From ugob at lubik.ca Mon Nov 3 16:33:23 2008 From: ugob at lubik.ca (Ugo Bellavance) Date: Mon, 03 Nov 2008 11:33:23 -0500 Subject: Anaconda not looking for the right package Message-ID: Hi, I'm trying to install RHEL 5 U2 via kickstart, but I get an error saying "The file sysklogd-1.4.1-32.2.x86_64.rpm cannot be opened. It is right, the file is not there, the file that is there is sysklogd-1.4.1-44.el5.x86_64.rpm. Where should I start looking for potential errors? Please let me know if you need more info. Regards, Ugo From kpowell at redhat.com Mon Nov 3 21:41:46 2008 From: kpowell at redhat.com (Kyle Powell) Date: Mon, 03 Nov 2008 16:41:46 -0500 Subject: Anaconda not looking for the right package In-Reply-To: References: Message-ID: <490F701A.9080807@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Ugo Bellavance wrote: > Hi, > > I'm trying to install RHEL 5 U2 via kickstart, but I get an error saying > "The file sysklogd-1.4.1-32.2.x86_64.rpm cannot be opened. > > It is right, the file is not there, the file that is there is > sysklogd-1.4.1-44.el5.x86_64.rpm. This is correct. The package that ships with RHEL 5 U2 is sysklogd-1.4.1-44.el5.x86_64.rpm > Where should I start looking for potential errors? I'd start with your kickstart file. > Please let me know if you need more info. If you'd like to provide your kickstart file, we may be able to assist you further. - -- Kyle Powell | Red Hat | Senior Consultant, RHCE | Cell: 571.215.4340 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJD3Aa7pTtanQdBU4RAl+fAJwJF56DGrGl0mZh4fzU2H9ZWm5yuACghf3t ZwXxL5YDLpg3TPgBlIFl96g= =0X2t -----END PGP SIGNATURE----- From tony.chamberlain at lemko.com Tue Nov 4 03:00:48 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Tue, 04 Nov 2008 03:00:48 +0000 Subject: extra packages Message-ID: After a bit of time I finally was able to make an installation DVD of CentOS4.5 with a ks.cfg file (made by anaconda during a previous installation). I placed it in the same directory as CentOS, isolinux, etc. (i.e. the top directory) and can do linux ks=cdrom:ks.cfg This works fine. After installation there are a couple other things I do: untar two files and install them. For instance, if the file is called xyz.tz I do a tar -zxvf xyz.tz cd xyz ./install these are my own install packages (one installs a VPN and the other just does stuff like add users, install java, etc). I want to use the %post of my ks.cfg file to do these automatically. On the same install disk, before doing the mkisofs, I made in the top directory a directory called "addon" and put my two tar files, preimage.tz and nppt.tz. Then from examples I modified the ks.cfg file like this: %post --nochroot mkdir -p /mnt/cdrom mount /tmp/cdrom /mnt/cdrom mkdir /tmp/inst cp /mnt/cdrom/addon/*.tz /tmp/inst/ cd /tmp/inst tar -zxvf preimage.tz cd preimage ./install cd .. tar -zxvf nppt.tz cd nppt ./install I wanted to access the addon directory on the CD and from what I could tell to do that you had to do the mount of /tmp/cdrom. During install it said it was executing the postinstall and I did not see any errors, but nothing worked. There was no /mnt/cdrom directory, and nothign was installed. Unfortunately, the install did not show any errors and the stuff was not in /root/install.log. I can experiment and do things like df >> /root/mylog.txt and mount /tmp/cdrom /mnt/cdrom 2>&1 >> /roote/mylog.txt to see where the files from the CD really are and what is going wrong, but every time I do debugging I have to waste another DVD making an installation DVD. So two questions: 1. How do I access the addon directory on the DVD? I know it is there because on a Linux system I put the DVD in the reader and see it. 2. How can I test thet %post without wasting a DVD each time? From Chip.Shabazian at bankofamerica.com Tue Nov 4 04:20:18 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Mon, 03 Nov 2008 20:20:18 -0800 Subject: extra packages In-Reply-To: References: Message-ID: Put a "sleep 9999999999" in your %post, then go to the alt-f2 console and take a look around. I forget where it is, but I believe it's already mounted, so you can just copy it using the --nochroot %post first. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of tony.chamberlain at lemko.com Sent: Monday, November 03, 2008 7:01 PM To: kickstart-list at redhat.com Subject: extra packages After a bit of time I finally was able to make an installation DVD of CentOS4.5 with a ks.cfg file (made by anaconda during a previous installation). I placed it in the same directory as CentOS, isolinux, etc. (i.e. the top directory) and can do linux ks=cdrom:ks.cfg This works fine. After installation there are a couple other things I do: untar two files and install them. For instance, if the file is called xyz.tz I do a tar -zxvf xyz.tz cd xyz ./install these are my own install packages (one installs a VPN and the other just does stuff like add users, install java, etc). I want to use the %post of my ks.cfg file to do these automatically. On the same install disk, before doing the mkisofs, I made in the top directory a directory called "addon" and put my two tar files, preimage.tz and nppt.tz. Then from examples I modified the ks.cfg file like this: %post --nochroot mkdir -p /mnt/cdrom mount /tmp/cdrom /mnt/cdrom mkdir /tmp/inst cp /mnt/cdrom/addon/*.tz /tmp/inst/ cd /tmp/inst tar -zxvf preimage.tz cd preimage ./install cd .. tar -zxvf nppt.tz cd nppt ./install I wanted to access the addon directory on the CD and from what I could tell to do that you had to do the mount of /tmp/cdrom. During install it said it was executing the postinstall and I did not see any errors, but nothing worked. There was no /mnt/cdrom directory, and nothign was installed. Unfortunately, the install did not show any errors and the stuff was not in /root/install.log. I can experiment and do things like df >> /root/mylog.txt and mount /tmp/cdrom /mnt/cdrom 2>&1 >> /roote/mylog.txt to see where the files from the CD really are and what is going wrong, but every time I do debugging I have to waste another DVD making an installation DVD. So two questions: 1. How do I access the addon directory on the DVD? I know it is there because on a Linux system I put the DVD in the reader and see it. 2. How can I test thet %post without wasting a DVD each time? _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From jarle at bjorgeengen.net Tue Nov 4 10:02:17 2008 From: jarle at bjorgeengen.net (=?ISO-8859-1?Q?Jarle_Bj=F8rgeengen?=) Date: Tue, 4 Nov 2008 11:02:17 +0100 Subject: extra packages In-Reply-To: References: Message-ID: <373754DD-8DD2-4BC3-A7DD-3DAA20F6AA3D@bjorgeengen.net> On Nov 4, 2008, at 5:20 , Shabazian, Chip wrote: > > tar -zxvf xyz.tz > cd xyz > ./install > > these are my own install packages (one installs a VPN and the other > just > does stuff like add users, install java, etc). > > I want to use the %post of my ks.cfg file to do these automatically. > On > the same install disk, before doing the mkisofs, I made in the top > directory a directory called "addon" and put my two tar files, > preimage.tz and nppt.tz. Then from examples I modified the ks.cfg file > like this: > > > %post --nochroot If you do this .... > ./install > ...this will install your software to the booted ramdisk , not to the mount sysimage. Maybe you need to change the install script so that it installs everything relative to /mnt/sysimage when you don't chroot during the %post -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Best regards .~. Jarle Bj?rgeengen /V\ Mob: +47 9155 7978 // \\ http://www.uio.no/sok?person=jb /( )\ while(<>){if(s/^(.*\?)$/42 !/){print "$1 $_"}} ^`~'^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From tony.chamberlain at lemko.com Tue Nov 4 14:14:10 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Tue, 04 Nov 2008 14:14:10 +0000 Subject: extra packages Message-ID: You seem not to be able to do ALT-F2 during the %post part. And it went very quickly (I just did sleep 5000 which I hope is about 8 minutes). It went way quicker than that, but the "Installing bootloader" is taking forver. I wonder whether my sleep statement is taking place there instead? I did do a couplf of "echo" and "ls" and "pwd" in the %post and redirected the output so I hope if this thing finishes that I will be able to read the log stuff. -----Original Message----- From: Shabazian, Chip [mailto:Chip.Shabazian at bankofamerica.com] Sent: Monday, November 3, 2008 10:20 PM To: 'Discussion list about Kickstart' Subject: RE: extra packages Put a "sleep 9999999999" in your %post, then go to the alt-f2 console and take a look around. I forget where it is, but I believe it's already mounted, so you can just copy it using the --nochroot %post first. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of tony.chamberlain at lemko.com Sent: Monday, November 03, 2008 7:01 PM To: kickstart-list at redhat.com Subject: extra packages After a bit of time I finally was able to make an installation DVD of CentOS4.5 with a ks.cfg file (made by anaconda during a previous installation). I placed it in the same directory as CentOS, isolinux, etc. (i.e. the top directory) and can do linux ks=cdrom:ks.cfg This works fine. After installation there are a couple other things I do: untar two files and install them. For instance, if the file is called xyz.tz I do a tar -zxvf xyz.tz cd xyz ./install these are my own install packages (one installs a VPN and the other just does stuff like add users, install java, etc). I want to use the %post of my ks.cfg file to do these automatically. On the same install disk, before doing the mkisofs, I made in the top directory a directory called "addon" and put my two tar files, preimage.tz and nppt.tz. Then from examples I modified the ks.cfg file like this: %post --nochroot mkdir -p /mnt/cdrom mount /tmp/cdrom /mnt/cdrom mkdir /tmp/inst cp /mnt/cdrom/addon/*.tz /tmp/inst/ cd /tmp/inst tar -zxvf preimage.tz cd preimage ./install cd .. tar -zxvf nppt.tz cd nppt ./install I wanted to access the addon directory on the CD and from what I could tell to do that you had to do the mount of /tmp/cdrom. During install it said it was executing the postinstall and I did not see any errors, but nothing worked. There was no /mnt/cdrom directory, and nothign was installed. Unfortunately, the install did not show any errors and the stuff was not in /root/install.log. I can experiment and do things like df >> /root/mylog.txt and mount /tmp/cdrom /mnt/cdrom 2>&1 >> /roote/mylog.txt to see where the files from the CD really are and what is going wrong, but every time I do debugging I have to waste another DVD making an installation DVD. So two questions: 1. How do I access the addon directory on the DVD? I know it is there because on a Linux system I put the DVD in the reader and see it. 2. How can I test thet %post without wasting a DVD each time? _______________________________________________ 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 ugob at lubik.ca Tue Nov 4 14:48:28 2008 From: ugob at lubik.ca (Ugo Bellavance) Date: Tue, 04 Nov 2008 09:48:28 -0500 Subject: Anaconda not looking for the right package In-Reply-To: <490F701A.9080807@redhat.com> References: <490F701A.9080807@redhat.com> Message-ID: Kyle Powell a ?crit : > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Ugo Bellavance wrote: >> Hi, >> >> I'm trying to install RHEL 5 U2 via kickstart, but I get an error saying >> "The file sysklogd-1.4.1-32.2.x86_64.rpm cannot be opened. >> >> It is right, the file is not there, the file that is there is >> sysklogd-1.4.1-44.el5.x86_64.rpm. > > This is correct. The package that ships with RHEL 5 U2 is > sysklogd-1.4.1-44.el5.x86_64.rpm That is what I thought as well... >> Please let me know if you need more info. > > If you'd like to provide your kickstart file, we may be able to assist you further. Here it is ======= install url --url http://192.168.x.x/pxe/rhel564 key 6f6ce0d56ee72ce0 lang en_US.UTF-8 keyboard us network --device eth0 --bootproto dhcp --hostname hostname.domain.com rootpw --iscrypted $$$$crypted firewall --ssh authconfig --enableshadow --enablemd5 vnc --host=192.168.x.x selinux --disabled timezone America/Montreal bootloader --location=mbr --driveorder=cciss/c0d0 # 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=cciss/c0d0 part /boot --fstype ext3 --size=100 --ondisk=cciss/c0d0 part pv.5 --size=0 --grow --ondisk=cciss/c0d0 volgroup VolGroup00 --pesize=32768 pv.5 logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=8192 logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=689984 %packages @ virtualization ========================= Thanks a lot, Ugo From kpowell at redhat.com Tue Nov 4 15:56:21 2008 From: kpowell at redhat.com (Kyle Powell) Date: Tue, 04 Nov 2008 10:56:21 -0500 Subject: Anaconda not looking for the right package In-Reply-To: References: <490F701A.9080807@redhat.com> Message-ID: <491070A5.5040706@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Ugo Bellavance wrote: > url --url http://192.168.x.x/pxe/rhel564 Since sysklogd-1.4.1-39.2 is the version of sysklogd that shipped with RHEL 5.0 (I can find no record of a sysklogd-1.4.1-32.2, are you sure it didn't report "The file sysklogd-1.4.1-39.2.x86_64.rpm cannot be opened"?), my suspicion is you have some old files from RHEL 5.0 at http://192.168.x.x/pxe/rhel564. Specifically primary.xml.gz, which is what tells anaconda what package versions it should use. Go to the host (192.168.x.x) and check the contents of the pxe/rhel564/repodata directory. Check the md5sums of the files. Here are the md5sums for RHEL 5.2 Server: c08afbc263c71a75bedd847ddfe6d5ce comps-rhel5-server-core.xml 3afdf321976aacb13bc0659e5874b0b3 filelists.xml.gz 95f121d308260d52bba6a6a8102b258e other.xml.gz 18b04e5dcfbe64eba59fea6e3750e9e2 primary.xml.gz 8816b87b53ffc8c6ff542a3728b1941c repomd.xml a966fa06180ba06fdda5178edbf95c75 TRANS.TBL And these are the md5sums for the RHEL 5 Server files: fbf1f3f9808d0c834fd1ca01b3dab24c comps-rhel5-server-core.xml 19f4a40e247715bf6ab6e84d87d8bfb9 filelists.xml.gz 028bad2bc9d0d818a5754540f0579539 other.xml.gz ebf520b937716327bf60f3ca887bb0ff primary.xml.gz 9ce1b5fcd721c0ada03f44dd3cab58c9 repomd.xml a966fa06180ba06fdda5178edbf95c75 TRANS.TBL If they're the older versions, you can grab the correct versions from the RHEL 5.2 iso. - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJEHCl7pTtanQdBU4RAufiAJ9/6iY7q1dvXUJICDzhRAcZSj2nVwCfeWbg k4ZtRnIGUmKBd6IWFX6txHM= =2fXz -----END PGP SIGNATURE----- From ugob at lubik.ca Tue Nov 4 16:43:02 2008 From: ugob at lubik.ca (Ugo Bellavance) Date: Tue, 04 Nov 2008 11:43:02 -0500 Subject: Anaconda not looking for the right package In-Reply-To: <491070A5.5040706@redhat.com> References: <490F701A.9080807@redhat.com> <491070A5.5040706@redhat.com> Message-ID: Kyle Powell a ?crit : > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Ugo Bellavance wrote: >> url --url http://192.168.x.x/pxe/rhel564 > > Since sysklogd-1.4.1-39.2 is the version of sysklogd that shipped with RHEL 5.0 > (I can find no record of a sysklogd-1.4.1-32.2, are you sure it didn't report > "The file sysklogd-1.4.1-39.2.x86_64.rpm cannot be opened"?), my suspicion is > you have some old files from RHEL 5.0 at http://192.168.x.x/pxe/rhel564. > Specifically primary.xml.gz, which is what tells anaconda what package versions > it should use. Go to the host (192.168.x.x) and check the contents of the > pxe/rhel564/repodata directory. Check the md5sums of the files. > > If they're the older versions, you can grab the correct versions from the RHEL > 5.2 iso. They're the right version on the filesystem, but when I access them via apache, the files are totally different. The change I did is that I mounted the 5.2 iso instead of the 5.0. I restarted httpd a few times since, but I'm confused about what could that happen. I've deleted my browser cache and I'm not going through a proxy. Any idea why that could happen? Thanks, Ugo From kpowell at redhat.com Tue Nov 4 17:39:54 2008 From: kpowell at redhat.com (Kyle Powell) Date: Tue, 04 Nov 2008 12:39:54 -0500 Subject: extra packages In-Reply-To: References: Message-ID: <491088EA.4070304@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 tony.chamberlain at lemko.com wrote: > You seem not to be able to do ALT-F2 during the %post part. And it > went very quickly (I just did sleep 5000 which I hope is about 8 minutes). > It went way quicker than that, but the "Installing bootloader" is taking > forver. I wonder whether my sleep statement is taking place there > instead? No, you certainly should be able to switch to VT 2 while the %post script is running. If anaconda seems to be finishing the post script quickly and you have a 'sleep 5000' in your post script, this would indicate the post script is not running completely, at least not far enough to reach the sleep (sleep 5000 would sleep for 5000 seconds, which is nearly an hour and a half). > %post --nochroot > mkdir -p /mnt/cdrom > mount /tmp/cdrom /mnt/cdrom > mkdir /tmp/inst > cp /mnt/cdrom/addon/*.tz /tmp/inst/ > cd /tmp/inst > tar -zxvf preimage.tz > cd preimage > ./install > cd .. > tar -zxvf nppt.tz > cd nppt > ./install > > 1. How do I access the addon directory on the DVD? I know it is there > because on a Linux system I put the DVD in the reader and see it. When installing from CD or DVD, it will already be mounted at /mnt/source. I would recommend you use two post scripts. Copy the files from the CD in a - --nochroot post script, then run the install in a regular (chroot'ed) post script. %post --nochroot mkdir -p /mnt/sysimage/tmp/inst cp /mnt/source/addon/* /mnt/sysimage/tmp/inst %post cd /tmp/inst tar -zxvf preimage.tz cd preimage ./install cd .. tar -zxvf nppt.tz cd nppt ./install > 2. How can I test thet %post without wasting a DVD each time? If you have a second machine you could pull the ks.cfg via http, or if you have a floppy, you could get the kickstart file from the floppy drive. These would allow you to change the kickstart file without burning a new DVD each time. If you absolutely want to pull the kickstart file from the iso each time (instead of just changing the ks= line to point to the cdrom once all your testing is complete) you could kickstart a virtual machine and present the iso file to the VM without actually burning a DVD. - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJEIjq7pTtanQdBU4RArnMAJ9IkpJ7MTibLhB+QXe2XWWLNAEZ7ACgh2hs JZ1VNQZe9yH2J5dWs2qvWD4= =WtHD -----END PGP SIGNATURE----- From kpowell at redhat.com Tue Nov 4 17:42:32 2008 From: kpowell at redhat.com (Kyle Powell) Date: Tue, 04 Nov 2008 12:42:32 -0500 Subject: Anaconda not looking for the right package In-Reply-To: References: <490F701A.9080807@redhat.com> <491070A5.5040706@redhat.com> Message-ID: <49108988.40309@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Ugo Bellavance wrote: > > They're the right version on the filesystem, but when I access them via > apache, the files are totally different. The change I did is that I > mounted the 5.2 iso instead of the 5.0. I restarted httpd a few times > since, but I'm confused about what could that happen. I've deleted my > browser cache and I'm not going through a proxy. > > Any idea why that could happen? Nope. I have no idea why apache would send you a different file than what's on the filesystem. - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJEImI7pTtanQdBU4RAoHsAKCIYh0F6EkGZR1I1zwfSREQowe7UACcCsDj L4pJSnXo5lYuADzXV36Iv+8= =A9A/ -----END PGP SIGNATURE----- From william_dennison at sra.com Tue