, March 13, 2008 10:19 AM > To: Discussion list about Kickstart > Subject: RE: Wrong packages being installed. > > Blech, so basically no way around getting rid of it. > > Any idea why the deployment guide refuses to not install itself? > > Like I said, I checked after I installed the system for > anything that would have been bringing it in, here's what you get: > > [root at titan ~]# rpm -q --whatrequires > system-config-securitylevel-tui no package requires > system-config-securitylevel-tui > > [root at titan ~]# rpm -q --whatrequires Deployment_Guide-en-US > no package requires Deployment_Guide-en-US > > I understand why the system-config-securitylevel-tui gets > installed now, but I don't see what's pulling in the > Deployment guide. > > -- > sh > > > -----Original Message----- > > From: kickstart-list-bounces at redhat.com > > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Chris Lumens > > Sent: Thursday, March 13, 2008 8:12 AM > > To: kickstart-list at redhat.com > > Subject: Re: Wrong packages being installed. > > > > > What am I missing here? Also, > > system-config-securitylevel-tui isn't > > > being brought in because of a requirement, it's just > > showing up. You > > > can remove the package as soon as the system comes up. It also > > > doesn't show in the anaconda.log as being added to fulfill a > > > requirement. It just suddenly shows up in the log, because > > slang and > > > newt do get added due to a requirement that > > system-config-securitylevel has. > > > > system-config-securitylevel-tui is installed because > anaconda itself > > needs it to be installed for setting the root password, configuring > > the firewall, etc. > > > > - Chris > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From steven.hajducko at digitalinsight.com Thu Mar 13 20:31:21 2008 From: steven.hajducko at digitalinsight.com (Hajducko, Steven) Date: Thu, 13 Mar 2008 13:31:21 -0700 Subject: Stupid dependencies.. ? Message-ID: <21F8C5D33731F44BB9DCC396A12F9F2109067955@WLVEXM01.corp.ad.diginsite.com> Is there a place to suggest better dependencies for future builds of RH? I'm not sure which list / bug tracking system I should be putting this into. For example, the core group lists authconfig as a mandatory package. authconfig requires rhpl, a package that contains python code used by other programs. rhpl requires wireless-tools. It was not this way in RH4. authconfig was part of core, but the RHPL package was not required by it, nor did the RHPL package require wireless-tools. So in essence, every server I install into my datacenter with RH5 is going to have the wireless-tools package, a completely arbitrary and unnecessary package, unless I uninstall authconfig, rhpl and wireless-tools in the %post. ..................................................................... Steven Hajducko Systems Engineer, Systems Integration w: 818.597.6443 m: 805.377.9074 This email may contain confidential and privileged material for the sole use of the intended recipient. Any review or distribution by others is strictly prohibited. If you are not the intended recipient please contact the sender and delete all copies. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: emailSign_logo2.gif Type: image/gif Size: 1293 bytes Desc: emailSign_logo2.gif URL: From mdehaan at redhat.com Thu Mar 13 20:41:53 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 13 Mar 2008 16:41:53 -0400 Subject: Stupid dependencies.. ? In-Reply-To: <21F8C5D33731F44BB9DCC396A12F9F2109067955@WLVEXM01.corp.ad.diginsite.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067955@WLVEXM01.corp.ad.diginsite.com> Message-ID: <47D99191.3020700@redhat.com> Hajducko, Steven wrote: > Is there a place to suggest better dependencies for future builds of > RH? I'm not sure which list / bug tracking system I should be putting > this into. > I would suggest bugging your support guys. > For example, > > the core group lists authconfig as a mandatory package. > > authconfig requires rhpl, a package that contains python code used by > other programs. Lots of things, including rhn register (IIRC) use rhpl, so that's reasonable. It also contains some translation code. > > rhpl requires wireless-tools. > Not sure why that is, but that seems busted. That being said, there may be a good reason for it and perhaps someone with more experience with rhpl can explain why. > It was not this way in RH4. authconfig was part of core, but the RHPL > package was not required by it, nor did the RHPL package require > wireless-tools. > > So in essence, every server I install into my datacenter with RH5 is > going to have the wireless-tools package, a completely arbitrary and > unnecessary package, unless I uninstall authconfig, rhpl and > wireless-tools in the %post. > > ..................................................................... > *Steven Hajducko* > Systems Engineer, Systems Integration > w: 818.597.6443 m: 805.377.9074 > > This email may contain confidential and privileged material for > the sole use of the intended recipient. Any review or distribution > by others is strictly prohibited. If you are not the intended > recipient please contact the sender and delete all copies. > > > ------------------------------------------------------------------------ > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From steven.hajducko at digitalinsight.com Thu Mar 13 20:51:21 2008 From: steven.hajducko at digitalinsight.com (Hajducko, Steven) Date: Thu, 13 Mar 2008 13:51:21 -0700 Subject: Stupid dependencies.. ? In-Reply-To: <47D99191.3020700@redhat.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067955@WLVEXM01.corp.ad.diginsite.com> <47D99191.3020700@redhat.com> Message-ID: <21F8C5D33731F44BB9DCC396A12F9F2109067962@WLVEXM01.corp.ad.diginsite.com> Yeah, the authconfig requiring rhpl does not bug me, as other things require rhpl as well. It's more that rhpl now requires wireless-tools that irks me. wireless-tools strikes me as a really 'as-needed' package, especially in the enterprise environment where many of the systems RHEL5 is being installed on have no wireless capability. > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Michael DeHaan > Sent: Thursday, March 13, 2008 1:42 PM > To: Discussion list about Kickstart > Subject: Re: Stupid dependencies.. ? > > Hajducko, Steven wrote: > > Is there a place to suggest better dependencies for future > builds of > > RH? I'm not sure which list / bug tracking system I should > be putting > > this into. > > > I would suggest bugging your support guys. > > > For example, > > > > the core group lists authconfig as a mandatory package. > > > > authconfig requires rhpl, a package that contains python > code used by > > other programs. > > Lots of things, including rhn register (IIRC) use rhpl, so that's > reasonable. It also contains some translation code. > > > > rhpl requires wireless-tools. > > > Not sure why that is, but that seems busted. That being said, there > may be a good reason for it and perhaps someone with more > experience with rhpl can explain why. > > > It was not this way in RH4. authconfig was part of core, > but the RHPL > > package was not required by it, nor did the RHPL package require > > wireless-tools. > > > > So in essence, every server I install into my datacenter > with RH5 is > > going to have the wireless-tools package, a completely > arbitrary and > > unnecessary package, unless I uninstall authconfig, rhpl and > > wireless-tools in the %post. > > > > > ..................................................................... > > *Steven Hajducko* > > Systems Engineer, Systems Integration > > w: 818.597.6443 m: 805.377.9074 > > > > This email may contain confidential and privileged material for the > > sole use of the intended recipient. Any review or distribution by > > others is strictly prohibited. If you are not the intended > recipient > > please contact the sender and delete all copies. > > > > > > > ---------------------------------------------------------------------- > > -- > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From debian at herakles.homelinux.org Fri Mar 14 09:02:44 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Fri, 14 Mar 2008 18:02:44 +0900 Subject: Stupid dependencies.. ? In-Reply-To: <21F8C5D33731F44BB9DCC396A12F9F2109067962@WLVEXM01.corp.ad.diginsite.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067955@WLVEXM01.corp.ad.diginsite.com> <47D99191.3020700@redhat.com> <21F8C5D33731F44BB9DCC396A12F9F2109067962@WLVEXM01.corp.ad.diginsite.com> Message-ID: <47DA3F34.8000307@herakles.homelinux.org> Hajducko, Steven wrote: > Yeah, the authconfig requiring rhpl does not bug me, as other things > require rhpl as well. It's more that rhpl now requires wireless-tools > that irks me. wireless-tools strikes me as a really 'as-needed' package, > especially in the enterprise environment where many of the systems RHEL5 > is being installed on have no wireless capability. > "strongly recommended" on laptops, "Are you sure you want this?" on servers and desktops. -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From steven.hajducko at digitalinsight.com Fri Mar 14 20:53:46 2008 From: steven.hajducko at digitalinsight.com (Hajducko, Steven) Date: Fri, 14 Mar 2008 13:53:46 -0700 Subject: Stupid dependencies.. ? In-Reply-To: <47DA3F34.8000307@herakles.homelinux.org> References: <21F8C5D33731F44BB9DCC396A12F9F2109067955@WLVEXM01.corp.ad.diginsite.com> <47D99191.3020700@redhat.com><21F8C5D33731F44BB9DCC396A12F9F2109067962@WLVEXM01.corp.ad.diginsite.com> <47DA3F34.8000307@herakles.homelinux.org> Message-ID: <21F8C5D33731F44BB9DCC396A12F9F2109067B7C@WLVEXM01.corp.ad.diginsite.com> Agreed, but not many laptops go into our datacenter. :) > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of John > Summerfield > Sent: Friday, March 14, 2008 2:03 AM > To: Discussion list about Kickstart > Subject: Re: Stupid dependencies.. ? > > Hajducko, Steven wrote: > > Yeah, the authconfig requiring rhpl does not bug me, as > other things > > require rhpl as well. It's more that rhpl now requires > wireless-tools > > that irks me. wireless-tools strikes me as a really 'as-needed' > > package, especially in the enterprise environment where many of the > > systems RHEL5 is being installed on have no wireless capability. > > > > "strongly recommended" on laptops, "Are you sure you want > this?" on servers and desktops. > > > -- > > Cheers > John > > -- spambait > 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu > -- Advice > http://webfoot.com/advice/email.top.php > http://www.catb.org/~esr/faqs/smart-questions.html > http://support.microsoft.com/kb/555375 > > You cannot reply off-list:-) > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From debian at herakles.homelinux.org Fri Mar 14 23:49:01 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Sat, 15 Mar 2008 08:49:01 +0900 Subject: Stupid dependencies.. ? In-Reply-To: <21F8C5D33731F44BB9DCC396A12F9F2109067B7C@WLVEXM01.corp.ad.diginsite.com> References: <21F8C5D33731F44BB9DCC396A12F9F2109067955@WLVEXM01.corp.ad.diginsite.com> <47D99191.3020700@redhat.com><21F8C5D33731F44BB9DCC396A12F9F2109067962@WLVEXM01.corp.ad.diginsite.com> <47DA3F34.8000307@herakles.homelinux.org> <21F8C5D33731F44BB9DCC396A12F9F2109067B7C@WLVEXM01.corp.ad.diginsite.com> Message-ID: <47DB0EED.6010202@herakles.homelinux.org> Hajducko, Steven wrote: > Agreed, but not many laptops go into our datacenter. :) OTOH Robert Day (Fedora) runs training courses and has admitted to having a few laptops, but not to a data centre. Imagine him in a hotel room with a switch full of laptops doing a ks install for tomorrow's course:-) > >> -----Original Message----- >> From: kickstart-list-bounces at redhat.com >> [mailto:kickstart-list-bounces at redhat.com] On Behalf Of John >> Summerfield >> Sent: Friday, March 14, 2008 2:03 AM >> To: Discussion list about Kickstart >> Subject: Re: Stupid dependencies.. ? >> >> Hajducko, Steven wrote: >>> Yeah, the authconfig requiring rhpl does not bug me, as >> other things >>> require rhpl as well. It's more that rhpl now requires >> wireless-tools >>> that irks me. wireless-tools strikes me as a really 'as-needed' >>> package, especially in the enterprise environment where many of the >>> systems RHEL5 is being installed on have no wireless capability. >>> >> "strongly recommended" on laptops, "Are you sure you want >> this?" on servers and desktops. >> >> >> -- >> >> Cheers >> John >> >> -- spambait >> 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu >> -- Advice >> http://webfoot.com/advice/email.top.php >> http://www.catb.org/~esr/faqs/smart-questions.html >> http://support.microsoft.com/kb/555375 >> >> You cannot reply off-list:-) >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From Gerrard.Geldenhuis at datacash.com Tue Mar 18 11:07:47 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Tue, 18 Mar 2008 11:07:47 -0000 Subject: Post script breakes kickstart file Message-ID: Hi I have the following few lines in a post script. When I add this to the post script running chrooted it breaks the kickstart. Meaning nothing after this lines gets executed. f [ $(grep -c eth /proc/net/dev) -eq 4 ]; then SECETH='eth2' else SECETH='eth1' I have encapsulated the script in brackets and piped the output to a log file but this is empty and provides no clue as to why this breaks the kickstart file. I am looking into the matter further but just thought maybe some one would at a glance know what is causing the problem. Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From debian at herakles.homelinux.org Tue Mar 18 11:16:27 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Tue, 18 Mar 2008 20:16:27 +0900 Subject: Post script breakes kickstart file In-Reply-To: References: Message-ID: <47DFA48B.2020107@herakles.homelinux.org> Gerrard Geldenhuis wrote: > Hi > > I have the following few lines in a post script. When I add this to the > post script running chrooted it breaks the kickstart. Meaning nothing > after this lines gets executed. Doesn't pass my scanner either > > > > f [ $(grep -c eth /proc/net/dev) -eq 4 ]; then should that be if? > > SECETH='eth2' > > else > > SECETH='eth1' a fi here maybe? > > > > I have encapsulated the script in brackets and piped the output to a log > file but this is empty and provides no clue as to why this breaks the > kickstart file. > > > > I am looking into the matter further but just thought maybe some one > would at a glance know what is causing the problem. > > > > Regards > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From rainer at ultra-secure.de Tue Mar 18 11:18:13 2008 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 18 Mar 2008 12:18:13 +0100 Subject: Post script breakes kickstart file In-Reply-To: References: Message-ID: <47DFA4F5.5080606@ultra-secure.de> Gerrard Geldenhuis schrieb: > > Hi > > I have the following few lines in a post script. When I add this to > the post script running chrooted it breaks the kickstart. Meaning > nothing after this lines gets executed. > > > > f [ $(grep -c eth /proc/net/dev) -eq 4 ]; then > > SECETH='eth2' > > else > > SECETH='eth1' > Is "f" equivalent to "if" or is this a typo? I've got several post-install steps and they actually do work, after some fiddling. cheers, Rainer From jwildebo at redhat.com Tue Mar 18 11:23:09 2008 From: jwildebo at redhat.com (Jan Wildeboer) Date: Tue, 18 Mar 2008 12:23:09 +0100 Subject: Post script breakes kickstart file In-Reply-To: References: Message-ID: <1205839389.24492.115.camel@jhwx61.muc.redhat.com> On Tue, 2008-03-18 at 11:07 +0000, Gerrard Geldenhuis wrote: I hope you just copy/pasted wrong, but: > f [ $(grep -c eth /proc/net/dev) -eq 4 ]; then Missing i > SECETH='eth2' > > else > > SECETH='eth1' Missing fi Jan -- Jan H Wildeboer | EMEA Open Source Affairs | Office: +49 (0)89 205071-207 Red Hat GmbH | Mobile: +49 (0)174 33 23 249 Otto-Hahn-Str.20 | Fax: +49 (0)89 205071-111 D-85609 Dornach/Munich | eMail: jan.wildeboer at redhat.com _____________________________________________________________________ GPG-Key-ID: 5DEBAFB0 GPG-Fingerprint: 6104 0F74 8513 F17E DFD5 E820 6F61 A078 5DEB AFB0 _____________________________________________________________________ Reg. Adresse: Red Hat GmbH, Hauptstaetter Strasse 58, 70178 Stuttgart Handelsregister: Amtsgericht Stuttgart HRB 153243 Geschaeftsfuehrer: Brendan Lane, Charlie Peters, Michael Cunningham, Werner Knoblich _____________________________________________________________________ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From Gerrard.Geldenhuis at datacash.com Tue Mar 18 11:23:51 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Tue, 18 Mar 2008 11:23:51 -0000 Subject: Post script breakes kickstart file In-Reply-To: <47DFA48B.2020107@herakles.homelinux.org> References: <47DFA48B.2020107@herakles.homelinux.org> Message-ID: I feel like an idiot... The f was a typo it was actually if I have forgotten the fi at the end. The test bed script I wrote did have the fi in it, it just seems I never copied it over. Here I was imagining all kind of strange bizarre reasons why my script were failing. Thanks > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of John Summerfield > Sent: 18 March 2008 11:16 > To: Discussion list about Kickstart > Subject: Re: Post script breakes kickstart file > > Gerrard Geldenhuis wrote: > > Hi > > > > I have the following few lines in a post script. When I add this to the > > post script running chrooted it breaks the kickstart. Meaning nothing > > after this lines gets executed. > Doesn't pass my scanner either > > > > > > > > > f [ $(grep -c eth /proc/net/dev) -eq 4 ]; then > should that be if? > > > > > SECETH='eth2' > > > > else > > > > SECETH='eth1' > > a fi here maybe? > > > > > > > > > I have encapsulated the script in brackets and piped the output to a log > > file but this is empty and provides no clue as to why this breaks the > > kickstart file. > > > > > > > > I am looking into the matter further but just thought maybe some one > > would at a glance know what is causing the problem. > > > > > > > > Regards > > > > > > > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > https://www.redhat.com/mailman/listinfo/kickstart-list > > > -- > > Cheers > John > > -- spambait > 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu > -- Advice > http://webfoot.com/advice/email.top.php > http://www.catb.org/~esr/faqs/smart-questions.html > http://support.microsoft.com/kb/555375 > > You cannot reply off-list:-) > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From debian at herakles.homelinux.org Tue Mar 18 12:09:22 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Tue, 18 Mar 2008 21:09:22 +0900 Subject: Post script breakes kickstart file In-Reply-To: References: <47DFA48B.2020107@herakles.homelinux.org> Message-ID: <47DFB0F2.9050207@herakles.homelinux.org> Gerrard Geldenhuis wrote: > I feel like an idiot... You took the words out of my fingers;-) -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From jasong at optonline.net Wed Mar 19 15:21:20 2008 From: jasong at optonline.net (jasong at optonline.net) Date: Wed, 19 Mar 2008 15:21:20 +0000 (GMT) Subject: Interactively prompting for a piece of install data Message-ID: Hi all,I'm new to kickstart and was wondering if it was possible to ask a single question, i.e. root password, during a kickstart install and automate anything else.? If it is, how is this accomplished? -------------- next part -------------- An HTML attachment was scrubbed... URL: From victor_sira at yahoo.com Wed Mar 19 15:47:42 2008 From: victor_sira at yahoo.com (Victor Sira) Date: Wed, 19 Mar 2008 08:47:42 -0700 (PDT) Subject: Interactively prompting for a piece of install data Message-ID: <897357.40001.qm@web51309.mail.re2.yahoo.com> Yes, possible. Take a look at this article for an example: http://www.redhatmagazine.com/2007/07/31/solutions-from-the-road-how-to-get-the-%e2%80%9creal%e2%80%9d-hostname-registered-during-satellite-provisioning-or-how-to-avoid-the-dreaded-%e2%80%9cunknown%e2%80%9d-hostname/ That asks for the hostname, but you could ask different questions. You can also parse /proc/cmdline for custom parameters passed at boot time. --Victor ----- Original Message ---- From: "jasong at optonline.net" To: kickstart-list at redhat.com Sent: Wednesday, March 19, 2008 8:21:20 AM Subject: Interactively prompting for a piece of install data Hi all, I'm new to kickstart and was wondering if it was possible to ask a single question, i.e. root password, during a kickstart install and automate anything else. If it is, how is this accomplished? ____________________________________________________________________________________ Looking for last minute shopping deals? Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping -------------- next part -------------- An HTML attachment was scrubbed... URL: From jasong at optonline.net Wed Mar 19 16:02:46 2008 From: jasong at optonline.net (jasong at optonline.net) Date: Wed, 19 Mar 2008 16:02:46 +0000 (GMT) Subject: Interactively prompting for a piece of install data In-Reply-To: <897357.40001.qm@web51309.mail.re2.yahoo.com> References: <897357.40001.qm@web51309.mail.re2.yahoo.com> Message-ID: Thanks Victor.So does this mean that Kickstart doesn't have a built in facility for doing this natively?----- Original Message -----From: Victor Sira Date: Wednesday, March 19, 2008 11:48 amSubject: Re: Interactively prompting for a piece of install dataTo: Discussion list about Kickstart > Yes, possible. Take a look at this article for an example:> > http://www.redhatmagazine.com/2007/07/31/solutions-from-the-road-> how-to-get-the-%e2%80%9creal%e2%80%9d-hostname-registered-during-> satellite-provisioning-or-how-to-avoid-the-dreaded-> %e2%80%9cunknown%e2%80%9d-hostname/> > That asks for the hostname, but you could ask different > questions. You can also parse /proc/cmdline for custom > parameters passed at boot time.> > --Victor> > ----- Original Message ----> From: "jasong at optonline.net" > To: kickstart-list at redhat.com> Sent: Wednesday, March 19, 2008 8:21:20 AM> Subject: Interactively prompting for a piece of install data> > Hi all,> > I'm new to kickstart and was wondering if it was possible to ask > a single question, i.e. root password, during a kickstart > install and automate anything else. If it is, how is this > accomplished?> > > > > > > ____________________________________________________________________________________Looking for last minute shopping deals? > Find them fast with Yahoo! Search. > http://tools.search.yahoo.com/newsearch/category.php?category=shopping -------------- next part -------------- An HTML attachment was scrubbed... URL: From victor_sira at yahoo.com Wed Mar 19 18:26:13 2008 From: victor_sira at yahoo.com (Victor Sira) Date: Wed, 19 Mar 2008 11:26:13 -0700 (PDT) Subject: Interactively prompting for a piece of install data Message-ID: <567765.95946.qm@web51310.mail.re2.yahoo.com> Not aware of any other facility to do this, but can't imagine a situation where these wouldn't be enough... --Victor ----- Original Message ---- From: "jasong at optonline.net" To: Discussion list about Kickstart Sent: Wednesday, March 19, 2008 9:02:46 AM Subject: Re: Interactively prompting for a piece of install data Thanks Victor. So does this mean that Kickstart doesn't have a built in facility for doing this natively? ----- Original Message ----- From: Victor Sira Date: Wednesday, March 19, 2008 11:48 am Subject: Re: Interactively prompting for a piece of install data To: Discussion list about Kickstart > Yes, possible. Take a look at this article for an example: > > http://www.redhatmagazine.com/2007/07/31/solutions-from-the-road- > how-to-get-the-%e2%80%9creal%e2%80%9d-hostname-registered-during- > satellite-provisioning-or-how-to-avoid-the-dreaded- > %e2%80%9cunknown%e2%80%9d-hostname/ > > That asks for the hostname, but you could ask different > questions. You can also parse /proc/cmdline for custom > parameters passed at boot time. > > --Victor > > ----- Original Message ---- > From: "jasong at optonline.net" > To: kickstart-list at redhat.com > Sent: Wednesday, March 19, 2008 8:21:20 AM > Subject: Interactively prompting for a piece of install data > > Hi all, > > I'm new to kickstart and was wondering if it was possible to ask > a single question, i.e. root password, during a kickstart > install and automate anything else. If it is, how is this > accomplished? > > > > > > > ____________________________________________________________________________________Looking for last minute shopping deals? > Find them fast with Yahoo! Search. > http://tools.search.yahoo.com/newsearch/category.php?category=shopping ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs -------------- next part -------------- An HTML attachment was scrubbed... URL: From jasong at optonline.net Wed Mar 19 18:29:07 2008 From: jasong at optonline.net (jasong at optonline.net) Date: Wed, 19 Mar 2008 18:29:07 +0000 (GMT) Subject: Interactively prompting for a piece of install data In-Reply-To: <567765.95946.qm@web51310.mail.re2.yahoo.com> References: <567765.95946.qm@web51310.mail.re2.yahoo.com> Message-ID: Ok.? Just curious, I'm having a debate with a colleague who uses SUSE and apparently their deployment tool can do this natively.----- Original Message -----From: Victor Sira Date: Wednesday, March 19, 2008 2:27 pmSubject: Re: Interactively prompting for a piece of install dataTo: Discussion list about Kickstart > Not aware of any other facility to do this, but can't imagine a > situation where these wouldn't be enough...> > --Victor> > > ----- Original Message ----> From: "jasong at optonline.net" > To: Discussion list about Kickstart > Sent: Wednesday, March 19, 2008 9:02:46 AM> Subject: Re: Interactively prompting for a piece of install data> > Thanks Victor.> > So does this mean that Kickstart doesn't have a built in > facility for doing this natively?> > ----- Original Message -----> From: Victor Sira > Date: Wednesday, March 19, 2008 11:48 am> Subject: Re: Interactively prompting for a piece of install data> To: Discussion list about Kickstart > > > Yes, possible. Take a look at this article for an example:> > > > http://www.redhatmagazine.com/2007/07/31/solutions-from-the-> road-> > how-to-get-the-%e2%80%9creal%e2%80%9d-hostname-registered-> during-> > satellite-provisioning-or-how-to-avoid-the-dreaded-> > %e2%80%9cunknown%e2%80%9d-hostname/> > > > That asks for the hostname, but you could ask different > > questions. You can also parse /proc/cmdline for custom > > parameters passed at boot time.> > > > --Victor> > > > ----- Original Message ----> > From: "jasong at optonline.net" > > To: kickstart-list at redhat.com> > Sent: Wednesday, March 19, 2008 8:21:20 AM> > Subject: Interactively prompting for a piece of install data> > > > Hi all,> > > > I'm new to kickstart and was wondering if it was possible to > ask > > a single question, i.e. root password, during a kickstart > > install and automate anything else. If it is, how is this > > accomplished?> > > > > > > > > > > > > > > ____________________________________________________________________________________Looking for last minute shopping deals? > > Find them fast with Yahoo! Search. > > > http://tools.search.yahoo.com/newsearch/category.php?category=shopping > > > > > > > ____________________________________________________________________________________Never miss a thing. Make Yahoo your home page. > http://www.yahoo.com/r/hs> -------------- next part -------------- An HTML attachment was scrubbed... URL: From kb4xley at gmail.com Mon Mar 24 15:37:59 2008 From: kb4xley at gmail.com (Kent Baxley) Date: Mon, 24 Mar 2008 11:37:59 -0400 Subject: using /dev/ram devices in RHEL5 kickstarts Message-ID: I'm researching a kickstart problem that deals with some of the differences between RHEL4 and RHEL5. In RHEL 4, I could create a RAM drive and copy files to this RAM drive in the %pre script portion of kickstart. I would then open the RAM drive in the %post section. This allowed me to pass information from the Pre to the Post processing scripts. Because the %pre scripts do not have a local drive to write to at this time in the installation process the became the easiest way tell the post process scripts what went on during the pre script installation choices. In the past, I was able to use the following code to create a ram drive during the %pre process of kickstart. mkdir /tmp/ramdisk mke2fs /dev/ram mount /dev/ram /tmp/ramdisk This code is right out of my kickstart. Once in the %post process mkdir /tmp/ramdisk mount /dev/ram /tmp/ramdisk Then I copy files from the /tmp/ramdisk mount point. The /dev/ram devices seem to be missing during the installation process, so, the above code snippents don't seem to work anymore on RHEL5. I've found I can work around this by doing something like: mknod /dev/ram b 1 0 Then continue with the steps above, however, why is /dev/ram seemingly not created by default anymore in RHEL5? Should I be using another method from here on out? Another possibility for passing data between %pre and %post environments, I've found, could be: %pre Create a tmpfs ram drive Copy files to tmpfs %post --nochroot copy files from tmpfs ram drive to /mnt/sysimage/path %post . . Notice that the first %post allows for a nochroot flag so I can copy data from the tmpfs. The second and last %post is where I would run my bash scripts after the data in tmpfs has been copied to the local hard drive. So, my question is, will either option (run mknod to create /dev/ram or use tmpfs) work, and what should I be doing for RHEL5 installs and beyond? I'm also open to any other suggestions. Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL: From ebrown at lanl.gov Mon Mar 24 15:54:15 2008 From: ebrown at lanl.gov (Ed Brown) Date: Mon, 24 Mar 2008 09:54:15 -0600 Subject: using /dev/ram devices in RHEL5 kickstarts In-Reply-To: References: Message-ID: <47E7CEA7.5050007@lanl.gov> Kent Baxley wrote: > I'm also open to any other suggestions. Any reason not to simply write to disk in %pre? %pre date >/tmp/date %post --nochroot cp /tmp/date /mnt/sysimage/tmp/date -Ed From kajtzu at basen.net Mon Mar 24 16:37:00 2008 From: kajtzu at basen.net (Kaj Niemi) Date: Mon, 24 Mar 2008 18:37:00 +0200 Subject: using /dev/ram devices in RHEL5 kickstarts In-Reply-To: References: Message-ID: <429D69D3-DB02-4585-BDEF-BF4CCEC2C137@basen.net> Hi, You could just create a subdirectory under /tmp and copy files there. Everything else (in %%post sections) is still applicable. Kaj On Mar 24, 2008, at 17:37, Kent Baxley wrote: > > So, my question is, will either option (run mknod to create /dev/ram > or use tmpfs) work, and what should I be doing for RHEL5 installs > and beyond? I'm also open to any other suggestions. HTH Kaj -- Kaj J. Niemi +358 45 63 12000 -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 3811 bytes Desc: not available URL: From ebrown at lanl.gov Mon Mar 24 17:00:03 2008 From: ebrown at lanl.gov (Ed Brown) Date: Mon, 24 Mar 2008 11:00:03 -0600 Subject: using /dev/ram devices in RHEL5 kickstarts In-Reply-To: <47E7CEA7.5050007@lanl.gov> References: <47E7CEA7.5050007@lanl.gov> Message-ID: <47E7DE13.8050009@lanl.gov> > Any reason not to simply write to disk in %pre? Should've quoted 'disk'. It's ramdisk of course, but it's available without having to set it up yourself, assuming it's large enough... -Ed From Chip.Shabazian at bankofamerica.com Mon Mar 24 17:15:39 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Mon, 24 Mar 2008 10:15:39 -0700 Subject: using /dev/ram devices in RHEL5 kickstarts In-Reply-To: References: Message-ID: The important thing to know about this is that you can run TWO %post sections. In the first, you can copy whatever you want from /tmp/your_dir to the chrooted /tmp, then you can use your current %post within a chrooted environment ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Kent Baxley Sent: Monday, March 24, 2008 8:38 AM To: kickstart-list at redhat.com Subject: using /dev/ram devices in RHEL5 kickstarts I'm researching a kickstart problem that deals with some of the differences between RHEL4 and RHEL5. In RHEL 4, I could create a RAM drive and copy files to this RAM drive in the %pre script portion of kickstart. I would then open the RAM drive in the %post section. This allowed me to pass information from the Pre to the Post processing scripts. Because the %pre scripts do not have a local drive to write to at this time in the installation process the became the easiest way tell the post process scripts what went on during the pre script installation choices. In the past, I was able to use the following code to create a ram drive during the %pre process of kickstart. mkdir /tmp/ramdisk mke2fs /dev/ram mount /dev/ram /tmp/ramdisk This code is right out of my kickstart. Once in the %post process mkdir /tmp/ramdisk mount /dev/ram /tmp/ramdisk Then I copy files from the /tmp/ramdisk mount point. The /dev/ram devices seem to be missing during the installation process, so, the above code snippents don't seem to work anymore on RHEL5. I've found I can work around this by doing something like: mknod /dev/ram b 1 0 Then continue with the steps above, however, why is /dev/ram seemingly not created by default anymore in RHEL5? Should I be using another method from here on out? Another possibility for passing data between %pre and %post environments, I've found, could be: %pre Create a tmpfs ram drive Copy files to tmpfs %post --nochroot copy files from tmpfs ram drive to /mnt/sysimage/path %post . . Notice that the first %post allows for a nochroot flag so I can copy data from the tmpfs. The second and last %post is where I would run my bash scripts after the data in tmpfs has been copied to the local hard drive. So, my question is, will either option (run mknod to create /dev/ram or use tmpfs) work, and what should I be doing for RHEL5 installs and beyond? I'm also open to any other suggestions. Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL: From stan.hearn at nscorp.com Mon Mar 24 17:40:01 2008 From: stan.hearn at nscorp.com (Hearn, Stan J.) Date: Mon, 24 Mar 2008 13:40:01 -0400 Subject: using /dev/ram devices in RHEL5 kickstarts In-Reply-To: Message-ID: So in order to not use a ram disk and to still have access to the files in the chrooted %post, you'd need to copy the files to a directory beneath /mnt/sysimage during the %pre stage, right? Something like /mnt/sysimage/var/tmp? This way no ram disk is needed and the data is written in %pre and readable by %post. ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Shabazian, Chip Sent: Monday, March 24, 2008 1:16 PM To: Discussion list about Kickstart Subject: RE: using /dev/ram devices in RHEL5 kickstarts The important thing to know about this is that you can run TWO %post sections. In the first, you can copy whatever you want from /tmp/your_dir to the chrooted /tmp, then you can use your current %post within a chrooted environment ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Kent Baxley Sent: Monday, March 24, 2008 8:38 AM To: kickstart-list at redhat.com Subject: using /dev/ram devices in RHEL5 kickstarts I'm researching a kickstart problem that deals with some of the differences between RHEL4 and RHEL5. In RHEL 4, I could create a RAM drive and copy files to this RAM drive in the %pre script portion of kickstart. I would then open the RAM drive in the %post section. This allowed me to pass information from the Pre to the Post processing scripts. Because the %pre scripts do not have a local drive to write to at this time in the installation process the became the easiest way tell the post process scripts what went on during the pre script installation choices. In the past, I was able to use the following code to create a ram drive during the %pre process of kickstart. mkdir /tmp/ramdisk mke2fs /dev/ram mount /dev/ram /tmp/ramdisk This code is right out of my kickstart. Once in the %post process mkdir /tmp/ramdisk mount /dev/ram /tmp/ramdisk Then I copy files from the /tmp/ramdisk mount point. The /dev/ram devices seem to be missing during the installation process, so, the above code snippents don't seem to work anymore on RHEL5. I've found I can work around this by doing something like: mknod /dev/ram b 1 0 Then continue with the steps above, however, why is /dev/ram seemingly not created by default anymore in RHEL5? Should I be using another method from here on out? Another possibility for passing data between %pre and %post environments, I've found, could be: %pre Create a tmpfs ram drive Copy files to tmpfs %post --nochroot copy files from tmpfs ram drive to /mnt/sysimage/path %post . . Notice that the first %post allows for a nochroot flag so I can copy data from the tmpfs. The second and last %post is where I would run my bash scripts after the data in tmpfs has been copied to the local hard drive. So, my question is, will either option (run mknod to create /dev/ram or use tmpfs) work, and what should I be doing for RHEL5 installs and beyond? I'm also open to any other suggestions. Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL: From Chip.Shabazian at bankofamerica.com Mon Mar 24 17:47:43 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Mon, 24 Mar 2008 10:47:43 -0700 Subject: using /dev/ram devices in RHEL5 kickstarts In-Reply-To: References: Message-ID: Exactly. The way I would do this is to create a directory in /tmp in the %pre and simply copy the contents of that directory to /mnt/sysimage/tmp in a %post -nochroot immediately preceeding your current %post. %pre mkdir /tmp/my_stuff_to_copy cp stuff /tmp/my_stuff_to_copy %post -nochroot cp -a /tmp/my_stuff_to_copy/* /mnt/sysimage/tmp %post Your_post_here ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Hearn, Stan J. Sent: Monday, March 24, 2008 10:40 AM To: Discussion list about Kickstart Subject: RE: using /dev/ram devices in RHEL5 kickstarts So in order to not use a ram disk and to still have access to the files in the chrooted %post, you'd need to copy the files to a directory beneath /mnt/sysimage during the %pre stage, right? Something like /mnt/sysimage/var/tmp? This way no ram disk is needed and the data is written in %pre and readable by %post. ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Shabazian, Chip Sent: Monday, March 24, 2008 1:16 PM To: Discussion list about Kickstart Subject: RE: using /dev/ram devices in RHEL5 kickstarts The important thing to know about this is that you can run TWO %post sections. In the first, you can copy whatever you want from /tmp/your_dir to the chrooted /tmp, then you can use your current %post within a chrooted environment ________________________________ From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Kent Baxley Sent: Monday, March 24, 2008 8:38 AM To: kickstart-list at redhat.com Subject: using /dev/ram devices in RHEL5 kickstarts I'm researching a kickstart problem that deals with some of the differences between RHEL4 and RHEL5. In RHEL 4, I could create a RAM drive and copy files to this RAM drive in the %pre script portion of kickstart. I would then open the RAM drive in the %post section. This allowed me to pass information from the Pre to the Post processing scripts. Because the %pre scripts do not have a local drive to write to at this time in the installation process the became the easiest way tell the post process scripts what went on during the pre script installation choices. In the past, I was able to use the following code to create a ram drive during the %pre process of kickstart. mkdir /tmp/ramdisk mke2fs /dev/ram mount /dev/ram /tmp/ramdisk This code is right out of my kickstart. Once in the %post process mkdir /tmp/ramdisk mount /dev/ram /tmp/ramdisk Then I copy files from the /tmp/ramdisk mount point. The /dev/ram devices seem to be missing during the installation process, so, the above code snippents don't seem to work anymore on RHEL5. I've found I can work around this by doing something like: mknod /dev/ram b 1 0 Then continue with the steps above, however, why is /dev/ram seemingly not created by default anymore in RHEL5? Should I be using another method from here on out? Another possibility for passing data between %pre and %post environments, I've found, could be: %pre Create a tmpfs ram drive Copy files to tmpfs %post --nochroot copy files from tmpfs ram drive to /mnt/sysimage/path %post . . Notice that the first %post allows for a nochroot flag so I can copy data from the tmpfs. The second and last %post is where I would run my bash scripts after the data in tmpfs has been copied to the local hard drive. So, my question is, will either option (run mknod to create /dev/ram or use tmpfs) work, and what should I be doing for RHEL5 installs and beyond? I'm also open to any other suggestions. Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL: From Gerrard.Geldenhuis at datacash.com Wed Mar 26 17:54:01 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Wed, 26 Mar 2008 17:54:01 -0000 Subject: Includes in pre scripts Message-ID: Hi I am trying to do the following in satellite: At the top of the kickstart file: device scsi cciss zerombr yes clearpart --all %include /tmp/partinfo In my pre script : %pre wget http://blah.com/pub/ks-scripts/post.sh -O /tmp/post.sh wget http://blah.com/pub/ks-scripts/pre.sh -O /tmp/pre.sh %include /tmp/pre.sh This works but my pre.sh script looks like the following: HOSTNAME=$(grep HOSTNAME /tmp/netinfo|cut -d= -f2|cut -d. -f1) vgn=vg_$HOSTNAME cat << EOF > /tmp/partinfo partition /boot --fstype=ext3 --size=200 partition pv.01 --size=1000 --grow partition swap --size=1000 --maxsize=2000 volgroup $vgn pv.01 logvol / --vgname=$vgn --name=rootvol --size=5120 logvol /home --vgname=$vgn --name=homevol --size=1024 logvol /tmp --vgname=$vgn --name=tmpvol --size=1024 logvol /var --vgname=$vgn --name=varvol --size=1536 logvol /var/log --vgname=$vgn --name=logvol --size=30720 logvol /var/tmp --vgname=$vgn --name=vartmpvol --size=1024 EOF So what I am expecting to happen is: 1. Download post and pre scripts from server. 2. Insert pre.sh as part of the pre script 3. Execute the pre.sh script 4. Continue on its mery way. However I get a No such file or directory: '/tmp/partinfo' Can anyone suggest a way of achieving this. I have 5 different kickstarts at the moment and I want them all to share the same post and pre scripts without having to edit the scripts on each one in the satellite config everytime I make a minor change. Any suggestions would be appreciated. Regards From michael at kmaclub.com Wed Mar 26 18:01:41 2008 From: michael at kmaclub.com (michael at kmaclub.com) Date: Wed, 26 Mar 2008 11:01:41 -0700 Subject: Includes in pre scripts In-Reply-To: References: Message-ID: <47EA8F85.7060907@kmaclub.com> Gerrard Geldenhuis wrote: > Hi > I am trying to do the following in satellite: > > At the top of the kickstart file: > device scsi cciss > zerombr yes > clearpart --all > %include /tmp/partinfo > > In my pre script : > %pre > wget http://blah.com/pub/ks-scripts/post.sh -O /tmp/post.sh > wget http://blah.com/pub/ks-scripts/pre.sh -O /tmp/pre.sh > %include /tmp/pre.sh You don't want to include /tmp/pre.sh, you want to run it sh /tmp/pre.sh Michael From stan.hearn at nscorp.com Wed Mar 26 18:40:21 2008 From: stan.hearn at nscorp.com (Hearn, Stan J.) Date: Wed, 26 Mar 2008 14:40:21 -0400 Subject: Includes in pre scripts In-Reply-To: <47EA8F85.7060907@kmaclub.com> Message-ID: What is the difference between "%include something" and "sh /something"? What is a good reason to use %include then? Are you saying this because the include is above the %pre? In other words, do you also want to use "sh /tmp/something" instead of include in a %pre or %post stanza? Stan -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of michael at kmaclub.com Sent: Wednesday, March 26, 2008 2:02 PM To: Discussion list about Kickstart Subject: Re: Includes in pre scripts Gerrard Geldenhuis wrote: > Hi > I am trying to do the following in satellite: > > At the top of the kickstart file: > device scsi cciss > zerombr yes > clearpart --all > %include /tmp/partinfo > > In my pre script : > %pre > wget http://blah.com/pub/ks-scripts/post.sh -O /tmp/post.sh > wget http://blah.com/pub/ks-scripts/pre.sh -O /tmp/pre.sh > %include /tmp/pre.sh You don't want to include /tmp/pre.sh, you want to run it sh /tmp/pre.sh Michael _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From rigg0022 at umn.edu Wed Mar 26 18:40:24 2008 From: rigg0022 at umn.edu (Ben Riggs) Date: Wed, 26 Mar 2008 13:40:24 -0500 Subject: Includes in pre scripts In-Reply-To: References: Message-ID: <47EA9898.7070202@umn.edu> %post #do some logic cat stuff > some_file %include some_file Ben Hearn, Stan J. wrote: > What is the difference between "%include something" and "sh /something"? > > > What is a good reason to use %include then? > > Are you saying this because the include is above the %pre? In other > words, do you also want to use "sh /tmp/something" instead of include in > a %pre or %post stanza? > > Stan > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of > michael at kmaclub.com > Sent: Wednesday, March 26, 2008 2:02 PM > To: Discussion list about Kickstart > Subject: Re: Includes in pre scripts > > Gerrard Geldenhuis wrote: > >> Hi >> I am trying to do the following in satellite: >> >> At the top of the kickstart file: >> device scsi cciss >> zerombr yes >> clearpart --all >> %include /tmp/partinfo >> >> In my pre script : >> %pre >> wget http://blah.com/pub/ks-scripts/post.sh -O /tmp/post.sh >> wget http://blah.com/pub/ks-scripts/pre.sh -O /tmp/pre.sh >> %include /tmp/pre.sh >> > > You don't want to include /tmp/pre.sh, you want to run it > > sh /tmp/pre.sh > > > Michael > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Pablo.Iranzo at redhat.com Wed Mar 26 18:46:27 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Wed, 26 Mar 2008 19:46:27 +0100 (CET) Subject: Includes in pre scripts In-Reply-To: References: Message-ID: On Wed, 26 Mar 2008, Hearn, Stan J. wrote: > What is the difference between "%include something" and "sh /something"? sh /something executes it %include just put contents into kickstart file passed to anaconda > What is a good reason to use %include then? For including scripts created or placed on a web server, maybe, generated dinamically based on values suplied on the wget call wget -O /tmp/ksdyn http://yourserver.com/cgi-bin/get-my-script.cgi?MAC=11:00:33:00:AA:FF&mem=128M %include /tmp/ksdyn For example, for create a bigger swap partition based on ram... > > Are you saying this because the include is above the %pre? In other > words, do you also want to use "sh /tmp/something" instead of include in > a %pre or %post stanza? It dependes on what do you really want, if you want to exec something, you can just wget it, then sh /yourscript and you're done, even mix in your scripts orders to create an include file that you'll later include with %include Regards Pablo From Gerrard.Geldenhuis at datacash.com Wed Mar 26 18:46:14 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Wed, 26 Mar 2008 18:46:14 -0000 Subject: Includes in pre scripts In-Reply-To: <47EA8F85.7060907@kmaclub.com> References: <47EA8F85.7060907@kmaclub.com> Message-ID: Thanks that has done the trick for me. Although the %include in the post section still worked as a an include. Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of michael at kmaclub.com > Sent: 26 March 2008 18:02 > To: Discussion list about Kickstart > Subject: Re: Includes in pre scripts > > Gerrard Geldenhuis wrote: > > Hi > > I am trying to do the following in satellite: > > > > At the top of the kickstart file: > > device scsi cciss > > zerombr yes > > clearpart --all > > %include /tmp/partinfo > > > > In my pre script : > > %pre > > wget http://blah.com/pub/ks-scripts/post.sh -O /tmp/post.sh > > wget http://blah.com/pub/ks-scripts/pre.sh -O /tmp/pre.sh > > %include /tmp/pre.sh > > You don't want to include /tmp/pre.sh, you want to run it > > sh /tmp/pre.sh > > > Michael > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From stan.hearn at nscorp.com Wed Mar 26 18:51:38 2008 From: stan.hearn at nscorp.com (Hearn, Stan J.) Date: Wed, 26 Mar 2008 14:51:38 -0400 Subject: Includes in pre scripts In-Reply-To: <47EA9898.7070202@umn.edu> Message-ID: Okay, I think I get it. "%include" can have anaconda specific commands where as "sh some_file" would specifically be shell scripting. But can't the "%include" contain shell commands? I don't mean to be obtuse, but I was surprised to hear that include wasn't working by executing the shell contents, but I'm assuming that is because it was outside a %pre or %post? Shell commands are not allowed outside %pre or %post, right? -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Ben Riggs Sent: Wednesday, March 26, 2008 2:40 PM To: Discussion list about Kickstart Subject: Re: Includes in pre scripts %post #do some logic cat stuff > some_file %include some_file Ben Hearn, Stan J. wrote: > What is the difference between "%include something" and "sh /something"? > > > What is a good reason to use %include then? > > Are you saying this because the include is above the %pre? In other > words, do you also want to use "sh /tmp/something" instead of include in > a %pre or %post stanza? > > Stan > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of > michael at kmaclub.com > Sent: Wednesday, March 26, 2008 2:02 PM > To: Discussion list about Kickstart > Subject: Re: Includes in pre scripts > > Gerrard Geldenhuis wrote: > >> Hi >> I am trying to do the following in satellite: >> >> At the top of the kickstart file: >> device scsi cciss >> zerombr yes >> clearpart --all >> %include /tmp/partinfo >> >> In my pre script : >> %pre >> wget http://blah.com/pub/ks-scripts/post.sh -O /tmp/post.sh >> wget http://blah.com/pub/ks-scripts/pre.sh -O /tmp/pre.sh >> %include /tmp/pre.sh >> > > You don't want to include /tmp/pre.sh, you want to run it > > sh /tmp/pre.sh > > > Michael > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From Chip.Shabazian at bankofamerica.com Wed Mar 26 19:18:21 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Wed, 26 Mar 2008 12:18:21 -0700 Subject: Includes in pre scripts In-Reply-To: References: <47EA9898.7070202@umn.edu> Message-ID: Think about it this way, the %include is replaced in the kickstart file by the file included. So, if your command section looked like this: install text lang en_US langsupport --default en_US en_US ... %include /tmp/timezone ... And in your %pre you had this: echo "TZ=America/New_York" > /tmp/timezone Anaconda would replace the %include line in the kickstart file so that this is what really got built: install text lang en_US langsupport --default en_US en_US ... TZ=America/New_York ... Does that explain what %include is doing? -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Hearn, Stan J. Sent: Wednesday, March 26, 2008 11:52 AM To: Discussion list about Kickstart Subject: RE: Includes in pre scripts Okay, I think I get it. "%include" can have anaconda specific commands where as "sh some_file" would specifically be shell scripting. But can't the "%include" contain shell commands? I don't mean to be obtuse, but I was surprised to hear that include wasn't working by executing the shell contents, but I'm assuming that is because it was outside a %pre or %post? Shell commands are not allowed outside %pre or %post, right? -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Ben Riggs Sent: Wednesday, March 26, 2008 2:40 PM To: Discussion list about Kickstart Subject: Re: Includes in pre scripts %post #do some logic cat stuff > some_file %include some_file Ben Hearn, Stan J. wrote: > What is the difference between "%include something" and "sh /something"? > > > What is a good reason to use %include then? > > Are you saying this because the include is above the %pre? In other > words, do you also want to use "sh /tmp/something" instead of include in > a %pre or %post stanza? > > Stan > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of > michael at kmaclub.com > Sent: Wednesday, March 26, 2008 2:02 PM > To: Discussion list about Kickstart > Subject: Re: Includes in pre scripts > > Gerrard Geldenhuis wrote: > >> Hi >> I am trying to do the following in satellite: >> >> At the top of the kickstart file: >> device scsi cciss >> zerombr yes >> clearpart --all >> %include /tmp/partinfo >> >> In my pre script : >> %pre >> wget http://blah.com/pub/ks-scripts/post.sh -O /tmp/post.sh >> wget http://blah.com/pub/ks-scripts/pre.sh -O /tmp/pre.sh >> %include /tmp/pre.sh >> > > You don't want to include /tmp/pre.sh, you want to run it > > sh /tmp/pre.sh > > > Michael > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From tibbs at math.uh.edu Wed Mar 26 20:00:18 2008 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: 26 Mar 2008 15:00:18 -0500 Subject: Behavior when kickstart file specifies a static IP Message-ID: I've noticed some changes between Fedora 7 and Fedora 8 when the kickstart file specifies a static IP. My normal install procedure generates a kickstart file for a specific host. The machine uses DHCP to get its initial IP, then grabs the kickstart file and proceeds to install via VNC. Under F7, the machine will change its IP to the one specified in the kickstart file once it's been downloaded, and the VNC window indicates the proper hostname of the device. Under F8, the machine keeps its DHCP address and the VNC window reflects whatever random address the DHCP server handed out, which is a bit suboptimal as now I can't track the progress of individual installs. I'm guessing that this was all intentional and that my expectations are not in line with reality, but is there any way of obtaining the previous behavior? - J< From debian at herakles.homelinux.org Wed Mar 26 23:19:35 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Thu, 27 Mar 2008 08:19:35 +0900 Subject: Includes in pre scripts In-Reply-To: <47EA8F85.7060907@kmaclub.com> References: <47EA8F85.7060907@kmaclub.com> Message-ID: <47EADA07.5020901@herakles.homelinux.org> michael at kmaclub.com wrote: > Gerrard Geldenhuis wrote: >> %include /tmp/pre.sh > > You don't want to include /tmp/pre.sh, you want to run it > > sh /tmp/pre.sh or . /tmp/pre.sh -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From debian at herakles.homelinux.org Wed Mar 26 23:21:20 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Thu, 27 Mar 2008 08:21:20 +0900 Subject: Includes in pre scripts In-Reply-To: References: Message-ID: <47EADA70.7070801@herakles.homelinux.org> Hearn, Stan J. wrote: > What is the difference between "%include something" and "sh /something"? > > > What is a good reason to use %include then? I'd be worried that it might not work, if one were to create a %pre script in a %pre script. -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From debian at herakles.homelinux.org Wed Mar 26 23:23:12 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Thu, 27 Mar 2008 08:23:12 +0900 Subject: Includes in pre scripts In-Reply-To: References: Message-ID: <47EADAE0.2000002@herakles.homelinux.org> Hearn, Stan J. wrote: > Okay, I think I get it. "%include" can have anaconda specific commands > where as "sh some_file" would specifically be shell scripting. But > can't the "%include" contain shell commands? > Provided it's run before the shell commands are required, yes. If the shell commands can be pre-determined,. I'd put them in the ks file. I generate my ks files by CGI. -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From pgroven at 2wire.com Thu Mar 27 17:43:59 2008 From: pgroven at 2wire.com (Phillip Groven) Date: Thu, 27 Mar 2008 10:43:59 -0700 Subject: UUID and kickstart In-Reply-To: <20080327160015.F11E9619C61@hormel.redhat.com> References: <20080327160015.F11E9619C61@hormel.redhat.com> Message-ID: <299E8BB0D5475E4FABB162E7DCA1BBE9023876E5@phxexch01.corp.2wire.com> I am trying to kickstart a server based on the UUID When I look in the HP ilo the UUID is listed as 368132USM53402MA The server is an HP DL360 When I run dmidecode I get the following UUID dmidecode |grep UUID UUID: 33363831-3334-5553-4D35-333430324D41 1. Why are they different? 2. Can I use the HPILO UUID to kickstart a box? 3. Should I be using the 32 bit number? I believe the UUID should be a file in /tftpboot/pxeboot/pxelinux.cfg 4. What should be in this file? -------------- next part -------------- An HTML attachment was scrubbed... URL: From ebrown at lanl.gov Thu Mar 27 18:39:12 2008 From: ebrown at lanl.gov (Ed Brown) Date: Thu, 27 Mar 2008 12:39:12 -0600 Subject: UUID and kickstart In-Reply-To: <299E8BB0D5475E4FABB162E7DCA1BBE9023876E5@phxexch01.corp.2wire.com> References: <20080327160015.F11E9619C61@hormel.redhat.com> <299E8BB0D5475E4FABB162E7DCA1BBE9023876E5@phxexch01.corp.2wire.com> Message-ID: <47EBE9D0.7000307@lanl.gov> A universal identifier can be used in various contexts, to identify different kinds of devices for example. This page I think answers your other questions: http://syslinux.zytor.com/pxe.php To make your life easier, if you're setting up a kickstart/PXE environment for the first time, you probably should check out: https://fedorahosted.org/cobbler/wiki/CobblerLiveCd -Ed From mdehaan at redhat.com Thu Mar 27 19:04:49 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 27 Mar 2008 15:04:49 -0400 Subject: UUID and kickstart In-Reply-To: <47EBE9D0.7000307@lanl.gov> References: <20080327160015.F11E9619C61@hormel.redhat.com> <299E8BB0D5475E4FABB162E7DCA1BBE9023876E5@phxexch01.corp.2wire.com> <47EBE9D0.7000307@lanl.gov> Message-ID: <47EBEFD1.9040602@redhat.com> Ed Brown wrote: > A universal identifier can be used in various contexts, to identify > different kinds of devices for example. This page I think answers > your other questions: > > http://syslinux.zytor.com/pxe.php > > To make your life easier, if you're setting up a kickstart/PXE > environment for the first time, you probably should check out: > > https://fedorahosted.org/cobbler/wiki/CobblerLiveCd > > -Ed > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list You should be kickstarting based on the MAC, and yes, Cobbler will help you do this. I would start here -- http://cobbler.et.redhat.com and not pay attention to the live CD docs until later if you need it. Not everyone does and if you have a working PXE setup, it's total overkill. --Michael From debian at herakles.homelinux.org Fri Mar 28 02:52:26 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Fri, 28 Mar 2008 11:52:26 +0900 Subject: UUID and kickstart In-Reply-To: <299E8BB0D5475E4FABB162E7DCA1BBE9023876E5@phxexch01.corp.2wire.com> References: <20080327160015.F11E9619C61@hormel.redhat.com> <299E8BB0D5475E4FABB162E7DCA1BBE9023876E5@phxexch01.corp.2wire.com> Message-ID: <47EC5D6A.6090209@herakles.homelinux.org> Phillip Groven wrote: > > > I am trying to kickstart a server based on the UUID > > When I look in the HP ilo the UUID is listed as > > 368132USM53402MA > > The server is an HP DL360 > > When I run dmidecode I get the following UUID > > dmidecode |grep UUID > > UUID: 33363831-3334-5553-4D35-333430324D41 > > 1. Why are they different? > > 2. Can I use the HPILO UUID to kickstart a box? > > > > 3. Should I be using the 32 bit number? > > I believe the UUID should be a file in > > /tftpboot/pxeboot/pxelinux.cfg Something like that. See what tftpd logs, and whether you can map that back to your documentation. > > > > 4. What should be in this file? > A standard pxelinux configuration, much like the syslinux config on your CD or DVD. -- Cheers John -- spambait 1aaaaaaa at coco.merseine.nu Z1aaaaaaa at coco.merseine.nu -- Advice http://webfoot.com/advice/email.top.php http://www.catb.org/~esr/faqs/smart-questions.html http://support.microsoft.com/kb/555375 You cannot reply off-list:-) From Rudi at SoftDux.com Fri Mar 28 13:55:39 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Fri, 28 Mar 2008 15:55:39 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? Message-ID: <47ECF8DB.2030004@SoftDux.com> Hi I'm trying to automate some server deployment, and would like to know how to tell kickstart to use a custom kernel, and install additional rpm's not normally found on the CD / DVD? -- Kind Regards Rudi Ahlers CEO, SoftDux Office: 087 805-9573 Fax No: 086 609 6128 Cell: 082 554 7532 Web: http://www.SoftDux.com Forum: http://Forum.SoftDux.com Join SA WebHostingTalk today, on http://www.WebHostingTalk.co.za From tom at ng23.net Fri Mar 28 14:15:25 2008 From: tom at ng23.net (Tom Brown) Date: Fri, 28 Mar 2008 14:15:25 +0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ECF8DB.2030004@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> Message-ID: <47ECFD7D.1030006@ng23.net> > > > I'm trying to automate some server deployment, and would like to know > how to tell kickstart to use a custom kernel, and install additional > rpm's not normally found on the CD / DVD? > rpm -i in the %post From Rudi at SoftDux.com Fri Mar 28 14:30:55 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Fri, 28 Mar 2008 16:30:55 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ECFD7D.1030006@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> Message-ID: <47ED011F.6090507@SoftDux.com> Tom Brown wrote: > >> >> >> I'm trying to automate some server deployment, and would like to know >> how to tell kickstart to use a custom kernel, and install additional >> rpm's not normally found on the CD / DVD? >> > > rpm -i in the %post > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > As follows? rpm -i xen-3.2.0-0xs.centos5.i386.rpm rpm -i xen-debuginfo-3.2.0-0xs.centos5.i386.rpm ..... etc? -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From tom at ng23.net Fri Mar 28 14:46:13 2008 From: tom at ng23.net (Tom Brown) Date: Fri, 28 Mar 2008 14:46:13 +0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED011F.6090507@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> Message-ID: <47ED04B5.6050904@ng23.net> > As follows? > > rpm -i xen-3.2.0-0xs.centos5.i386.rpm > rpm -i xen-debuginfo-3.2.0-0xs.centos5.i386.rpm > ..... > etc? > providing they are available at that location yes - personally i install over http for this but its up to you From Chip.Shabazian at bankofamerica.com Fri Mar 28 15:01:19 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Fri, 28 Mar 2008 08:01:19 -0700 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED04B5.6050904@ng23.net> Message-ID: The other thing you can do, if you are using RHEL 5, is to setup your own yum repo, include it in the command section, and then simply add the packages to the package section like the any other build. A bit more work to setup, but it works well. If you just want the quick and easy method, then doing it over http is the easiest as Tom said. Remember, you can load the rpm directly via http, no need to transfer it to the local system first. ex: rpm -Uvh http://my_server/rpm/location/xen-3.2.0-0xs.centos5.i386.rpm -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Tom Brown Sent: Friday, March 28, 2008 7:46 AM To: Discussion list about Kickstart Subject: Re: how to install a custom XEN kernel & rpm's from kickstart? > As follows? > > rpm -i xen-3.2.0-0xs.centos5.i386.rpm > rpm -i xen-debuginfo-3.2.0-0xs.centos5.i386.rpm > ..... > etc? > providing they are available at that location yes - personally i install over http for this but its up to you _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From robin-lists at robinbowes.com Fri Mar 28 15:02:45 2008 From: robin-lists at robinbowes.com (Robin Bowes) Date: Fri, 28 Mar 2008 15:02:45 +0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED011F.6090507@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> Message-ID: Rudi Ahlers wrote: > Tom Brown wrote: >> >>> >>> >>> I'm trying to automate some server deployment, and would like to know >>> how to tell kickstart to use a custom kernel, and install additional >>> rpm's not normally found on the CD / DVD? >>> >> >> rpm -i in the %post >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> > As follows? > > rpm -i xen-3.2.0-0xs.centos5.i386.rpm > rpm -i xen-debuginfo-3.2.0-0xs.centos5.i386.rpm Personally, I set up my own repo and add it to the kickstart: repo --name=myrepo --baseurl=http://example.com/centos/5/x86_64/RPMS Specify any additional RPMs in the %packages list Any RPMs in that repo that are older than the ones on the install media will be installed. Where did you get xen 3.2 for CentOS? Is that from the CentOS site? Ah, I see - it's i386. I ended up building 3.1.2 from Fedora for my x86_64 boxes. R. From tom at ng23.net Fri Mar 28 15:07:47 2008 From: tom at ng23.net (Tom Brown) Date: Fri, 28 Mar 2008 15:07:47 +0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> Message-ID: <47ED09C3.9010808@ng23.net> > > > Where did you get xen 3.2 for CentOS? Is that from the CentOS site? > Ah, I see - it's i386. I ended up building 3.1.2 from Fedora for my > x86_64 boxes. > i rebuilt this one for x86_64 http://bits.xensource.com/oss-xen/release/3.2.0/centos-5.1/xen-3.2.0-0xs.centos5.src.rpm From Gerrard.Geldenhuis at datacash.com Fri Mar 28 15:17:09 2008 From: Gerrard.Geldenhuis at datacash.com (Gerrard Geldenhuis) Date: Fri, 28 Mar 2008 15:17:09 -0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: References: <47ED04B5.6050904@ng23.net> Message-ID: Assuming you were using satellite server what would be prefferable? Custom channels in Satellite server for your sofware or your own yum repo's...? If you are using satellite I can't see the reason for using separate repos unless you have multiple satellite server which all subscribe to this custom repo. Regards > -----Original Message----- > From: kickstart-list-bounces at redhat.com [mailto:kickstart-list- > bounces at redhat.com] On Behalf Of Shabazian, Chip > Sent: 28 March 2008 15:01 > To: Discussion list about Kickstart > Subject: RE: how to install a custom XEN kernel & rpm's from kickstart? > > The other thing you can do, if you are using RHEL 5, is to setup your > own yum repo, include it in the command section, and then simply add the > packages to the package section like the any other build. A bit more > work to setup, but it works well. > > If you just want the quick and easy method, then doing it over http is > the easiest as Tom said. Remember, you can load the rpm directly via > http, no need to transfer it to the local system first. ex: > rpm -Uvh http://my_server/rpm/location/xen-3.2.0-0xs.centos5.i386.rpm > From Rudi at SoftDux.com Fri Mar 28 15:29:04 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Fri, 28 Mar 2008 17:29:04 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED09C3.9010808@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> Message-ID: <47ED0EC0.7080804@SoftDux.com> Tom Brown wrote: > >> >> >> Where did you get xen 3.2 for CentOS? Is that from the CentOS site? >> Ah, I see - it's i386. I ended up building 3.1.2 from Fedora for my >> x86_64 boxes. >> > > i rebuilt this one for x86_64 > > http://bits.xensource.com/oss-xen/release/3.2.0/centos-5.1/xen-3.2.0-0xs.centos5.src.rpm > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > That's the one I've just downloaded as well :) -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From Rudi at SoftDux.com Fri Mar 28 15:30:45 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Fri, 28 Mar 2008 17:30:45 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> Message-ID: <47ED0F25.5030909@SoftDux.com> Robin Bowes wrote: > Rudi Ahlers wrote: >> Tom Brown wrote: >>> >>>> >>>> >>>> I'm trying to automate some server deployment, and would like to >>>> know how to tell kickstart to use a custom kernel, and install >>>> additional rpm's not normally found on the CD / DVD? >>>> >>> >>> rpm -i in the %post >>> >>> _______________________________________________ >>> Kickstart-list mailing list >>> Kickstart-list at redhat.com >>> https://www.redhat.com/mailman/listinfo/kickstart-list >>> >> As follows? >> >> rpm -i xen-3.2.0-0xs.centos5.i386.rpm >> rpm -i xen-debuginfo-3.2.0-0xs.centos5.i386.rpm > > Personally, I set up my own repo and add it to the kickstart: > > repo --name=myrepo --baseurl=http://example.com/centos/5/x86_64/RPMS > > Specify any additional RPMs in the %packages list > > Any RPMs in that repo that are older than the ones on the install > media will be installed. > > Where did you get xen 3.2 for CentOS? Is that from the CentOS site? > Ah, I see - it's i386. I ended up building 3.1.2 from Fedora for my > x86_64 boxes. > > R. > > _______________________________________________ I need to make a CD in order to install remote servers & colocation providers, so I won't be able to setup a repo there, unless I put another server in, which will cost extra. So I need to keep everything on the CD, or even DVD -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From Rudi at SoftDux.com Fri Mar 28 15:32:31 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Fri, 28 Mar 2008 17:32:31 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED04B5.6050904@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED04B5.6050904@ng23.net> Message-ID: <47ED0F8F.9040309@SoftDux.com> Tom Brown wrote: > >> As follows? >> >> rpm -i xen-3.2.0-0xs.centos5.i386.rpm >> rpm -i xen-debuginfo-3.2.0-0xs.centos5.i386.rpm >> ..... >> etc? >> > > providing they are available at that location yes - personally i > install over http for this but its up to you > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > So, so I just copy them to the CentOS folder on the CD, or where else should they go? I need to keep everything on the CD, since it will be at remote locations like colocation providers or client's without internet & extra repo server(s) -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stugg From tom at ng23.net Fri Mar 28 15:38:01 2008 From: tom at ng23.net (Tom Brown) Date: Fri, 28 Mar 2008 15:38:01 +0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED0F8F.9040309@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED04B5.6050904@ng23.net> <47ED0F8F.9040309@SoftDux.com> Message-ID: <47ED10D9.9070005@ng23.net> > So, so I just copy them to the CentOS folder on the CD, or where else > should they go? I need to keep everything on the CD, since it will be > at remote locations like colocation providers or client's without > internet & extra repo server(s) > put them in a folder on the cd if you like and make sure your paths are right in the %post - they can be in any mount(able) area From Rudi at SoftDux.com Fri Mar 28 20:59:33 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Fri, 28 Mar 2008 22:59:33 +0200 Subject: some help needed with python script to detect number of drives in kickstart file Message-ID: <47ED5C35.5050401@SoftDux.com> Hi all I would like to put together a kickstart script which will allow me to install CentOS on many different platforms, regardless of the number & types of drive installed. If the system has 3 drives, I'd like to setup RAID 1 + 1 hot spare, for example. If it has 4 HDD's, then I'd like to setup RAID10, and for 10 HDD's, RAID 50. So, after trying many different options, I still can't get it to work, but recently found the following website, http://evuraan.blogspot.com/2005/02/auto-finding-your-hard-drives-for.html which basically outlines a way to determine how many drives are in the machine, but it's very basic. %pre --interpreter /usr/bin/python import os, sys sys.path.append('/usr/lib/anaconda') import isys # get a sorted list of drives drives = isys.hardDriveDict().keys() drives.sort() # write the include file to /tmp/kspart, drives[0] is the first drive, # drives[1] is the second, etc. To get the filet to be used, put # '%include /tmp/partitions' in your kickstart configuration. print "Writing partition details" f = open("/tmp/partitions", "w") f.write("part /boot --size 400 --ondisk %s\n" % drives[0]) f.write("part / --size 6144 --ondisk %s\n" % drives[0]) f.write("part swap --size 2048 --ondisk %s\n" % drives[0]) f.write("part /var --size 3072 --ondisk %s\n" % drives[0]) f.write("part /home --size 2048 --ondisk %s\n" % drives[0]) f.write("part /tmp --size 4096 --ondisk %s\n" % drives[0]) f.write("part /data --size 6144 --ondisk %s\n" % drives[0]) f.close() So, with this, I'd like to do different things, according to the number of drives, with an if/elseif loop (is there a better way?) This is what I've tried: if [ drives[] == "5" ] ; then f = open("/tmp/num_drives","w") f.write("5") f.close() elif [ drives[] == "4" ] ; then f = open("/tmp/num_drives","w") f.write("4") f.close() elif [ drives[] == "3" ] ; then f = open("/tmp/num_drives","w") f.write("3") f.close() <--snip--> but it fails. Can someone please help with this? -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From mdehaan at redhat.com Fri Mar 28 21:01:58 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Mar 2008 17:01:58 -0400 Subject: some help needed with python script to detect number of drives in kickstart file In-Reply-To: <47ED5C35.5050401@SoftDux.com> References: <47ED5C35.5050401@SoftDux.com> Message-ID: <47ED5CC6.8030409@redhat.com> Rudi Ahlers wrote: > Hi all > > I would like to put together a kickstart script which will allow me to > install CentOS on many different platforms, regardless of the number & > types of drive installed. If the system has 3 drives, I'd like to > setup RAID 1 + 1 hot spare, for example. If it has 4 HDD's, then I'd > like to setup RAID10, and for 10 HDD's, RAID 50. > > So, after trying many different options, I still can't get it to work, > but recently found the following website, > http://evuraan.blogspot.com/2005/02/auto-finding-your-hard-drives-for.html > which basically outlines a way to determine how many drives are in the > machine, but it's very basic. > This script is a bit overkill. From cobbler's snippets directory, I have this for default drive selection code. You will have do a bit more get RAID auto-configured, but that is left as an exercise for the reader. %include /tmp/partinfo %pre # Determine how many drives we have set \$(list-harddrives) let numd=\$#/2 d1=\$1 d2=\$3 cat << EOF > /tmp/partinfo part / --fstype ext3 --size=1024 --grow --ondisk=\$d1 --asprimary part swap --size=1024 --ondisk=\$d1 --asprimary EOF > > %pre --interpreter /usr/bin/python > import os, sys > sys.path.append('/usr/lib/anaconda') > import isys > > # get a sorted list of drives > drives = isys.hardDriveDict().keys() > drives.sort() > > # write the include file to /tmp/kspart, drives[0] is the first drive, > # drives[1] is the second, etc. To get the filet to be used, put > # '%include /tmp/partitions' in your kickstart configuration. > print "Writing partition details" > f = open("/tmp/partitions", "w") > f.write("part /boot --size 400 --ondisk %s\n" % drives[0]) > f.write("part / --size 6144 --ondisk %s\n" % drives[0]) > f.write("part swap --size 2048 --ondisk %s\n" % drives[0]) > f.write("part /var --size 3072 --ondisk %s\n" % drives[0]) > f.write("part /home --size 2048 --ondisk %s\n" % drives[0]) > f.write("part /tmp --size 4096 --ondisk %s\n" % drives[0]) > f.write("part /data --size 6144 --ondisk %s\n" % drives[0]) > > f.close() > > > > > > So, with this, I'd like to do different things, according to the > number of drives, with an if/elseif loop (is there a better way?) > > This is what I've tried: > > if [ drives[] == "5" ] ; then > f = open("/tmp/num_drives","w") > f.write("5") > f.close() > > elif [ drives[] == "4" ] ; then > f = open("/tmp/num_drives","w") > f.write("4") > f.close() > > elif [ drives[] == "3" ] ; then > f = open("/tmp/num_drives","w") > f.write("3") > f.close() > > <--snip--> > > but it fails. Can someone please help with this? > > From Rudi at SoftDux.com Fri Mar 28 21:15:58 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Fri, 28 Mar 2008 23:15:58 +0200 Subject: some help needed with python script to detect number of drives in kickstart file In-Reply-To: <47ED5CC6.8030409@redhat.com> References: <47ED5C35.5050401@SoftDux.com> <47ED5CC6.8030409@redhat.com> Message-ID: <47ED600E.6020503@SoftDux.com> Michael DeHaan wrote: > Rudi Ahlers wrote: >> Hi all >> >> I would like to put together a kickstart script which will allow me >> to install CentOS on many different platforms, regardless of the >> number & types of drive installed. If the system has 3 drives, I'd >> like to setup RAID 1 + 1 hot spare, for example. If it has 4 HDD's, >> then I'd like to setup RAID10, and for 10 HDD's, RAID 50. >> >> So, after trying many different options, I still can't get it to >> work, but recently found the following website, >> http://evuraan.blogspot.com/2005/02/auto-finding-your-hard-drives-for.html >> which basically outlines a way to determine how many drives are in >> the machine, but it's very basic. >> > This script is a bit overkill. From cobbler's snippets directory, I > have this for default drive selection code. You will have do a bit > more get RAID auto-configured, but that is left as an exercise for the > reader. > > %include /tmp/partinfo > > %pre > # Determine how many drives we have > set \$(list-harddrives) > let numd=\$#/2 > d1=\$1 > d2=\$3 > > cat << EOF > /tmp/partinfo > part / --fstype ext3 --size=1024 --grow --ondisk=\$d1 --asprimary > part swap --size=1024 --ondisk=\$d1 --asprimary > EOF > > I've tried that example already, as well, but the partitions never get created, with the more advanced code below: set $(list-harddrives) let numd=$#/2 # This will provide the total # of drives d1=$1 # This is the device of disk 1 d2=$3 # This is the device of disk 2, etc. d3=$5 d4=$7 d5=$9 S1=$2 # This is the size of disk 1 S2=$4 # This is the size of disk 2, etc. S3=$6 S4=$8 S5=$10 # This