Nov 4 18:06:13 2008 From: william_dennison at sra.com (William J. Dennison) Date: Tue, 4 Nov 2008 13:06:13 -0500 Subject: LOGVOL --grow Seems to Be Broken Message-ID: <200811041306.13470.william_dennison@sra.com> The Kickstart documentation located here (http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation_Guide-en-US/s1-kickstart2-options.html) specifically notes that you can use "--grow" with a "logvol" parameter. When I do this the RHEL5 installation blows up. Is this an issue with Cobbler, Kickstart, RHEL 5 limitations, etc.? Below is my partitioning area in my kickstart file. -------------------------------------------------------------------------------------------------------------------- clearpart --all --initlabel partition /boot --size=1000 --ondisk=sda --asprimary --fstype=ext3 partition pv.01 --size=22000 --ondisk=sda --asprimary partition pv.02 --size=1024 --grow --ondisk=sda --asprimary volgroup vg_system pv.01 volgroup vg_data pv.02 logvol / --vgname=vg_system --size=2000 --name=lv_root --fstype=ext3 logvol /swap --vgname=vg_system --size=4000 --name=lv_swap --fstype=swap logvol /usr --vgname=vg_system --size=4000 --name=lv_usr --fstype=ext3 logvol /var --vgname=vg_system --size=2000 --name=lv_var --fstype=ext3 logvol /var_log --vgname=vg_system --size=2000 --name=lv_var_log --fstype=ext3 logvol /var_log_audit --vgname=vg_system --size=2000 --name=lv_var_log_audit -- fstype=ext3 logvol /var/tmp --vgname=vg_system --size=2000 --name=lv_var_tmp --fstype=ext3 logvol /tmp --vgname=vg_system --size=2000 --name=lv_tmp --fstype=ext3 logvol /opt --vgname=vg_system --size=1000 --name=lv_opt --fstype=ext3 logvol /home --vgname=vg_data --size=10000 --name=lv_home --fstype=ext3 logvol /opt/store/incomming_data --vgname=vg_data --size=1000 --name=lv_incom ming_data --fstype=ext3 logvol /opt/store/saved_data --vgname=vg_data --size=1000 --name=lv_saved_data --fstype=ext3 logvol /opt/store/upload_data --vgname=vg_data --size=1000 --name=lv_upload_data --fstype=ext3 logvol /opt/store/export_data --vgname=vg_data --size=1000 --name=lv_export_data --fstype=ext3 logvol /opt/store/webserver --vgname=vg_data --size=1000 --name=lv_appserver --fstype=ext3 logvol /opt/store/dbserver --vgname=vg_data --size=1000 --name=lv_dbserver --fstype=ext3 logvol /opt/store/virtual_machines --vgname=vg_data --size=1 --grow --name=lv_virtual_machines --fstype=ext3 ---------------------------------------------------------------------------------------------------------------------------------------------------------------- If I remove the --grow from "virtual_machines", all works fine, if I leave it in the Anaconda load blows up. When I run an interactive install, I notice that the --grow feature is not allowed on logvols. ?Does the logvol RED HAT documentation have to be edited? -- William J. Dennison Principal Consultant SRA International Inc. - RABA Center 8830 Stanford Blvd, Suite 205 Columbia, Maryland 21045 Phone: (410) 715-9399 x2351 Mobile: (410) 218-5866 Email: william_dennison at sra.com Email2: wjdennison at gmail.com Email3: wjdennison at cablespeed.com Jabber: wjdennison at jabber.org AIM: dennisonwjonly GTalk: wjdennison at gmail.com From CallahanT at tessco.com Tue Nov 4 18:15:12 2008 From: CallahanT at tessco.com (Tom Callahan) Date: Tue, 04 Nov 2008 13:15:12 -0500 Subject: LOGVOL --grow Seems to Be Broken In-Reply-To: <200811041306.13470.william_dennison@sra.com> Message-ID: It looks like --grow requires an attribute, documentation shows "--grow=" Try --grow=max maybe? Or --grow=y I have not played with --grow on logvol yet. On 11/4/08 1:06 PM, "William J. Dennison" wrote: > The Kickstart documentation located here > (http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation_Guid > e-en-US/s1-kickstart2-options.html) > specifically notes that you can use "--grow" with a "logvol" parameter. When > I do this the RHEL5 installation blows up. Is this an issue with Cobbler, > Kickstart, RHEL 5 limitations, etc.? > > Below is my partitioning area in my kickstart file. > > ------------------------------------------------------------------------------ > -------------------------------------- > clearpart --all --initlabel > > partition /boot --size=1000 --ondisk=sda --asprimary --fstype=ext3 > > partition pv.01 --size=22000 --ondisk=sda --asprimary > partition pv.02 --size=1024 --grow --ondisk=sda --asprimary > > volgroup vg_system pv.01 > volgroup vg_data pv.02 > > logvol / --vgname=vg_system --size=2000 --name=lv_root --fstype=ext3 > logvol /swap --vgname=vg_system --size=4000 --name=lv_swap --fstype=swap > logvol /usr --vgname=vg_system --size=4000 --name=lv_usr --fstype=ext3 > logvol /var --vgname=vg_system --size=2000 --name=lv_var --fstype=ext3 > logvol /var_log --vgname=vg_system --size=2000 --name=lv_var_log --fstype=ext3 > logvol /var_log_audit --vgname=vg_system --size=2000 --name=lv_var_log_audit > -- > fstype=ext3 > logvol /var/tmp --vgname=vg_system --size=2000 --name=lv_var_tmp --fstype=ext3 > logvol /tmp --vgname=vg_system --size=2000 --name=lv_tmp --fstype=ext3 > logvol /opt --vgname=vg_system --size=1000 --name=lv_opt --fstype=ext3 > > logvol /home --vgname=vg_data --size=10000 --name=lv_home --fstype=ext3 > > logvol /opt/store/incomming_data --vgname=vg_data --size=1000 --name=lv_incom > ming_data --fstype=ext3 > logvol /opt/store/saved_data --vgname=vg_data --size=1000 --name=lv_saved_data > --fstype=ext3 > logvol /opt/store/upload_data --vgname=vg_data --size=1000 > --name=lv_upload_data --fstype=ext3 > logvol /opt/store/export_data --vgname=vg_data --size=1000 > --name=lv_export_data --fstype=ext3 > logvol /opt/store/webserver --vgname=vg_data --size=1000 --name=lv_appserver > --fstype=ext3 > logvol /opt/store/dbserver --vgname=vg_data --size=1000 --name=lv_dbserver > --fstype=ext3 > logvol /opt/store/virtual_machines --vgname=vg_data --size=1 --grow > --name=lv_virtual_machines --fstype=ext3 > ------------------------------------------------------------------------------ > ------------------------------------------------------------------------------ > ---- > > If I remove the --grow from "virtual_machines", all works fine, if I leave it > in the Anaconda load blows up. When I run an interactive install, I notice > that the --grow feature is not allowed on logvols. ?Does the logvol RED HAT > documentation have to be edited? From CallahanT at tessco.com Tue Nov 4 18:18:11 2008 From: CallahanT at tessco.com (Tom Callahan) Date: Tue, 04 Nov 2008 13:18:11 -0500 Subject: LOGVOL --grow Seems to Be Broken In-Reply-To: Message-ID: Doing some quick googling of "logvol" AND "--grow" Gives me a couple examples where people are using just --grow on logvol definitions. Try making it the last argument in the definition string. logvol /opt/store/virtual_machines --vgname=vg_data --name=lv_virtual_machines --fstype=ext3 --size=1 --grow On 11/4/08 1:15 PM, "Thomas Callahan" wrote: > It looks like --grow requires an attribute, documentation shows "--grow=" > > Try --grow=max maybe? Or --grow=y > > I have not played with --grow on logvol yet. > > > > On 11/4/08 1:06 PM, "William J. Dennison" wrote: > >> The Kickstart documentation located here >> (http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation_Gui>> d >> e-en-US/s1-kickstart2-options.html) >> specifically notes that you can use "--grow" with a "logvol" parameter. When >> I do this the RHEL5 installation blows up. Is this an issue with Cobbler, >> Kickstart, RHEL 5 limitations, etc.? >> >> Below is my partitioning area in my kickstart file. >> >> ----------------------------------------------------------------------------->> - >> -------------------------------------- >> clearpart --all --initlabel >> >> partition /boot --size=1000 --ondisk=sda --asprimary --fstype=ext3 >> >> partition pv.01 --size=22000 --ondisk=sda --asprimary >> partition pv.02 --size=1024 --grow --ondisk=sda --asprimary >> >> volgroup vg_system pv.01 >> volgroup vg_data pv.02 >> >> logvol / --vgname=vg_system --size=2000 --name=lv_root --fstype=ext3 >> logvol /swap --vgname=vg_system --size=4000 --name=lv_swap --fstype=swap >> logvol /usr --vgname=vg_system --size=4000 --name=lv_usr --fstype=ext3 >> logvol /var --vgname=vg_system --size=2000 --name=lv_var --fstype=ext3 >> logvol /var_log --vgname=vg_system --size=2000 --name=lv_var_log >> --fstype=ext3 >> logvol /var_log_audit --vgname=vg_system --size=2000 --name=lv_var_log_audit >> -- >> fstype=ext3 >> logvol /var/tmp --vgname=vg_system --size=2000 --name=lv_var_tmp >> --fstype=ext3 >> logvol /tmp --vgname=vg_system --size=2000 --name=lv_tmp --fstype=ext3 >> logvol /opt --vgname=vg_system --size=1000 --name=lv_opt --fstype=ext3 >> >> logvol /home --vgname=vg_data --size=10000 --name=lv_home --fstype=ext3 >> >> logvol /opt/store/incomming_data --vgname=vg_data --size=1000 --name=lv_incom >> ming_data --fstype=ext3 >> logvol /opt/store/saved_data --vgname=vg_data --size=1000 >> --name=lv_saved_data >> --fstype=ext3 >> logvol /opt/store/upload_data --vgname=vg_data --size=1000 >> --name=lv_upload_data --fstype=ext3 >> logvol /opt/store/export_data --vgname=vg_data --size=1000 >> --name=lv_export_data --fstype=ext3 >> logvol /opt/store/webserver --vgname=vg_data --size=1000 --name=lv_appserver >> --fstype=ext3 >> logvol /opt/store/dbserver --vgname=vg_data --size=1000 --name=lv_dbserver >> --fstype=ext3 >> logvol /opt/store/virtual_machines --vgname=vg_data --size=1 --grow >> --name=lv_virtual_machines --fstype=ext3 >> ----------------------------------------------------------------------------->> - >> ----------------------------------------------------------------------------->> - >> ---- >> >> If I remove the --grow from "virtual_machines", all works fine, if I leave it >> in the Anaconda load blows up. When I