would be a partition scheme for three or more drives if [ $numd -ge 3 ] ; then cat << EOF >> /tmp/partinfo part raid.11 --size 200 --asprimary --ondrive=$d1 part raid.12 --size 2048 --asprimary --ondrive=$d1 part raid.13 --size 1 --grow --ondrive=$d1 part raid.21 --size 200 --asprimary --ondrive=$d2 part raid.22 --size 2048 --asprimary --ondrive=$d2 part raid.23 --size 1 --grow --ondrive=$d2 part raid.31 --size 200 --asprimary --ondrive=$d3 part raid.32 --size 2048 --asprimary --ondrive=$d3 part raid.33 --size 1 --grow --ondrive=$d3 raid /boot --fstype ext3 --device md0 --level=RAID1 raid.11 raid.21 raid.31 --spare=1 raid swap --fstype swap --device md1 --level=RAID1 raid.12 raid.22 raid.32 --spare=1 raid pv.01 --fstype ext3 --device md2 --level=RAID1 raid.13 raid.23 raid.33 --spare=1 logvol / --vgname=sysvg --size=3072 --name=root logvol /usr --vgname=sysvg --size=8192 --name=usr logvol /tmp --vgname=sysvg --size=1024 --name=tmp logvol /var --vgname=sysvg --size=8192 --name=var logvol /home --vgname=sysvg --size=20480 --name=home logvol /bck --vgname=pv.02 --size=1 --grow --name=bck EOF elif [ $numd - ge 2 ]; then cat << EOF >> /tmp/partinfo part /boot --fstype ext3 --size=150 --ondisk=$d1 part swap --size=4092 --ondisk=$d1,$d2 part pv.01 --fstype ext3 --size=100 --grow --ondisk=$d1 part pv.02 --fstype ext3 --size=100 --grow --ondisk=$d2 logvol / --vgname=sysvg --size=3072 --name=root logvol /usr --vgname=sysvg --size=8192 --name=usr logvol /tmp --vgname=sysvg --size=1024 --name=tmp logvol /var --vgname=sysvg --size=8192 --name=var logvol /home --vgname=sysvg --size=20480 --name=home EOF else cat << EOF >> /tmp/partinfo part /boot --fstype ext3 --size=150 --ondisk=$d1 part swap --size=4092 --ondisk=$d1 part pv.01 --fstype ext3 --size=100 --grow --ondisk=$d1 logvol / --vgname=sysvg --size=3072 --name=root logvol /usr --vgname=sysvg --size=8192 --name=usr logvol /tmp --vgname=sysvg --size=1024 --name=tmp logvol /var --vgname=sysvg --size=8192 --name=var logvol /home --vgname=sysvg --size=20480 --name=home EOF fi -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From mdehaan at redhat.com Fri Mar 28 21:48:39 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Mar 2008 17:48:39 -0400 Subject: some help needed with python script to detect number of drives in kickstart file In-Reply-To: <47ED600E.6020503@SoftDux.com> References: <47ED5C35.5050401@SoftDux.com> <47ED5CC6.8030409@redhat.com> <47ED600E.6020503@SoftDux.com> Message-ID: <47ED67B7.4000307@redhat.com> Rudi Ahlers wrote: > Michael DeHaan wrote: >> Rudi Ahlers wrote: >>> Hi all >>> >>> I would like to put together a kickstart script which will allow me >>> to install CentOS on many different platforms, regardless of the >>> number & types of drive installed. If the system has 3 drives, I'd >>> like to setup RAID 1 + 1 hot spare, for example. If it has 4 HDD's, >>> then I'd like to setup RAID10, and for 10 HDD's, RAID 50. >>> >>> So, after trying many different options, I still can't get it to >>> work, but recently found the following website, >>> http://evuraan.blogspot.com/2005/02/auto-finding-your-hard-drives-for.html >>> which basically outlines a way to determine how many drives are in >>> the machine, but it's very basic. >>> >> This script is a bit overkill. From cobbler's snippets directory, I >> have this for default drive selection code. You will have do a bit >> more get RAID auto-configured, but that is left as an exercise for >> the reader. >> >> %include /tmp/partinfo >> >> %pre >> # Determine how many drives we have >> set \$(list-harddrives) >> let numd=\$#/2 >> d1=\$1 >> d2=\$3 >> >> cat << EOF > /tmp/partinfo >> part / --fstype ext3 --size=1024 --grow --ondisk=\$d1 --asprimary >> part swap --size=1024 --ondisk=\$d1 --asprimary >> EOF >> >> > I've tried that example already, as well, but the partitions never get > created, with the more advanced code below: > This is not the problem of $(list-harddrives) but more likely what you were doing in your kickstart. From Chip.Shabazian at bankofamerica.com Fri Mar 28 21:51:07 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Fri, 28 Mar 2008 14:51:07 -0700 Subject: some help needed with python script to detect number of drives in kickstart file In-Reply-To: <47ED600E.6020503@SoftDux.com> Message-ID: Your problem is probably in your partitioning schemes. Try building with each scheme via kickstart. If they work then the problem is probably in your logic, where I already see one issue. set two drives to -eq, not -ge as 3 is always greater than 2. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Rudi Ahlers Sent: Friday, March 28, 2008 2:16 PM To: Discussion list about Kickstart Subject: Re: some help needed with python script to detect number of drives in kickstart file Michael DeHaan wrote: > Rudi Ahlers wrote: >> Hi all >> >> I would like to put together a kickstart script which will allow me >> to install CentOS on many different platforms, regardless of the >> number & types of drive installed. If the system has 3 drives, I'd >> like to setup RAID 1 + 1 hot spare, for example. If it has 4 HDD's, >> then I'd like to setup RAID10, and for 10 HDD's, RAID 50. >> >> So, after trying many different options, I still can't get it to >> work, but recently found the following website, >> http://evuraan.blogspot.com/2005/02/auto-finding-your-hard-drives-for >> .html which basically outlines a way to determine how many drives are >> in the machine, but it's very basic. >> > This script is a bit overkill. From cobbler's snippets directory, I > have this for default drive selection code. You will have do a bit > more get RAID auto-configured, but that is left as an exercise for the > reader. > > %include /tmp/partinfo > > %pre > # Determine how many drives we have > set \$(list-harddrives) > let numd=\$#/2 > d1=\$1 > d2=\$3 > > cat << EOF > /tmp/partinfo > part / --fstype ext3 --size=1024 --grow --ondisk=\$d1 --asprimary part > swap --size=1024 --ondisk=\$d1 --asprimary EOF > > I've tried that example already, as well, but the partitions never get created, with the more advanced code below: set $(list-harddrives) let numd=$#/2 # This will provide the total # of drives d1=$1 # This is the device of disk 1 d2=$3 # This is the device of disk 2, etc. d3=$5 d4=$7 d5=$9 S1=$2 # This is the size of disk 1 S2=$4 # This is the size of disk 2, etc. S3=$6 S4=$8 S5=$10 # This would be a partition scheme for three or more drives if [ $numd -ge 3 ] ; then cat << EOF >> /tmp/partinfo part raid.11 --size 200 --asprimary --ondrive=$d1 part raid.12 --size 2048 --asprimary --ondrive=$d1 part raid.13 --size 1 --grow --ondrive=$d1 part raid.21 --size 200 --asprimary --ondrive=$d2 part raid.22 --size 2048 --asprimary --ondrive=$d2 part raid.23 --size 1 --grow --ondrive=$d2 part raid.31 --size 200 --asprimary --ondrive=$d3 part raid.32 --size 2048 --asprimary --ondrive=$d3 part raid.33 --size 1 --grow --ondrive=$d3 raid /boot --fstype ext3 --device md0 --level=RAID1 raid.11 raid.21 raid.31 --spare=1 raid swap --fstype swap --device md1 --level=RAID1 raid.12 raid.22 raid.32 --spare=1 raid pv.01 --fstype ext3 --device md2 --level=RAID1 raid.13 raid.23 raid.33 --spare=1 logvol / --vgname=sysvg --size=3072 --name=root logvol /usr --vgname=sysvg --size=8192 --name=usr logvol /tmp --vgname=sysvg --size=1024 --name=tmp logvol /var --vgname=sysvg --size=8192 --name=var logvol /home --vgname=sysvg --size=20480 --name=home logvol /bck --vgname=pv.02 --size=1 --grow --name=bck EOF elif [ $numd - ge 2 ]; then cat << EOF >> /tmp/partinfo part /boot --fstype ext3 --size=150 --ondisk=$d1 part swap --size=4092 --ondisk=$d1,$d2 part pv.01 --fstype ext3 --size=100 --grow --ondisk=$d1 part pv.02 --fstype ext3 --size=100 --grow --ondisk=$d2 logvol / --vgname=sysvg --size=3072 --name=root logvol /usr --vgname=sysvg --size=8192 --name=usr logvol /tmp --vgname=sysvg --size=1024 --name=tmp logvol /var --vgname=sysvg --size=8192 --name=var logvol /home --vgname=sysvg --size=20480 --name=home EOF else cat << EOF >> /tmp/partinfo part /boot --fstype ext3 --size=150 --ondisk=$d1 part swap --size=4092 --ondisk=$d1 part pv.01 --fstype ext3 --size=100 --grow --ondisk=$d1 logvol / --vgname=sysvg --size=3072 --name=root logvol /usr --vgname=sysvg --size=8192 --name=usr logvol /tmp --vgname=sysvg --size=1024 --name=tmp logvol /var --vgname=sysvg --size=8192 --name=var logvol /home --vgname=sysvg --size=20480 --name=home EOF fi -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From Rudi at SoftDux.com Sat Mar 29 10:41:53 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Sat, 29 Mar 2008 12:41:53 +0200 Subject: please help, CentOS 5.1 kickstart installation looking for stuff from CD2??? Message-ID: <47EE1CF1.90402@SoftDux.com> Hi I'm putting together a CentOS 5.1 kickstart CD, with some specific software for my needs. Almost at the end, it's looking for a file from CD2, Deployment_Guide-en-US-5.1.0-11.el5.centos.1.noarch.rpm - which I have downloaded and put into the /CentOS folder on the CD already. Here's my packages selection from the ks.cfg file: %packages --nobase openssh-server #@virtualization acpid wget mutt links ftp mlocate man -aspell -aspell-en -bluez-utils -gpm -finger -irda-utils -isdn4k-utils -logwatch -ppp -rdate -rdist -rp-pppoe -rsh -wvdial -ypbind -yp-tools What is telling the installation process to install Deployment_Guide-en-US-5.1.0-11.el5.centos.1.noarch.rpm from CD2? And how do I make it install from the same CD? I'd like to keep to 1 CD only. -- Kind Regards Rudi Ahlers CEO, SoftDux Office: 087 805-9573 Fax No: 086 609 6128 Cell: 082 554 7532 Web: http://www.SoftDux.com Forum: http://Forum.SoftDux.com Join SA WebHostingTalk today, on http://www.WebHostingTalk.co.za From Pablo.Iranzo at redhat.com Sat Mar 29 10:47:05 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Sat, 29 Mar 2008 11:47:05 +0100 (CET) Subject: please help, CentOS 5.1 kickstart installation looking for stuff from CD2??? In-Reply-To: <47EE1CF1.90402@SoftDux.com> References: <47EE1CF1.90402@SoftDux.com> Message-ID: I suppose that you just copied the 1st install CD to a folder and try to kickstart from it, so it contains information about more "CD's" and dependencies are asking for Deployment_Guide which is on next CD... so it ask for it. If you just want an installation from one CD, you can REMOVE any package not needed nor it's dependencies, then run createrepo for updating the files used by anaconda/yum to setup packages and dependencies, and then, burn a CD with only those packages as well as the rest of the CentOS 5.1 CD except the RPMS folder as you'll be using yours. Regards Pablo -- Pablo Iranzo G??mez (http://Alufis35.uv.es/~iranzo/) (PGPKey Available on http://www.uv.es/~iranzop/PGPKey.pgp) -- Postulado de Boling sobre la Ley de Murphy: Si se encuentra bien, no se preocupe. Se le pasar?? On Sat, 29 Mar 2008, Rudi Ahlers wrote: > Hi > > I'm putting together a CentOS 5.1 kickstart CD, with some specific > software for my needs. Almost at the end, it's looking for a file from > CD2, Deployment_Guide-en-US-5.1.0-11.el5.centos.1.noarch.rpm - which I > have downloaded and put into the /CentOS folder on the CD already. > > Here's my packages selection from the ks.cfg file: > > > %packages --nobase > openssh-server > #@virtualization > acpid > wget > mutt > links > ftp > mlocate > man > -aspell > -aspell-en > -bluez-utils > -gpm > -finger > -irda-utils > -isdn4k-utils > -logwatch > -ppp > -rdate > -rdist > -rp-pppoe > -rsh > -wvdial > -ypbind > -yp-tools > > > What is telling the installation process to install > Deployment_Guide-en-US-5.1.0-11.el5.centos.1.noarch.rpm from CD2? And > how do I make it install from the same CD? I'd like to keep to 1 CD only. > > -- > > Kind Regards > Rudi Ahlers > CEO, SoftDux > Office: 087 805-9573 > Fax No: 086 609 6128 > Cell: 082 554 7532 > > > Web: http://www.SoftDux.com > Forum: http://Forum.SoftDux.com > > Join SA WebHostingTalk today, on http://www.WebHostingTalk.co.za > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Rudi at SoftDux.com Sat Mar 29 10:52:38 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Sat, 29 Mar 2008 12:52:38 +0200 Subject: please help, CentOS 5.1 kickstart installation looking for stuff from CD2??? In-Reply-To: References: <47EE1CF1.90402@SoftDux.com> Message-ID: <47EE1F76.7010007@SoftDux.com> Pablo Iranzo G?mez wrote: > I suppose that you just copied the 1st install CD to a folder and > try to kickstart from it, so it contains information about more "CD's" and > dependencies are asking for Deployment_Guide which is on next CD... so it > ask for it. > > If you just want an installation from one CD, you can REMOVE any > package not needed nor it's dependencies, then run createrepo for updating > the files used by anaconda/yum to setup packages and dependencies, and > then, burn a CD with only those packages as well as the rest of the CentOS > 5.1 CD except the RPMS folder as you'll be using yours. > > Regards > Pablo > Yes, I just copied cd.iso to a folder and removed a few things I didn't want. Later on I'm going to add a few rpm's which I want to the kickstart to copy to the installed system. I currently have all the files on a SME 7.3 server (which runs on a very modified version of CentOS 4.6) so I don't know if the createrepo command will work, or actually do what it's supposed todo. Is there any way of sorting this out manually? I'm happy to edit the files for the CD -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From Pablo.Iranzo at redhat.com Sat Mar 29 11:02:32 2008 From: Pablo.Iranzo at redhat.com (=?UTF-8?Q?Pablo_Iranzo_G=C3=B3mez?=) Date: Sat, 29 Mar 2008 12:02:32 +0100 (CET) Subject: please help, CentOS 5.1 kickstart installation looking for stuff from CD2??? In-Reply-To: <47EE1F76.7010007@SoftDux.com> References: <47EE1CF1.90402@SoftDux.com> <47EE1F76.7010007@SoftDux.com> Message-ID: Rudi, On Sat, 29 Mar 2008, Rudi Ahlers wrote: > Yes, I just copied cd.iso to a folder and removed a few things I didn't > want. Later on I'm going to add a few rpm's which I want to the > kickstart to copy to the installed system. > I currently have all the files on a SME 7.3 server (which runs on a > very modified version of CentOS 4.6) so I don't know if the createrepo > command will work, or actually do what it's supposed todo. You'll need :) createrepo updates files in repodata/ that are used during installation. > Is there any way of sorting this out manually? I'm happy to edit the > files for the CD What I would do personally, if I were at your situation would be to use reposync (in yum-utils) to take care of syncing needed packages, for example: I would need joe, mc, screen, ssh, so I would ask reposync to sync from a CentOS mirror those packages (and reposync will also download all needed dependencies). In order to have a more real setup, I will install a system with just: %packages --nobase to have the minimal files that @Core needs, then will rpm -qa to know which ones are needed and add them to my personal preference list in order to reposync to get the rest. At the end, reposync will have setup for me a copy of that repo with just what I need and it's dependencies and run createrepo on them... (That's what cobbler does ;) ) Regards Pablo From robin-lists at robinbowes.com Sat Mar 29 17:14:07 2008 From: robin-lists at robinbowes.com (Robin Bowes) Date: Sat, 29 Mar 2008 17:14:07 +0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED0F25.5030909@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED0F25.5030909@SoftDux.com> Message-ID: Rudi Ahlers wrote: > I need to make a CD in order to install remote servers & colocation > providers, so I won't be able to setup a repo there, unless I put > another server in, which will cost extra. So I need to keep everything > on the CD, or even DVD I have a xen instance that I use just for that purpose - building RPMs and hosting my repo. Alternatively, it should be possible to just add the updated RPMs to the CD/DVD image and to rebuild the repo information before you burn the disc. R. From robin-lists at robinbowes.com Sat Mar 29 17:16:19 2008 From: robin-lists at robinbowes.com (Robin Bowes) Date: Sat, 29 Mar 2008 17:16:19 +0000 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47ED09C3.9010808@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> Message-ID: Tom Brown wrote: > >> >> >> Where did you get xen 3.2 for CentOS? Is that from the CentOS site? >> Ah, I see - it's i386. I ended up building 3.1.2 from Fedora for >> my x86_64 boxes. >> > > i rebuilt this one for x86_64 > > http://bits.xensource.com/oss-xen/release/3.2.0/centos-5.1/xen-3.2.0-0xs.centos5.src.rpm > > Hrm. It's been a while since I tried, but I seem to recall I had problems with either building that SRPM, or problems running the resulting RPMs. Perhaps I'll give it another try... R. From Rudi at SoftDux.com Sun Mar 30 10:44:14 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Sun, 30 Mar 2008 12:44:14 +0200 Subject: please help, CentOS 5.1 kickstart installation looking for stuff from CD2??? In-Reply-To: References: <47EE1CF1.90402@SoftDux.com> <47EE1F76.7010007@SoftDux.com> Message-ID: <47EF6EFE.9040401@SoftDux.com> Pablo Iranzo G?mez wrote: > Rudi, > > On Sat, 29 Mar 2008, Rudi Ahlers wrote: > >> Yes, I just copied cd.iso to a folder and removed a few things I didn't >> want. Later on I'm going to add a few rpm's which I want to the >> kickstart to copy to the installed system. >> I currently have all the files on a SME 7.3 server (which runs on a >> very modified version of CentOS 4.6) so I don't know if the createrepo >> command will work, or actually do what it's supposed todo. >> > > You'll need :) createrepo updates files in repodata/ that are used > during installation. > Would this still work if the only other machine I have to my disposal to setup the ISO (where I'm currently creating the kickstart file & ISO file) is on a different OS & architecture? For example, I'm writing & testing a kickstart for CentOS 5.1 x64, and I only have 1 64bit PC at this stage. But our LAN server, as I said, runs on SME 7.3, which is CentOS 4.6 i386 - will this make any difference? > >> Is there any way of sorting this out manually? I'm happy to edit the >> files for the CD >> > > What I would do personally, if I were at your situation would be > to use reposync (in yum-utils) to take care of syncing needed packages, > for example: > > I would need joe, mc, screen, ssh, so I would ask reposync to sync > from a CentOS mirror those packages (and reposync will also download all > needed dependencies). In order to have a more real setup, I will install a > system with just: > > %packages --nobase to have the minimal files that @Core needs, > then will rpm -qa to know which ones are needed and add them to my > personal preference list in order to reposync to get the rest. > > At the end, reposync will have setup for me a copy of that repo > with just what I need and it's dependencies and run createrepo on them... > (That's what cobbler does ;) ) > > Regards > Pablo > Well, this is the thing. I already use "%packages --nobase " in my kickstart file, which worked fine previously, without wanting to install this particular rpm, but now it's looking for it on CD2. odd -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From debian at herakles.homelinux.org Sun Mar 30 11:21:21 2008 From: debian at herakles.homelinux.org (John Summerfield) Date: Sun, 30 Mar 2008 19:21:21 +0800 Subject: please help, CentOS 5.1 kickstart installation looking for stuff from CD2??? In-Reply-To: <47EF6EFE.9040401@SoftDux.com> References: <47EE1CF1.90402@SoftDux.com> <47EE1F76.7010007@SoftDux.com> <47EF6EFE.9040401@SoftDux.com> Message-ID: <47EF77B1.5000000@herakles.homelinux.org> Rudi Ahlers wrote: > Pablo Iranzo G?mez wrote: >> Rudi, >> >> On Sat, 29 Mar 2008, Rudi Ahlers wrote: >> >>> Yes, I just copied cd.iso to a folder and removed a few things I didn't >>> want. Later on I'm going to add a few rpm's which I want to the >>> kickstart to copy to the installed system. >>> I currently have all the files on a SME 7.3 server (which runs on a >>> very modified version of CentOS 4.6) so I don't know if the createrepo >>> command will work, or actually do what it's supposed todo. >>> >> >> You'll need :) createrepo updates files in repodata/ that are used >> during installation. >> > Would this still work if the only other machine I have to my disposal to > setup the ISO (where I'm currently creating the kickstart file & ISO > file) is on a different OS & architecture? Er. Windows? > For example, I'm writing & testing a kickstart for CentOS 5.1 x64, and I > only have 1 64bit PC at this stage. But our LAN server, as I said, runs > on SME 7.3, which is CentOS 4.6 i386 - will this make any difference? Probably:-) AFAIk C4 repos work with C5. I don't believe the architecture is important, but the version of createrepo probably is. createrepo is noarch, so it can be installed anywhere you like, assuming you have the tools (and on SME I expect you have) rpm2cpio References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> Message-ID: <47EFB02E.9020200@SoftDux.com> Tom Brown wrote: > >> >> >> Where did you get xen 3.2 for CentOS? Is that from the CentOS site? >> Ah, I see - it's i386. I ended up building 3.1.2 from Fedora for my >> x86_64 boxes. >> > > i rebuilt this one for x86_64 > > http://bits.xensource.com/oss-xen/release/3.2.0/centos-5.1/xen-3.2.0-0xs.centos5.src.rpm > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > Hi Tom What about the other files? On the Xen downloads page, http://xen.org/download/dl_32rhel5.html I see the following: xen-3.2.0.i386.rpm (hypervisor and management tools) xen-3.2.0-0.src.rpm (source) xen-debuginfo-3.2.0.i386.rpm (debugging information) xen-devel-3.2.0.i386.rpm (development library) xen-libs-3.2.0.i386.rpm (support libraries) I don't think I can install the i386 rpms on a CentOS 5.1 x64 system, right? Did you have problems with these files? Do I need them? What about the xen-3.2.0.i386.rpm (hypervisor and management tools) & xen-libs-3.2.0.i386.rpm (support libraries) ? -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From Rudi at SoftDux.com Sun Mar 30 16:52:22 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Sun, 30 Mar 2008 18:52:22 +0200 Subject: please help, CentOS 5.1 kickstart installation looking for stuff from CD2??? [SOLVED] In-Reply-To: <47EF77B1.5000000@herakles.homelinux.org> References: <47EE1CF1.90402@SoftDux.com> <47EE1F76.7010007@SoftDux.com> <47EF6EFE.9040401@SoftDux.com> <47EF77B1.5000000@herakles.homelinux.org> Message-ID: <47EFC546.5090009@SoftDux.com> John Summerfield wrote: > Rudi Ahlers wrote: >> Pablo Iranzo G???mez wrote: >>> Rudi, >>> >>> On Sat, 29 Mar 2008, Rudi Ahlers wrote: >>> >>>> Yes, I just copied cd.iso to a folder and removed a few things I >>>> didn't >>>> want. Later on I'm going to add a few rpm's which I want to the >>>> kickstart to copy to the installed system. >>>> I currently have all the files on a SME 7.3 server (which runs on a >>>> very modified version of CentOS 4.6) so I don't know if the >>>> createrepo >>>> command will work, or actually do what it's supposed todo. >>>> >>> >>> You'll need :) createrepo updates files in repodata/ that are used >>> during installation. >>> >> Would this still work if the only other machine I have to my disposal >> to setup the ISO (where I'm currently creating the kickstart file & >> ISO file) is on a different OS & architecture? > > Er. Windows? Ok, ok. Not windows, I meant different Linux flavours, say like Debian / Slackware / etc? > >> For example, I'm writing & testing a kickstart for CentOS 5.1 x64, >> and I only have 1 64bit PC at this stage. But our LAN server, as I >> said, runs on SME 7.3, which is CentOS 4.6 i386 - will this make any >> difference? > > Probably:-) AFAIk C4 repos work with C5. > > I don't believe the architecture is important, but the version of > createrepo probably is. createrepo is noarch, so it can be installed > anywhere you like, assuming you have the tools (and on SME I expect > you have) > rpm2cpio > You could use that technique to put createrepo into an area within > your local mirror and so, if needs be, have several different versions > for several different repos. thanx, this has worked well. createrepo -g repodata/comps.xml . is that command I used and it works quite well. Thank you very much :) -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From rmj at ast.cam.ac.uk Sun Mar 30 18:51:30 2008 From: rmj at ast.cam.ac.uk (Roderick Johnstone) Date: Sun, 30 Mar 2008 19:51:30 +0100 Subject: How to do a kickstart install to specific disk Message-ID: Hi Different linux kernels seem to name disks in a different order. I need to be able to specify the right disk for kickstart to install on to. Previously we have always used eg: part / --onpart sda1 --fstype ext3 --size 1 --grow but sda1 now may not be on the right disk! It has been suggested that I use --onbiosdisk to be sure I get the correct disk. How do I determine what the right argument to this option is? Is this the only way to be sure I'm installing to the right disk? Thanks Roderick Johnstone From tom at ng23.net Sun Mar 30 20:18:54 2008 From: tom at ng23.net (Tom Brown) Date: Sun, 30 Mar 2008 21:18:54 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47EFB02E.9020200@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> Message-ID: <47EFF5AE.1040502@ng23.net> >> > Hi Tom > > What about the other files? > > On the Xen downloads page, http://xen.org/download/dl_32rhel5.html I > see the following: > > > xen-3.2.0.i386.rpm (hypervisor and management tools) > > > xen-3.2.0-0.src.rpm (source) > > > xen-debuginfo-3.2.0.i386.rpm (debugging information) > > > xen-devel-3.2.0.i386.rpm (development library) > > > xen-libs-3.2.0.i386.rpm (support libraries) > > > > > > I don't think I can install the i386 rpms on a CentOS 5.1 x64 system, > right? Did you have problems with these files? Do I need them? What > about the xen-3.2.0.i386.rpm (hypervisor and management tools) > > > & xen-libs-3.2.0.i386.rpm (support libraries) > ? > > if you rebuild the x86_64 src rpm then these will be available to you From Rudi at SoftDux.com Sun Mar 30 20:21:36 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Sun, 30 Mar 2008 22:21:36 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47EFF5AE.1040502@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> Message-ID: <47EFF650.4090403@SoftDux.com> Tom Brown wrote: > >>> >> Hi Tom >> >> What about the other files? >> >> On the Xen downloads page, http://xen.org/download/dl_32rhel5.html I >> see the following: >> >> >> xen-3.2.0.i386.rpm (hypervisor and management tools) >> >> >> xen-3.2.0-0.src.rpm (source) >> >> >> xen-debuginfo-3.2.0.i386.rpm (debugging information) >> >> >> xen-devel-3.2.0.i386.rpm (development library) >> >> >> xen-libs-3.2.0.i386.rpm (support libraries) >> >> >> >> >> >> I don't think I can install the i386 rpms on a CentOS 5.1 x64 system, >> right? Did you have problems with these files? Do I need them? What >> about the xen-3.2.0.i386.rpm (hypervisor and management tools) >> >> >> & xen-libs-3.2.0.i386.rpm (support libraries) >> ? >> >> > > if you rebuild the x86_64 src rpm then these will be available to you Thanx, unfortunately I don't have a standard install of CentOS 5.1 x64. What do I need to install, in order to rebuild them, and how do I rebuild them? this is a first for me -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From tom at ng23.net Sun Mar 30 20:28:42 2008 From: tom at ng23.net (Tom Brown) Date: Sun, 30 Mar 2008 21:28:42 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47EFF650.4090403@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> Message-ID: <47EFF7FA.2090105@ng23.net> > > Thanx, unfortunately I don't have a standard install of CentOS 5.1 > x64. What do I need to install, in order to rebuild them, and how do I > rebuild them? this is a first for me > http://www.google.co.uk/search?q=rebuild+src+rpm&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a From dan at half-asleep.com Sun Mar 30 21:16:49 2008 From: dan at half-asleep.com (Daniel Segall) Date: Sun, 30 Mar 2008 17:16:49 -0400 Subject: How to do a kickstart install to specific disk In-Reply-To: References: Message-ID: <47F00341.701@half-asleep.com> You just need --ondisk=sdx -Dan Roderick Johnstone wrote: > Hi > > Different linux kernels seem to name disks in a different order. > > I need to be able to specify the right disk for kickstart to install on > to. Previously we have always used eg: > > part / --onpart sda1 --fstype ext3 --size 1 --grow > > but sda1 now may not be on the right disk! > > It has been suggested that I use --onbiosdisk to be sure I get the > correct disk. > > How do I determine what the right argument to this option is? > > Is this the only way to be sure I'm installing to the right disk? > > Thanks > > Roderick Johnstone > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Rudi at SoftDux.com Mon Mar 31 09:43:03 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 31 Mar 2008 11:43:03 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47EFF7FA.2090105@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> Message-ID: <47F0B227.1060004@SoftDux.com> Tom Brown wrote: > >> >> Thanx, unfortunately I don't have a standard install of CentOS 5.1 >> x64. What do I need to install, in order to rebuild them, and how do >> I rebuild them? this is a first for me >> > > http://www.google.co.uk/search?q=rebuild+src+rpm&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > Thanx, but what packages do I need to install to get rpmbuild? Like I said, I have a barebone CentOS 5.1 x64 system, so it's not installed and "yum seacrch rpmbuild" didn't return anything -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From tom at ng23.net Mon Mar 31 09:59:21 2008 From: tom at ng23.net (Tom Brown) Date: Mon, 31 Mar 2008 10:59:21 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0B227.1060004@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> Message-ID: <47F0B5F9.70002@ng23.net> >> > Thanx, but what packages do I need to install to get rpmbuild? Like I > said, I have a barebone CentOS 5.1 x64 system, so it's not installed > and "yum seacrch rpmbuild" didn't return anything > have you actually tried to find out? http://www.google.co.uk/search?q=rpmbuild+rpm&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a From Rudi at SoftDux.com Mon Mar 31 10:06:46 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 31 Mar 2008 12:06:46 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0B5F9.70002@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> Message-ID: <47F0B7B6.50601@SoftDux.com> Tom Brown wrote: > >>> >> Thanx, but what packages do I need to install to get rpmbuild? Like I >> said, I have a barebone CentOS 5.1 x64 system, so it's not installed >> and "yum seacrch rpmbuild" didn't return anything >> > > have you actually tried to find out? > > http://www.google.co.uk/search?q=rpmbuild+rpm&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > No, I