run an interactive install, I notice >> that the --grow feature is not allowed on logvols. ?Does the logvol RED HAT >> documentation have to be edited? > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From william_dennison at sra.com Tue Nov 4 18:25:32 2008 From: william_dennison at sra.com (William J. Dennison) Date: Tue, 4 Nov 2008 13:25:32 -0500 Subject: LOGVOL --grow Seems to Be Broken In-Reply-To: References: Message-ID: <200811041325.32617.william_dennison@sra.com> You are correct sir, I completely missed the "=". However it would really be nice to know what = should be. In the documentation it says "--grow= Tells the logical volume to grow to fill available space (if any), or up to the maximum size setting. My sda3 partition is 250GB The volume group is the entire size, I am using 13GB or so prior to the call for "virtual machines", therefore the "virtual_machines" partition should be 230GB or so. On Tuesday 04 November 2008 01:15:12 pm Tom Callahan wrote: > It looks like --grow requires an attribute, documentation shows "--grow=" > > Try --grow=max maybe? Or --grow=y > > I have not played with --grow on logvol yet. > > On 11/4/08 1:06 PM, "William J. Dennison" wrote: > > The Kickstart documentation located here > > (http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation > >_Guid e-en-US/s1-kickstart2-options.html) > > specifically notes that you can use "--grow" with a "logvol" parameter. > > When I do this the RHEL5 installation blows up. Is this an issue with > > Cobbler, Kickstart, RHEL 5 limitations, etc.? > > > > Below is my partitioning area in my kickstart file. > > > > ------------------------------------------------------------------------- > >----- -------------------------------------- > > clearpart --all --initlabel > > > > partition /boot --size=1000 --ondisk=sda --asprimary --fstype=ext3 > > > > partition pv.01 --size=22000 --ondisk=sda --asprimary > > partition pv.02 --size=1024 --grow --ondisk=sda --asprimary > > > > volgroup vg_system pv.01 > > volgroup vg_data pv.02 > > > > logvol / --vgname=vg_system --size=2000 --name=lv_root --fstype=ext3 > > logvol /swap --vgname=vg_system --size=4000 --name=lv_swap --fstype=swap > > logvol /usr --vgname=vg_system --size=4000 --name=lv_usr --fstype=ext3 > > logvol /var --vgname=vg_system --size=2000 --name=lv_var --fstype=ext3 > > logvol /var_log --vgname=vg_system --size=2000 --name=lv_var_log > > --fstype=ext3 logvol /var_log_audit --vgname=vg_system --size=2000 > > --name=lv_var_log_audit -- > > fstype=ext3 > > logvol /var/tmp --vgname=vg_system --size=2000 --name=lv_var_tmp > > --fstype=ext3 logvol /tmp --vgname=vg_system --size=2000 --name=lv_tmp > > --fstype=ext3 logvol /opt --vgname=vg_system --size=1000 --name=lv_opt > > --fstype=ext3 > > > > logvol /home --vgname=vg_data --size=10000 --name=lv_home --fstype=ext3 > > > > logvol /opt/store/incomming_data --vgname=vg_data --size=1000 > > --name=lv_incom ming_data --fstype=ext3 > > logvol /opt/store/saved_data --vgname=vg_data --size=1000 > > --name=lv_saved_data --fstype=ext3 > > logvol /opt/store/upload_data --vgname=vg_data --size=1000 > > --name=lv_upload_data --fstype=ext3 > > logvol /opt/store/export_data --vgname=vg_data --size=1000 > > --name=lv_export_data --fstype=ext3 > > logvol /opt/store/webserver --vgname=vg_data --size=1000 > > --name=lv_appserver --fstype=ext3 > > logvol /opt/store/dbserver --vgname=vg_data --size=1000 > > --name=lv_dbserver --fstype=ext3 > > logvol /opt/store/virtual_machines --vgname=vg_data --size=1 --grow > > --name=lv_virtual_machines --fstype=ext3 > > ------------------------------------------------------------------------- > >----- > > ------------------------------------------------------------------------- > >----- ---- > > > > If I remove the --grow from "virtual_machines", all works fine, if I > > leave it in the Anaconda load blows up. When I run an interactive > > install, I notice that the --grow feature is not allowed on logvols. > > ?Does the logvol RED HAT documentation have to be edited? > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- William J. Dennison Principal Consultant SRA International Inc. - RABA Center 8830 Stanford Blvd, Suite 205 Columbia, Maryland 21045 Phone: (410) 715-9399 x2351 Mobile: (410) 218-5866 Email: william_dennison at sra.com Email2: wjdennison at gmail.com Email3: wjdennison at cablespeed.com Jabber: wjdennison at jabber.org AIM: dennisonwjonly GTalk: wjdennison at gmail.com From william_dennison at sra.com Tue Nov 4 18:35:51 2008 From: william_dennison at sra.com (William J. Dennison) Date: Tue, 4 Nov 2008 13:35:51 -0500 Subject: LOGVOL --grow Seems to Be Broken In-Reply-To: References: Message-ID: <200811041335.51456.william_dennison@sra.com> Attempted and failed, the error is "lvcreate failed for lv_upload_data" REBOOT, this is the same error from prior attempts. On Tuesday 04 November 2008 01:18:11 pm Tom Callahan wrote: > Doing some quick googling of > > "logvol" AND "--grow" > > Gives me a couple examples where people are using just --grow on logvol > definitions. Try making it the last argument in the definition string. > > logvol /opt/store/virtual_machines --vgname=vg_data > --name=lv_virtual_machines --fstype=ext3 --size=1 --grow > > On 11/4/08 1:15 PM, "Thomas Callahan" wrote: > > It looks like --grow requires an attribute, documentation shows "--grow=" > > > > Try --grow=max maybe? Or --grow=y > > > > I have not played with --grow on logvol yet. > > > > On 11/4/08 1:06 PM, "William J. Dennison" wrote: > >> The Kickstart documentation located here > > (http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation_G >ui>> d > > >> e-en-US/s1-kickstart2-options.html) > >> specifically notes that you can use "--grow" with a "logvol" parameter. > >> When I do this the RHEL5 installation blows up. Is this an issue with > >> Cobbler, Kickstart, RHEL 5 limitations, etc.? > >> > >> Below is my partitioning area in my kickstart file. > > --------------------------------------------------------------------------- >-->> - > > >> -------------------------------------- > >> clearpart --all --initlabel > >> > >> partition /boot --size=1000 --ondisk=sda --asprimary --fstype=ext3 > >> > >> partition pv.01 --size=22000 --ondisk=sda --asprimary > >> partition pv.02 --size=1024 --grow --ondisk=sda --asprimary > >> > >> volgroup vg_system pv.01 > >> volgroup vg_data pv.02 > >> > >> logvol / --vgname=vg_system --size=2000 --name=lv_root --fstype=ext3 > >> logvol /swap --vgname=vg_system --size=4000 --name=lv_swap --fstype=swap > >> logvol /usr --vgname=vg_system --size=4000 --name=lv_usr --fstype=ext3 > >> logvol /var --vgname=vg_system --size=2000 --name=lv_var --fstype=ext3 > >> logvol /var_log --vgname=vg_system --size=2000 --name=lv_var_log > >> --fstype=ext3 > >> logvol /var_log_audit --vgname=vg_system --size=2000 > >> --name=lv_var_log_audit -- > >> fstype=ext3 > >> logvol /var/tmp --vgname=vg_system --size=2000 --name=lv_var_tmp > >> --fstype=ext3 > >> logvol /tmp --vgname=vg_system --size=2000 --name=lv_tmp --fstype=ext3 > >> logvol /opt --vgname=vg_system --size=1000 --name=lv_opt --fstype=ext3 > >> > >> logvol /home --vgname=vg_data --size=10000 --name=lv_home --fstype=ext3 > >> > >> logvol /opt/store/incomming_data --vgname=vg_data --size=1000 > >> --name=lv_incom ming_data --fstype=ext3 > >> logvol /opt/store/saved_data --vgname=vg_data --size=1000 > >> --name=lv_saved_data > >> --fstype=ext3 > >> logvol /opt/store/upload_data --vgname=vg_data --size=1000 > >> --name=lv_upload_data --fstype=ext3 > >> logvol /opt/store/export_data --vgname=vg_data --size=1000 > >> --name=lv_export_data --fstype=ext3 > >> logvol /opt/store/webserver --vgname=vg_data --size=1000 > >> --name=lv_appserver --fstype=ext3 > >> logvol /opt/store/dbserver --vgname=vg_data --size=1000 > >> --name=lv_dbserver --fstype=ext3 > >> logvol /opt/store/virtual_machines --vgname=vg_data --size=1 --grow > >> --name=lv_virtual_machines --fstype=ext3 > > --------------------------------------------------------------------------- >-->> - > > --------------------------------------------------------------------------- >-->> - > > >> ---- > >> > >> If I remove the --grow from "virtual_machines", all works fine, if I > >> leave it in the Anaconda load blows up. When I run an interactive > >> install, I notice that the --grow feature is not allowed on logvols. > >> ?Does the logvol RED HAT documentation have to be edited? > > > > _______________________________________________ > > 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 -- William J. Dennison Principal Consultant SRA International Inc. - RABA Center 8830 Stanford Blvd, Suite 205 Columbia, Maryland 21045 Phone: (410) 715-9399 x2351 Mobile: (410) 218-5866 Email: william_dennison at sra.com Email2: wjdennison at gmail.com Email3: wjdennison at cablespeed.com Jabber: wjdennison at jabber.org AIM: dennisonwjonly GTalk: wjdennison at gmail.com From william_dennison at sra.com Tue Nov 4 18:26:12 2008 From: william_dennison at sra.com (William J. Dennison) Date: Tue, 4 Nov 2008 13:26:12 -0500 Subject: LOGVOL --grow Seems to Be Broken In-Reply-To: References: Message-ID: <200811041326.12273.william_dennison@sra.com> I will attempt that now. On Tuesday 04 November 2008 01:18:11 pm Tom Callahan wrote: > Doing some quick googling of > > "logvol" AND "--grow" > > Gives me a couple examples where people are using just --grow on logvol > definitions. Try making it the last argument in the definition string. > > logvol /opt/store/virtual_machines --vgname=vg_data > --name=lv_virtual_machines --fstype=ext3 --size=1 --grow > > On 11/4/08 1:15 PM, "Thomas Callahan" wrote: > > It looks like --grow requires an attribute, documentation shows "--grow=" > > > > Try --grow=max maybe? Or --grow=y > > > > I have not played with --grow on logvol yet. > > > > On 11/4/08 1:06 PM, "William J. Dennison" wrote: > >> The Kickstart documentation located here > > (http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation_G >ui>> d > > >> e-en-US/s1-kickstart2-options.html) > >> specifically notes that you can use "--grow" with a "logvol" parameter. > >> When I do this the RHEL5 installation blows up. Is this an issue with > >> Cobbler, Kickstart, RHEL 5 limitations, etc.? > >> > >> Below is my partitioning area in my kickstart file. > > --------------------------------------------------------------------------- >-->> - > > >> -------------------------------------- > >> clearpart --all --initlabel > >> > >> partition /boot --size=1000 --ondisk=sda --asprimary --fstype=ext3 > >> > >> partition pv.01 --size=22000 --ondisk=sda --asprimary > >> partition pv.02 --size=1024 --grow --ondisk=sda --asprimary > >> > >> volgroup vg_system pv.01 > >> volgroup vg_data pv.02 > >> > >> logvol / --vgname=vg_system --size=2000 --name=lv_root --fstype=ext3 > >> logvol /swap --vgname=vg_system --size=4000 --name=lv_swap --fstype=swap > >> logvol /usr --vgname=vg_system --size=4000 --name=lv_usr --fstype=ext3 > >> logvol /var --vgname=vg_system --size=2000 --name=lv_var --fstype=ext3 > >> logvol /var_log --vgname=vg_system --size=2000 --name=lv_var_log > >> --fstype=ext3 > >> logvol /var_log_audit --vgname=vg_system --size=2000 > >> --name=lv_var_log_audit -- > >> fstype=ext3 > >> logvol /var/tmp --vgname=vg_system --size=2000 --name=lv_var_tmp > >> --fstype=ext3 > >> logvol /tmp --vgname=vg_system --size=2000 --name=lv_tmp --fstype=ext3 > >> logvol /opt --vgname=vg_system --size=1000 --name=lv_opt --fstype=ext3 > >> > >> logvol /home --vgname=vg_data --size=10000 --name=lv_home --fstype=ext3 > >> > >> logvol /opt/store/incomming_data --vgname=vg_data --size=1000 > >> --name=lv_incom ming_data --fstype=ext3 > >> logvol /opt/store/saved_data --vgname=vg_data --size=1000 > >> --name=lv_saved_data > >> --fstype=ext3 > >> logvol /opt/store/upload_data --vgname=vg_data --size=1000 > >> --name=lv_upload_data --fstype=ext3 > >> logvol /opt/store/export_data --vgname=vg_data --size=1000 > >> --name=lv_export_data --fstype=ext3 > >> logvol /opt/store/webserver --vgname=vg_data --size=1000 > >> --name=lv_appserver --fstype=ext3 > >> logvol /opt/store/dbserver --vgname=vg_data --size=1000 > >> --name=lv_dbserver --fstype=ext3 > >> logvol /opt/store/virtual_machines --vgname=vg_data --size=1 --grow > >> --name=lv_virtual_machines --fstype=ext3 > > --------------------------------------------------------------------------- >-->> - > > --------------------------------------------------------------------------- >-->> - > > >> ---- > >> > >> If I remove the --grow from "virtual_machines", all works fine, if I > >> leave it in the Anaconda load blows up. When I run an interactive > >> install, I notice that the --grow feature is not allowed on logvols. > >> ?Does the logvol RED HAT documentation have to be edited? > > > > _______________________________________________ > > 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 -- William J. Dennison Principal Consultant SRA International Inc. - RABA Center 8830 Stanford Blvd, Suite 205 Columbia, Maryland 21045 Phone: (410) 715-9399 x2351 Mobile: (410) 218-5866 Email: william_dennison at sra.com Email2: wjdennison at gmail.com Email3: wjdennison at cablespeed.com Jabber: wjdennison at jabber.org AIM: dennisonwjonly GTalk: wjdennison at gmail.com From andrew.deagman at callfinity.com Tue Nov 4 18:23:54 2008 From: andrew.deagman at callfinity.com (Andrew Deagman) Date: Tue, 04 Nov 2008 13:23:54 -0500 Subject: Issue with RHEL 5 Kickstart hanging Message-ID: I have setup a kickstart server for RHEL 5. The install starts (pxeboot) properly, get?s an ip from dhcp, finds the ks.cfg (http) file and starts to load modules. It looks like modules load and then runs the anaconda script, as soon as that runs it just hangs. From the debug screen I have the following: INFO: Load module set done INFO: Running anaconda script /usr/bin/anaconda INFO: Using only Install Class Red Hat Enterprise Linux INFO: Using only Install Class Red Hat Enterprise Linux Then it will proceed to hang never timing out. I have tried installing directly from cd/dvd and that works fine. I have also tried cutting down the ks.cfg to a known working version and I get the same results. I can see the files being pulled from my http logs, but there are no errors... The install media is also located without any issues.. Any ideas as to what is going wrong? Thanks Andrew Copy of ks.cfg: #Kickstart file for RedHat ES 5 x64 rhel5x64-ks.cfg # Install from http url install # Location of the install media url --url http://10.10.10.1/kickstart/media/x64/ # Perform a text based install text # Set up Locale lang en_US # Keyboard Type keymap us # auto mouse mouse # Configure eth0 (should be dhcp) network --device eth0 --bootproto dhcp # Configure eth1 #network --device eth1 --onboot no --bootproto dhcp # Set the root password rootpw --iscrypted # Disable the firewall firewall --disabled # Use shadow passwords and md5 authconfig --enableshadow --enablemd5 # Enable SELinux set in permissive mode (logging only) selinux --permissive # Set the time zone timezone --utc America/New_York # Disbale firstboot screen firstboot --disabled # Do not set and installation key key --skip # Set the bootloader bootloader --location=mbr --driveorder=sda # Ignore all drives except sda ingnoredisk --drives=sdb,sdc,sdd,sde,sdf # Set the partition tables # Clear the partition table first clearpart --drives=sda --all --initlabel # Partition sda part /boot --fstype ext2 --size=128 --asprimary part swap --size=16384 --asprimary part / --fstype ext3 --size=6144 --asprimary part /var --fstype ext3 --size=4096 part /tmp --fstype ext3 --size=8192 part /opt --fstype ext3 --size=100 --grow # Packages to install %packages @base @legacy-software-support @development-libs - at base-x - at dialup - at admin-tools ntp ruby screen subversion mysql-connector-odbc unixODBC unixODBC-devel expat-devel gnutls-utils gnutls-devel curl-devel libtool-ltdl-devel libtool MySQL-python rcs rpm-build strace strace64 ltrace expect pfmon sysstat oprofile pstack elfutils gdb %post -- Andrew Deagman System Administrator | Callfinity Inc. E: andrew.deagman at callfinity.com O: 585-278-1940 x8126 M: 585-703-8621 Quietly implementing RFC 1925 wherever I go.... -------------- next part -------------- An HTML attachment was scrubbed... URL: From jarle at bjorgeengen.net Tue Nov 4 19:29:43 2008 From: jarle at bjorgeengen.net (=?ISO-8859-1?Q?Jarle_Bj=F8rgeengen?=) Date: Tue, 4 Nov 2008 20:29:43 +0100 Subject: extra packages In-Reply-To: <491088EA.4070304@redhat.com> References: <491088EA.4070304@redhat.com> Message-ID: On Nov 4, 2008, at 6:39 , Kyle Powell wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > tony.chamberlain at lemko.com wrote: >> You seem not to be able to do ALT-F2 during the %post part. And it >> went very quickly (I just did sleep 5000 which I hope is about 8 >> minutes). >> It went way quicker than that, but the "Installing bootloader" is >> taking >> forver. I wonder whether my sleep statement is taking place there >> instead? > > No, you certainly should be able to switch to VT 2 while the %post > script is > running. If anaconda seems to be finishing the post script quickly > and you have > a 'sleep 5000' in your post script, this would indicate the post > script is not > running completely, at least not far enough to reach the sleep > (sleep 5000 would > sleep for 5000 seconds, which is nearly an hour and a half). > >> %post --nochroot >> mkdir -p /mnt/cdrom >> mount /tmp/cdrom /mnt/cdrom >> mkdir /tmp/inst >> cp /mnt/cdrom/addon/*.tz /tmp/inst/ >> cd /tmp/inst >> tar -zxvf preimage.tz >> cd preimage >> ./install >> cd .. >> tar -zxvf nppt.tz >> cd nppt >> ./install >> >> 1. How do I access the addon directory on the DVD? I know it is >> there >> because on a Linux system I put the DVD in the reader and see it. > > When installing from CD or DVD, it will already be mounted at /mnt/ > source. I > would recommend you use two post scripts. Copy the files from the CD > in a > - --nochroot post script, then run the install in a regular > (chroot'ed) post script. > > %post --nochroot > mkdir -p /mnt/sysimage/tmp/inst > cp /mnt/source/addon/* /mnt/sysimage/tmp/inst > > %post > cd /tmp/inst > tar -zxvf preimage.tz > cd preimage > ./install > cd .. > tar -zxvf nppt.tz > cd nppt > ./install > Neat . Didn'r know you could have 2 %posts >> 2. How can I test thet %post without wasting a DVD each time? > > If you have a second machine you could pull the ks.cfg via http, or > if you have > a floppy, you could get the kickstart file from the floppy drive. > These would > allow you to change the kickstart file without burning a new DVD > each time. If > you absolutely want to pull the kickstart file from the iso each > time (instead > of just changing the ks= line to point to the cdrom once all your > testing is > complete) you could kickstart a virtual machine and present the iso > file to the > VM without actually burning a DVD. You can also use usb-stick linux ks=hd: Just edit isolinux config if you want this to be default. Once I pre-made a DVD image which wrote host-specific options to a temporary file in the %pre section, based on the output of dmidecode (machine serial number) , and later included the autogenerated file in the main ks.config. -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Best regards .~. Jarle Bj?rgeengen /V\ Mob: +47 9155 7978 // \\ http://www.uio.no/sok?person=jb /( )\ while(<>){if(s/^(.*\?)$/42 !/){print "$1 $_"}} ^`~'^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ From Tim.Mooney at ndsu.edu Tue Nov 4 20:22:22 2008 From: Tim.Mooney at ndsu.edu (Tim Mooney) Date: Tue, 4 Nov 2008 14:22:22 -0600 (CST) Subject: Issue with RHEL 5 Kickstart hanging In-Reply-To: References: Message-ID: In regard to: Issue with RHEL 5 Kickstart hanging, Andrew Deagman said (at...: > # Ignore all drives except sda > ingnoredisk --drives=sdb,sdc,sdd,sde,sdf Likely not your problem, but you have a typo there. Tim -- Tim Mooney Tim.Mooney at ndsu.edu Enterprise Computing & Infrastructure 701-231-1076 (Voice) Room 242-J6, IACC Building 701-231-8541 (Fax) North Dakota State University, Fargo, ND 58105-5164 From andrew.deagman at callfinity.com Tue Nov 4 20:34:45 2008 From: andrew.deagman at callfinity.com (Andrew Deagman) Date: Tue, 04 Nov 2008 15:34:45 -0500 Subject: Issue with RHEL 5 Kickstart hanging In-Reply-To: Message-ID: Thanks for catching that. However it does not fix my issue... Andrew On 11/4/08 3:22 PM, "Tim Mooney" wrote: > In regard to: Issue with RHEL 5 Kickstart hanging, Andrew Deagman said (at...: > >> > # Ignore all drives except sda >> > ingnoredisk --drives=sdb,sdc,sdd,sde,sdf > > Likely not your problem, but you have a typo there. > > Tim > -- > Tim Mooney Tim.Mooney at ndsu.edu > Enterprise Computing & Infrastructure 701-231-1076 (Voice) > Room 242-J6, IACC Building 701-231-8541 (Fax) > North Dakota State University, Fargo, ND 58105-5164 > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > > -- Andrew Deagman System Administrator | Callfinity Inc. E: andrew.deagman at callfinity.com O: 585-278-1940 x8126 M: 585-703-8621 Quietly implementing RFC 1925 wherever I go.... -------------- next part -------------- An HTML attachment was scrubbed... URL: From ugob at lubik.ca Tue Nov 4 21:53:05 2008 From: ugob at lubik.ca (Ugo Bellavance) Date: Tue, 04 Nov 2008 16:53:05 -0500 Subject: Anaconda not looking for the right package In-Reply-To: <49108988.40309@redhat.com> References: <490F701A.9080807@redhat.com> <491070A5.5040706@redhat.com> <49108988.40309@redhat.com> Message-ID: Kyle Powell a ?crit : > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Ugo Bellavance wrote: >> They're the right version on the filesystem, but when I access them via >> apache, the files are totally different. The change I did is that I >> mounted the 5.2 iso instead of the 5.0. I restarted httpd a few times >> since, but I'm confused about what could that happen. I've deleted my >> browser cache and I'm not going through a proxy. >> >> Any idea why that could happen? > > Nope. I have no idea why apache would send you a different file than what's on > the filesystem. I got it, it was an alias... Alias /pxe/rhel564/Server/repodata "/var/www/html/pxe/reporhel564/repodata" I wonder why it's there... I'm not the one who configured it... Ugo From robin-lists at robinbowes.com Wed Nov 5 00:45:50 2008 From: robin-lists at robinbowes.com (Robin Bowes) Date: Wed, 05 Nov 2008 00:45:50 +0000 Subject: set repo priority for initial install Message-ID: Hi, Is it possible to set yum repo priority during the initial (kickstart) install? I'm adding additional repos to my install (using cobbler, actually) and would like to prioritise them to ensure that the correct version is installed if a package exists in >1 repo R. From tony.chamberlain at lemko.com Wed Nov 5 02:37:06 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Wed, 05 Nov 2008 02:37:06 +0000 Subject: extra packages Message-ID: -----Original Message----- From: Jarle Bj?rgeengen [mailto:jarle at bjorgeengen.net] Sent: Tuesday, November 4, 2008 01:29 PM To: 'Discussion list about Kickstart' Subject: Re: extra packages On Nov 4, 2008, at 6:39 , Kyle Powell wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > tony.chamberlain at lemko.com wrote: >> You seem not to be able to do ALT-F2 during the %post part. And it >> went very quickly (I just did sleep 5000 which I hope is about 8 >> minutes). >> It went way quicker than that, but the "Installing bootloader" is >> taking >> forver. I wonder whether my sleep statement is taking place there >> instead? > > No, you certainly should be able to switch to VT 2 while the %post > script is > running. If anaconda seems to be finishing the post script quickly > and you have > a 'sleep 5000' in your post script, this would indicate the post > script is not > running completely, at least not far enough to reach the sleep > (sleep 5000 would > sleep for 5000 seconds, which is nearly an hour and a half). > >> %post --nochroot >> mkdir -p /mnt/cdrom >> mount /tmp/cdrom /mnt/cdrom >> mkdir /tmp/inst >> cp /mnt/cdrom/addon/*.tz /tmp/inst/ >> cd /tmp/inst >> tar -zxvf preimage.tz >> cd preimage >> ./install >> cd .. >> tar -zxvf nppt.tz >> cd nppt >> ./install >> >> 1. How do I access the addon directory on the DVD? I know it is >> there >> because on a Linux system I put the DVD in the reader and see it. > > When installing from CD or DVD, it will already be mounted at /mnt/ > source. I > would recommend you use two post scripts. Copy the files from the CD > in a > - --nochroot post script, then run the install in a regular > (chroot'ed) post script. > > %post --nochroot > mkdir -p /mnt/sysimage/tmp/inst > cp /mnt/source/addon/* /mnt/sysimage/tmp/inst > > %post > cd /tmp/inst > tar -zxvf preimage.tz > cd preimage > ./install > cd .. > tar -zxvf nppt.tz > cd nppt > ./install > Neat . Didn'r know you could have 2 %posts >> 2. How can I test thet %post without wasting a DVD each time? > > If you have a second machine you could pull the ks.cfg via http, or > if you have > a floppy, you could get the kickstart file from the floppy drive. > These would > allow you to change the kickstart file without burning a new DVD > each time. If > you absolutely want to pull the kickstart file from the iso each > time (instead > of just changing the ks= line to point to the cdrom once all your > testing is > complete) you could kickstart a virtual machine and present the iso > file to the > VM without actually burning a DVD. You can also use usb-stick linux ks=hd: Just edit isolinux config if you want this to be default. Once I pre-made a DVD image which wrote host-specific options to a temporary file in the %pre section, based on the output of dmidecode (machine serial number) , and later included the autogenerated file in the main ks.config. ========== I tried many values for "" but none seemed to work. From bjs at redhat.com Wed Nov 5 03:25:14 2008 From: bjs at redhat.com (Bryan J Smith) Date: Tue, 04 Nov 2008 22:25:14 -0500 Subject: Remedial Question: How do I import additional GPG keys (but not later in post)? Message-ID: <1225855514.4258.442.camel@localhost.localdomain> I have a seemingly remedial question ... How do I import additional GPG keys in a kickstart file so they are available during the initial package install? I want to do this largely in combination with the "repo" option (to tap other YUM repositories). Just curious and I know this is probably remedial. But I haven't found it in any Fedora docs (nor the RHEL 5 manuals for that matter, as it supports the "repo" option as well) or via basic searches. I guess I could have overlooked something. TIA! -- Bryan J Smith - Senior Consultant - Red Hat GPS SE US mailto:bjs at redhat.com +1 (407) 489-7013 (Mobile) mailto:b.j.smith at ieee.org (non-RH/ext to Blackberry) ----------------------------------------------------- For every dollar you spend on Red Hat solutions, you not only fund the leading community development re- source, but you receive the #1 IT industry leader in corporate value. http://www.redhat.com/promo/vendor/ From skvidal at fedoraproject.org Wed Nov 5 03:35:18 2008 From: skvidal at fedoraproject.org (Seth Vidal) Date: Tue, 4 Nov 2008 22:35:18 -0500 (EST) Subject: Remedial Question: How do I import additional GPG keys (but not later in post)? In-Reply-To: <1225855514.4258.442.camel@localhost.localdomain> References: <1225855514.4258.442.camel@localhost.localdomain> Message-ID: On Tue, 4 Nov 2008, Bryan J Smith wrote: > I have a seemingly remedial question ... > > How do I import additional GPG keys in a kickstart file so they are > available during the initial package install? I want to do this largely > in combination with the "repo" option (to tap other YUM repositories). > > Just curious and I know this is probably remedial. But I haven't found > it in any Fedora docs (nor the RHEL 5 manuals for that matter, as it > supports the "repo" option as well) or via basic searches. I guess I > could have overlooked something. > 1. anaconda doesn't check gpg keys during the kickstart install. 