haven't, cause I don't have oodles of bandwidth to download everything on every page from a google search hoping something would work. Thanx for the help though -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From tom at ng23.net Mon Mar 31 10:15:37 2008 From: tom at ng23.net (Tom Brown) Date: Mon, 31 Mar 2008 11:15:37 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0B7B6.50601@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> Message-ID: <47F0B9C9.8050904@ng23.net> > No, I haven't, cause I don't have oodles of bandwidth to download > everything on every page from a google search hoping something would > work. Thanx for the help though > errr you dont have bandwidth to be able to view web pages? reading is pretty lightweight From Rudi at SoftDux.com Mon Mar 31 10:18:59 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 31 Mar 2008 12:18:59 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0B9C9.8050904@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> Message-ID: <47F0BA93.3000503@SoftDux.com> Tom Brown wrote: > >> No, I haven't, cause I don't have oodles of bandwidth to download >> everything on every page from a google search hoping something would >> work. Thanx for the help though >> > > errr you dont have bandwidth to be able to view web pages? reading is > pretty lightweight > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > No, I don't have bandwidth to spare on downloading package xyz or abc cause one guy says this, and another says that. I haven't done this before, and I have a CentOS 5.1 x64 machine with very very little installed, so I need to download a lot of stuff for it to work probably. anyway, I'll look into it later, thank you -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From tom at ng23.net Mon Mar 31 10:30:33 2008 From: tom at ng23.net (Tom Brown) Date: Mon, 31 Mar 2008 11:30:33 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0BA93.3000503@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0BA93.3000503@SoftDux.com> Message-ID: <47F0BD49.2000908@ng23.net> >> > No, I don't have bandwidth to spare on downloading package xyz or abc > cause one guy says this, and another says that. I haven't done this > before, and I have a CentOS 5.1 x64 machine with very very little > installed, so I need to download a lot of stuff for it to work > probably. anyway, I'll look into it later, thank you > but by reading you'll find out what package to download without wasting bandwidth. If you have to pay for the transfer then its perhaps best to do a yum install rpm-build and find out what packages are required and then install these off the CD rather than from a repo. Or configure a local repo. From Rudi at SoftDux.com Mon Mar 31 10:36:12 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 31 Mar 2008 12:36:12 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0BD49.2000908@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0BA93.3000503@SoftDux.com> <47F0BD49.2000908@ng23.net> Message-ID: <47F0BE9C.4060107@SoftDux.com> Tom Brown wrote: > >>> >> No, I don't have bandwidth to spare on downloading package xyz or abc >> cause one guy says this, and another says that. I haven't done this >> before, and I have a CentOS 5.1 x64 machine with very very little >> installed, so I need to download a lot of stuff for it to work >> probably. anyway, I'll look into it later, thank you >> > > but by reading you'll find out what package to download without > wasting bandwidth. If you have to pay for the transfer then its > perhaps best to do a yum install rpm-build and find out what packages > are required and then install these off the CD rather than from a > repo. Or configure a local repo. > > _______________________________________________ But you're implying that it's wrong to ask the question on a list where I could possibly get a direct answer? With regard to configuring a local repo, it's going to cost me a lot of money to download the full CentOS 5 repo. I know the DVD is 3GB, which will cost me a fortune, since I pay per MB. Anyway, I'm not going to bother with it. OpenSuse 10.3 has a x64 rpm already so I'm going to use that instead. Thank you for your help -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From tom at ng23.net Mon Mar 31 10:54:24 2008 From: tom at ng23.net (Tom Brown) Date: Mon, 31 Mar 2008 11:54:24 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0BE9C.4060107@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0BA93.3000503@SoftDux.com> <47F0BD49.2000908@ng23.net> <47F0BE9C.4060107@SoftDux.com> Message-ID: <47F0C2E0.4060408@ng23.net> > But you're implying that it's wrong to ask the question on a list > where I could possibly get a direct answer? nope - being spoon fed info is no way to learn. Try things first and then come back with questions explaining whats not working. You first asked how to rebuild a src rpm whereas if you put that into a google search it would have told you in a matter of seconds, and even in the results without even needing to click on any links, same applies for the rpmbuild question. From Rudi at SoftDux.com Mon Mar 31 11:39:07 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 31 Mar 2008 13:39:07 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0B9C9.8050904@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> Message-ID: <47F0CD5B.2080504@SoftDux.com> Tom Brown wrote: > >> No, I haven't, cause I don't have oodles of bandwidth to download >> everything on every page from a google search hoping something would >> work. Thanx for the help though >> > > errr you dont have bandwidth to be able to view web pages? reading is > pretty lightweight > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > So, after installing 160MB's worth of package, just to install rpm-build, I still can't rebuild the package I'm now updating the whole syste, which will install another 53MB worth of packages. Sadly, google don't have an answer to my problems. [root at temp01 /]# rpmbuild --rebuild xen-3.2.0-0xs.centos5.src.rpm Installing xen-3.2.0-0xs.centos5.src.rpm warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root warning: user machbuild does not exist - using root warning: group machbuild does not exist - using root error: Failed build dependencies: /usr/include/gnu/stubs-32.h is needed by xen-3.2.0-0xs.x86_64 -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From tom at ng23.net Mon Mar 31 12:03:40 2008 From: tom at ng23.net (Tom Brown) Date: Mon, 31 Mar 2008 13:03:40 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0CD5B.2080504@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0CD5B.2080504@SoftDux.com> Message-ID: <47F0D31C.3090603@ng23.net> > So, after installing 160MB's worth of package, just to install > rpm-build, I still can't rebuild the package > I'm now updating the whole syste, which will install another 53MB > worth of packages. Sadly, google don't have an answer to my problems. > > > [root at temp01 /]# rpmbuild --rebuild xen-3.2.0-0xs.centos5.src.rpm > Installing xen-3.2.0-0xs.centos5.src.rpm > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > warning: user machbuild does not exist - using root > warning: group machbuild does not exist - using root > error: Failed build dependencies: > /usr/include/gnu/stubs-32.h is needed by xen-3.2.0-0xs.x86_64 > > http://www.google.co.uk/search?q=stubs-32.h+rpm&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a 2nd hit - maybe even first but not clicked on any links From gary at mlbassoc.com Mon Mar 31 12:22:04 2008 From: gary at mlbassoc.com (Gary Thomas) Date: Mon, 31 Mar 2008 06:22:04 -0600 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0D31C.3090603@ng23.net> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0CD5B.2080504@SoftDux.com> <47F0D31C.3090603@ng23.net> Message-ID: <47F0D76C.3060800@mlbassoc.com> Tom Brown wrote: > >> So, after installing 160MB's worth of package, just to install >> rpm-build, I still can't rebuild the package >> I'm now updating the whole syste, which will install another 53MB >> worth of packages. Sadly, google don't have an answer to my problems. >> >> >> [root at temp01 /]# rpmbuild --rebuild xen-3.2.0-0xs.centos5.src.rpm >> Installing xen-3.2.0-0xs.centos5.src.rpm >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> warning: user machbuild does not exist - using root >> warning: group machbuild does not exist - using root >> error: Failed build dependencies: >> /usr/include/gnu/stubs-32.h is needed by xen-3.2.0-0xs.x86_64 >> >> > > http://www.google.co.uk/search?q=stubs-32.h+rpm&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a > > > 2nd hit - maybe even first but not clicked on any links > Or, even more useful/explicit: # yum whatprovides /usr/include/gnu/stubs-32.h glibc-devel.i386 : Object files for development using standard C libraries. -- ------------------------------------------------------------ Gary Thomas | Consulting for the MLB Associates | Embedded world ------------------------------------------------------------ From rainer at ultra-secure.de Mon Mar 31 13:02:32 2008 From: rainer at ultra-secure.de (Rainer Duffner) Date: Mon, 31 Mar 2008 15:02:32 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0CD5B.2080504@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0CD5B.2080504@SoftDux.com> Message-ID: <47F0E0E8.4070005@ultra-secure.de> Rudi Ahlers schrieb: > Tom Brown wrote: >> >>> No, I haven't, cause I don't have oodles of bandwidth to download >>> everything on every page from a google search hoping something would >>> work. Thanx for the help though >>> >> >> errr you dont have bandwidth to be able to view web pages? reading is >> pretty lightweight >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list >> > So, after installing 160MB's worth of package, just to install > rpm-build, I still can't rebuild the package > I'm now updating the whole syste, which will install another 53MB > worth of packages. Sadly, google don't have an answer to my problems. > > > [root at temp01 /]# rpmbuild --rebuild xen-3.2.0-0xs.centos5.src.rpm You should not run rpmbuild as root (unless xen needs that - I am not sure, I don't use xen). rpmbuild requires some additional RPMs. You should really (!) try out all this stuff in some sort of VM. I assume you have enough (umetered) bandwidth at home/work and only the colo-traffic is paid by MB. Yes, it's wasteful to download a DVD image and we should all use less resources to save the planet. But even on our old 600/500 (kBit/s, not MBit/s) line at the old office, the download of a DVD ISO would just take a weekend. And if even that is an issue, two minutes of searching via Google found this: http://www.pricecheck.co.za/offers/2632237/Linux%3A+CentOS+5.1+DVD+%2B+Free+postage%21/ That's http://finance.yahoo.com/currency/convert?amt=55&from=ZAR&to=EUR&submit=Convert about 4 Euro and 20 Euro-Cents. Is your time really worth so little? cheers, Rainer From Rudi at SoftDux.com Mon Mar 31 13:20:27 2008 From: Rudi at SoftDux.com (Rudi Ahlers) Date: Mon, 31 Mar 2008 15:20:27 +0200 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0E0E8.4070005@ultra-secure.de> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0CD5B.2080504@SoftDux.com> <47F0E0E8.4070005@ultra-secure.de> Message-ID: <47F0E51B.5030103@SoftDux.com> Rainer Duffner wrote: > Rudi Ahlers schrieb: >> Tom Brown wrote: >>> >>>> No, I haven't, cause I don't have oodles of bandwidth to download >>>> everything on every page from a google search hoping something >>>> would work. Thanx for the help though >>>> >>> >>> errr you dont have bandwidth to be able to view web pages? reading >>> is pretty lightweight >>> >>> _______________________________________________ >>> Kickstart-list mailing list >>> Kickstart-list at redhat.com >>> https://www.redhat.com/mailman/listinfo/kickstart-list >>> >> So, after installing 160MB's worth of package, just to install >> rpm-build, I still can't rebuild the package >> I'm now updating the whole syste, which will install another 53MB >> worth of packages. Sadly, google don't have an answer to my problems. >> >> >> [root at temp01 /]# rpmbuild --rebuild xen-3.2.0-0xs.centos5.src.rpm > > > You should not run rpmbuild as root (unless xen needs that - I am not > sure, I don't use xen). > rpmbuild requires some additional RPMs. > > You should really (!) try out all this stuff in some sort of VM. > I assume you have enough (umetered) bandwidth at home/work and only > the colo-traffic is paid by MB. > > Yes, it's wasteful to download a DVD image and we should all use less > resources to save the planet. > But even on our old 600/500 (kBit/s, not MBit/s) line at the old > office, the download of a DVD ISO would just take a weekend. > > And if even that is an issue, two minutes of searching via Google > found this: > http://www.pricecheck.co.za/offers/2632237/Linux%3A+CentOS+5.1+DVD+%2B+Free+postage%21/ > > That's > http://finance.yahoo.com/currency/convert?amt=55&from=ZAR&to=EUR&submit=Convert > > about 4 Euro and 20 Euro-Cents. > Is your time really worth so little? > Hi Rainer No, I pay for bandwidth everywhere in my country, at home / office / colocation / etc. Needless to say, I'm trying to setup a VM for stuff like this. I only have one 64bit PC, and would like to use it for both 32bit & 64bit environments, so I was told that Xen 3.0 (which is the default in CentOS 5.1) won't work, I need Xen 3.2 - but there's no 64bit rpm's available, and it seems no one actually has 64bit rpm's ( I find this very very strange), and people also think that bandwidth all over the world is free, or even cheap So far, I've downloaded 254MB's worth of stuff just to install rpm-build, and I still can't get it to work. Unfortunately those who have managed to rebuild it think it's too much to share in their experience, so Im running around in circles, downloading everything I can think of to try and get it to work. [root at temp01 /]# du --si /var/cache/yum/*/packages 4.1k /var/cache/yum/addons/packages 123M /var/cache/yum/base/packages 4.1k /var/cache/yum/extras/packages 131M /var/cache/yum/updates/packages [root at temp01 /]# P.S. This was a new installation this morning, so I didn't have any updates or newly installed packages in those folders before stating this project. -- Kind Regards Rudi Ahlers CEO, SoftDux Web: http://www.SoftDux.com Check out my technical blog, http://blog.softdux.com for Linux or other technical stuff, or visit http://www.WebHostingTalk.co.za for Web Hosting stuff From tom at ng23.net Mon Mar 31 13:23:39 2008 From: tom at ng23.net (Tom Brown) Date: Mon, 31 Mar 2008 14:23:39 +0100 Subject: how to install a custom XEN kernel & rpm's from kickstart? In-Reply-To: <47F0E51B.5030103@SoftDux.com> References: <47ECF8DB.2030004@SoftDux.com> <47ECFD7D.1030006@ng23.net> <47ED011F.6090507@SoftDux.com> <47ED09C3.9010808@ng23.net> <47EFB02E.9020200@SoftDux.com> <47EFF5AE.1040502@ng23.net> <47EFF650.4090403@SoftDux.com> <47EFF7FA.2090105@ng23.net> <47F0B227.1060004@SoftDux.com> <47F0B5F9.70002@ng23.net> <47F0B7B6.50601@SoftDux.com> <47F0B9C9.8050904@ng23.net> <47F0CD5B.2080504@SoftDux.com> <47F0E0E8.4070005@ultra-secure.de> <47F0E51B.5030103@SoftDux.com> Message-ID: <47F0E5DB.20901@ng23.net> > > > So far, I've downloaded 254MB's worth of stuff just to install > rpm-build, and I still can't get it to work. Unfortunately those who > have managed to rebuild it think it's too much to share in their > experience, so Im running around in circles, downloading everything I > can think of to try and get it to work. > what is the error you are now seeing?