2. if you want gpg keys imported for repositories for use in in %post via yum just put gpgkey=pathtokey in the .repo section and if you pass a -y to yum it will auto-import the gpgkey. -sv From bjs at redhat.com Wed Nov 5 03:46:55 2008 From: bjs at redhat.com (Bryan J Smith) Date: Tue, 04 Nov 2008 22:46:55 -0500 Subject: Remedial Question: How do I import additional GPG keys (but not later in post)? In-Reply-To: References: <1225855514.4258.442.camel@localhost.localdomain> Message-ID: <1225856815.4258.464.camel@localhost.localdomain> On Tue, 2008-11-04 at 22:35 -0500, Seth Vidal wrote: > 1. anaconda doesn't check gpg keys during the kickstart install. Okay, that's what I wanted to verify, but didn't want to say. ;) > 2. if you want gpg keys imported for repositories for use in in %post via > yum just put gpgkey=pathtokey in the .repo section and if you pass a -y to > yum it will auto-import the gpgkey. I think I'd just rather to a "rpm --import http://" in %post instead. Any advantages to not doing such? -- Bryan J Smith - Senior Consultant - Red Hat GPS SE US mailto:bjs at redhat.com +1 (407) 489-7013 (Mobile) mailto:b.j.smith at ieee.org (non-RH/ext to Blackberry) ----------------------------------------------------- For every dollar you spend on Red Hat solutions, you not only fund the leading community development re- source, but you receive the #1 IT industry leader in corporate value. http://www.redhat.com/promo/vendor/ From skvidal at fedoraproject.org Wed Nov 5 04:03:22 2008 From: skvidal at fedoraproject.org (Seth Vidal) Date: Tue, 4 Nov 2008 23:03:22 -0500 (EST) Subject: Remedial Question: How do I import additional GPG keys (but not later in post)? In-Reply-To: <1225856815.4258.464.camel@localhost.localdomain> References: <1225855514.4258.442.camel@localhost.localdomain> <1225856815.4258.464.camel@localhost.localdomain> Message-ID: On Tue, 4 Nov 2008, Bryan J Smith wrote: > On Tue, 2008-11-04 at 22:35 -0500, Seth Vidal wrote: >> 1. anaconda doesn't check gpg keys during the kickstart install. > > Okay, that's what I wanted to verify, but didn't want to say. ;) https://bugzilla.redhat.com/show_bug.cgi?id=998 >> 2. if you want gpg keys imported for repositories for use in in %post via >> yum just put gpgkey=pathtokey in the .repo section and if you pass a -y to >> yum it will auto-import the gpgkey. > > I think I'd just rather to a "rpm --import http://" in %post instead. > Any advantages to not doing such? At some point soon (fedora11?, rhel6?) it won't do what you want anymore. rpm gpg keys won't be at the rpmdb. yay -sv From bjs at redhat.com Wed Nov 5 04:09:21 2008 From: bjs at redhat.com (Bryan J Smith) Date: Tue, 04 Nov 2008 23:09:21 -0500 Subject: Remedial Question: How do I import additional GPG keys (but not later in post)? In-Reply-To: References: <1225855514.4258.442.camel@localhost.localdomain> <1225856815.4258.464.camel@localhost.localdomain> Message-ID: <1225858161.4258.467.camel@localhost.localdomain> On Tue, 2008-11-04 at 23:03 -0500, Seth Vidal wrote: > At some point soon (fedora11?, rhel6?) it won't do what you want anymore. > rpm gpg keys won't be at the rpmdb. Good to know! -- Bryan J Smith - Senior Consultant - Red Hat GPS SE US mailto:bjs at redhat.com +1 (407) 489-7013 (Mobile) mailto:b.j.smith at ieee.org (non-RH/ext to Blackberry) ----------------------------------------------------- For every dollar you spend on Red Hat solutions, you not only fund the leading community development re- source, but you receive the #1 IT industry leader in corporate value. http://www.redhat.com/promo/vendor/ From mats.o.karlsson at gmail.com Wed Nov 5 06:09:05 2008 From: mats.o.karlsson at gmail.com (Mats Karlsson) Date: Wed, 5 Nov 2008 07:09:05 +0100 Subject: extra packages In-Reply-To: References: Message-ID: <32d5c55a0811042209w2db71b94r22ad259f6516be11@mail.gmail.com> Alt. 1. Invest in an rewritable DVD Alt. 2. Place the ks.cfg on a http/ftp server under the development phase On Wed, Nov 5, 2008 at 3:37 AM, wrote: > > -----Original Message----- > From: Jarle Bj?rgeengen [mailto:jarle at bjorgeengen.net] > Sent: Tuesday, November 4, 2008 01:29 PM > To: 'Discussion list about Kickstart' > Subject: Re: extra packages > > > On Nov 4, 2008, at 6:39 , Kyle Powell wrote: > > > -----BEGIN PGP SIGNED MESSAGE----- > > Hash: SHA1 > > > > tony.chamberlain at lemko.com wrote: > >> You seem not to be able to do ALT-F2 during the %post part. And it > >> went very quickly (I just did sleep 5000 which I hope is about 8 > >> minutes). > >> It went way quicker than that, but the "Installing bootloader" is > >> taking > >> forver. I wonder whether my sleep statement is taking place there > >> instead? > > > > No, you certainly should be able to switch to VT 2 while the %post > > script is > > running. If anaconda seems to be finishing the post script quickly > > and you have > > a 'sleep 5000' in your post script, this would indicate the post > > script is not > > running completely, at least not far enough to reach the sleep > > (sleep 5000 would > > sleep for 5000 seconds, which is nearly an hour and a half). > > > >> %post --nochroot > >> mkdir -p /mnt/cdrom > >> mount /tmp/cdrom /mnt/cdrom > >> mkdir /tmp/inst > >> cp /mnt/cdrom/addon/*.tz /tmp/inst/ > >> cd /tmp/inst > >> tar -zxvf preimage.tz > >> cd preimage > >> ./install > >> cd .. > >> tar -zxvf nppt.tz > >> cd nppt > >> ./install > >> > >> 1. How do I access the addon directory on the DVD? I know it is > >> there > >> because on a Linux system I put the DVD in the reader and see it. > > > > When installing from CD or DVD, it will already be mounted at /mnt/ > > source. I > > would recommend you use two post scripts. Copy the files from the CD > > in a > > - --nochroot post script, then run the install in a regular > > (chroot'ed) post script. > > > > %post --nochroot > > mkdir -p /mnt/sysimage/tmp/inst > > cp /mnt/source/addon/* /mnt/sysimage/tmp/inst > > > > %post > > cd /tmp/inst > > tar -zxvf preimage.tz > > cd preimage > > ./install > > cd .. > > tar -zxvf nppt.tz > > cd nppt > > ./install > > > > Neat . Didn'r know you could have 2 %posts > > >> 2. How can I test thet %post without wasting a DVD each time? > > > > If you have a second machine you could pull the ks.cfg via http, or > > if you have > > a floppy, you could get the kickstart file from the floppy drive. > > These would > > allow you to change the kickstart file without burning a new DVD > > each time. If > > you absolutely want to pull the kickstart file from the iso each > > time (instead > > of just changing the ks= line to point to the cdrom once all your > > testing is > > complete) you could kickstart a virtual machine and present the iso > > file to the > > VM without actually burning a DVD. > > You can also use usb-stick > > linux ks=hd: > > Just edit isolinux config if you want this to be default. > > Once I pre-made a DVD image which wrote host-specific options to a > temporary file in the %pre section, based on the output of dmidecode > (machine serial number) , and later included the autogenerated file in > the main ks.config. > > ========== > I tried many values for "" but none > seemed to work. > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > -------------- next part -------------- An HTML attachment was scrubbed... URL: From tony.chamberlain at lemko.com Wed Nov 5 16:53:06 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Wed, 05 Nov 2008 16:53:06 +0000 Subject: extra packages Message-ID: -----Original Message----- From: Kyle Powell [mailto:kpowell at redhat.com] Sent: Tuesday, November 4, 2008 11:39 AM To: 'Discussion list about Kickstart' Subject: Re: extra packages -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 tony.chamberlain at lemko.com wrote: > You seem not to be able to do ALT-F2 during the %post part. And it > went very quickly (I just did sleep 5000 which I hope is about 8 minutes). > It went way quicker than that, but the "Installing bootloader" is taking > forver. I wonder whether my sleep statement is taking place there > instead? No, you certainly should be able to switch to VT 2 while the %post script is running. If anaconda seems to be finishing the post script quickly and you have a 'sleep 5000' in your post script, this would indicate the post script is not running completely, at least not far enough to reach the sleep (sleep 5000 would sleep for 5000 seconds, which is nearly an hour and a half). > %post --nochroot > mkdir -p /mnt/cdrom > mount /tmp/cdrom /mnt/cdrom > mkdir /tmp/inst > cp /mnt/cdrom/addon/*.tz /tmp/inst/ > cd /tmp/inst > tar -zxvf preimage.tz > cd preimage > ./install > cd .. > tar -zxvf nppt.tz > cd nppt > ./install > > 1. How do I access the addon directory on the DVD? I know it is there > because on a Linux system I put the DVD in the reader and see it. When installing from CD or DVD, it will already be mounted at /mnt/source. I would recommend you use two post scripts. Copy the files from the CD in a - --nochroot post script, then run the install in a regular (chroot'ed) post script. %post --nochroot mkdir -p /mnt/sysimage/tmp/inst cp /mnt/source/addon/* /mnt/sysimage/tmp/inst %post cd /tmp/inst tar -zxvf preimage.tz cd preimage ./install cd .. tar -zxvf nppt.tz cd nppt ./install > 2. How can I test thet %post without wasting a DVD each time? If you have a second machine you could pull the ks.cfg via http, or if you have a floppy, you could get the kickstart file from the floppy drive. These would allow you to change the kickstart file without burning a new DVD each time. If you absolutely want to pull the kickstart file from the iso each time (instead of just changing the ks= line to point to the cdrom once all your testing is complete) you could kickstart a virtual machine and present the iso file to the VM without actually burning a DVD. - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJEIjq7pTtanQdBU4RArnMAJ9IkpJ7MTibLhB+QXe2XWWLNAEZ7ACgh2hs JZ1VNQZe9yH2J5dWs2qvWD4= =WtHD -----END PGP SIGNATURE----- _______________________________________________ Under both %post and %post --nochroot I did finds and could not find my stuff (preimage.tz, etc). Later in the %post I did a mkdir /tmp/disk mount /dev/hdc /tmp/disk mkdir /tmp/addon cp /tmp/disk/addon/* /tmp/addon and that worked. I could cd and untar and install. The thing is, if I make this disk how will I know the cdrom is always /dev/hdc? And why doesn't it show up in /mnt/source/ ? Also it looks like the %post is not being run in the postinstall but rather in the place where it says "installing bootloader" From tony.chamberlain at lemko.com Wed Nov 5 19:32:38 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Wed, 05 Nov 2008 19:32:38 +0000 Subject: extra packages Message-ID: -----Original Message----- From: Jarle Bj?rgeengen [mailto:jarle at bjorgeengen.net] Sent: Tuesday, November 4, 2008 01:29 PM To: 'Discussion list about Kickstart' Subject: Re: extra packages On Nov 4, 2008, at 6:39 , Kyle Powell wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > tony.chamberlain at lemko.com wrote: >> You seem not to be able to do ALT-F2 during the %post part. And it >> went very quickly (I just did sleep 5000 which I hope is about 8 >> minutes). >> It went way quicker than that, but the "Installing bootloader" is >> taking >> forver. I wonder whether my sleep statement is taking place there >> instead? > > No, you certainly should be able to switch to VT 2 while the %post > script is > running. If anaconda seems to be finishing the post script quickly > and you have > a 'sleep 5000' in your post script, this would indicate the post > script is not > running completely, at least not far enough to reach the sleep > (sleep 5000 would > sleep for 5000 seconds, which is nearly an hour and a half). > >> %post --nochroot >> mkdir -p /mnt/cdrom >> mount /tmp/cdrom /mnt/cdrom >> mkdir /tmp/inst >> cp /mnt/cdrom/addon/*.tz /tmp/inst/ >> cd /tmp/inst >> tar -zxvf preimage.tz >> cd preimage >> ./install >> cd .. >> tar -zxvf nppt.tz >> cd nppt >> ./install >> >> 1. How do I access the addon directory on the DVD? I know it is >> there >> because on a Linux system I put the DVD in the reader and see it. > > When installing from CD or DVD, it will already be mounted at /mnt/ > source. I > would recommend you use two post scripts. Copy the files from the CD > in a > - --nochroot post script, then run the install in a regular > (chroot'ed) post script. > > %post --nochroot > mkdir -p /mnt/sysimage/tmp/inst > cp /mnt/source/addon/* /mnt/sysimage/tmp/inst > > %post > cd /tmp/inst > tar -zxvf preimage.tz > cd preimage > ./install > cd .. > tar -zxvf nppt.tz > cd nppt > ./install > Neat . Didn'r know you could have 2 %posts >> 2. How can I test thet %post without wasting a DVD each time? > > If you have a second machine you could pull the ks.cfg via http, or > if you have > a floppy, you could get the kickstart file from the floppy drive. > These would > allow you to change the kickstart file without burning a new DVD > each time. If > you absolutely want to pull the kickstart file from the iso each > time (instead > of just changing the ks= line to point to the cdrom once all your > testing is > complete) you could kickstart a virtual machine and present the iso > file to the > VM without actually burning a DVD. You can also use usb-stick linux ks=hd: Just edit isolinux config if you want this to be default. Once I pre-made a DVD image which wrote host-specific options to a temporary file in the %pre section, based on the output of dmidecode (machine serial number) , and later included the autogenerated file in the main ks.config. ====================== OK I made something called lemkoks.cfg and put it in the top level directory of the dvd. If I do linux ks=cdrom:lemkoks.cfg everything works fine and it uses the lemkoks.cfg. I wanted to do like the above (without first the autogenerated stuff from %pre. Just isolinux stuff). So from the internet I found a web site and copied what they did into isolinux.cfg with my modifications like this: label prelem kernel vmlinuz append ks=cdrom:/lemkoks.cfg initrd=initrd.img ramdisk_size=8192 and then I redid mkisofs. When I boot and give the command linux prelem It loads everything, but it does not get the ks file. (I can always tell if at the beginning of installation I get the menu to test the disk or skip, that the ks file was not found). Did I do something wrong in the iso.cfg file? Or what am I missing? From tony.chamberlain at lemko.com Thu Nov 6 03:15:31 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Thu, 06 Nov 2008 03:15:31 +0000 Subject: extra packages -- additional parameters Message-ID: -----Original Message----- From: Mats Karlsson [mailto:mats.o.karlsson at gmail.com] Sent: Wednesday, November 5, 2008 12:09 AM To: 'Discussion list about Kickstart' Subject: Re: extra packages Alt. 1. Invest in an rewritable DVD Alt. 2. Place the ks.cfg on a http/ftp server under the development phase On Wed, Nov 5, 2008 at 3:37 AM, wrote: > > -----Original Message----- > From: Jarle Bj?rgeengen [mailto:jarle at bjorgeengen.net] > Sent: Tuesday, November 4, 2008 01:29 PM > To: 'Discussion list about Kickstart' > Subject: Re: extra packages > > > On Nov 4, 2008, at 6:39 , Kyle Powell wrote: > > > -----BEGIN PGP SIGNED MESSAGE----- > > Hash: SHA1 > > > > tony.chamberlain at lemko.com wrote: > >> You seem not to be able to do ALT-F2 during the %post part. And it > >> went very quickly (I just did sleep 5000 which I hope is about 8 > >> minutes). > >> It went way quicker than that, but the "Installing bootloader" is > >> taking > >> forver. I wonder whether my sleep statement is taking place there > >> instead? > > > > No, you certainly should be able to switch to VT 2 while the %post > > script is > > running. If anaconda seems to be finishing the post script quickly > > and you have > > a 'sleep 5000' in your post script, this would indicate the post > > script is not > > running completely, at least not far enough to reach the sleep > > (sleep 5000 would > > sleep for 5000 seconds, which is nearly an hour and a half). > > > >> %post --nochroot > >> mkdir -p /mnt/cdrom > >> mount /tmp/cdrom /mnt/cdrom > >> mkdir /tmp/inst > >> cp /mnt/cdrom/addon/*.tz /tmp/inst/ > >> cd /tmp/inst > >> tar -zxvf preimage.tz > >> cd preimage > >> ./install > >> cd .. > >> tar -zxvf nppt.tz > >> cd nppt > >> ./install > >> > >> 1. How do I access the addon directory on the DVD? I know it is > >> there > >> because on a Linux system I put the DVD in the reader and see it. > > > > When installing from CD or DVD, it will already be mounted at /mnt/ > > source. I > > would recommend you use two post scripts. Copy the files from the CD > > in a > > - --nochroot post script, then run the install in a regular > > (chroot'ed) post script. > > > > %post --nochroot > > mkdir -p /mnt/sysimage/tmp/inst > > cp /mnt/source/addon/* /mnt/sysimage/tmp/inst > > > > %post > > cd /tmp/inst > > tar -zxvf preimage.tz > > cd preimage > > ./install > > cd .. > > tar -zxvf nppt.tz > > cd nppt > > ./install > > > > Neat . Didn'r know you could have 2 %posts > > >> 2. How can I test thet %post without wasting a DVD each time? > > > > If you have a second machine you could pull the ks.cfg via http, or > > if you have > > a floppy, you could get the kickstart file from the floppy drive. > > These would > > allow you to change the kickstart file without burning a new DVD > > each time. If > > you absolutely want to pull the kickstart file from the iso each > > time (instead > > of just changing the ks= line to point to the cdrom once all your > > testing is > > complete) you could kickstart a virtual machine and present the iso > > file to the > > VM without actually burning a DVD. > > You can also use usb-stick > > linux ks=hd: > > Just edit isolinux config if you want this to be default. > > Once I pre-made a DVD image which wrote host-specific options to a > temporary file in the %pre section, based on the output of dmidecode > (machine serial number) , and later included the autogenerated file in > the main ks.config. > > ========== > I tried many values for "" but none > seemed to work. > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > ------------------------------ I found out the problem with the isolinux. Instead of typing "linux prelem" I just type "prelem" (the new label I added). Another couple questions. I saw someone do linux text nodmraid selinux=2 Is there a way for the ks.cfg to get these parameters? I mean if I want to add some of my own, like say prelem ip=10.0.0.66 mask=255.0.0.0 onboot=yes then my kickstart file could somehow reference these and set them, activate eth0 automatically. And then I could also have some of my own like installcdma=yes (or just installcdma). From barry at orodor.org.uk Fri Nov 7 10:39:25 2008 From: barry at orodor.org.uk (Barry O'Rourke) Date: Fri, 07 Nov 2008 10:39:25 +0000 Subject: Using kickstart to resize existing LVM partitions during a rebuild Message-ID: <49141ADD.7080709@orodor.org.uk> Hi, We have around 300 RHEL4 boxes that we are looking to upgrade to Scientific Linux 5, as part of the rebuild we would like to double the size of our current root partition to allow for future expansion and because they are getting a bit short on free space. Our current LVM set up looks like; /dev/mapper/ifp-slash 15Gb /dev/mapper/ifp-var 2Gb /dev/mapper/ifp-log 2Gb /dev/mapper/ifp-scratch fills the rest of the disk. We intend to script the shrinking of the ifp-scratch LV before starting the rebuild to free up 15Gb free space within the volume group which we would then like to add to the ifp-scratch LV during the rebuild. During the rebuild the ifp-scratch LV must stay intact as our users dump a lot of there data there. The problem that I've been having is that I can't figure out the syntax required for the logvol line in the kickstart file that will resize the ifp-slash to 30Gb. Every attempt that I've made so far seems to resize it to 15Gb as it was before. If you need any more information please let me know, and thanks in advance for any help that is offered. Barry From ugob at lubik.ca Fri Nov 7 15:01:30 2008 From: ugob at lubik.ca (Ugo Bellavance) Date: Fri, 07 Nov 2008 10:01:30 -0500 Subject: yum in kickstart %post and HP Insight Mgr In-Reply-To: <1117568770.3743.32.camel@localhost.localdomain> References: <1117568770.3743.32.camel@localhost.localdomain> Message-ID: Brian Long a ?crit : > I'm trying to use yum in the %post section of kickstart. With well- > written RPMs, this works great. However, I've come across the lovely HP > Insight Manager RPMS (hpasm, hprsm, etc) which seem to run "service > hpasm start" during the RPM postinstall (yuck). This wouldn't be too > bad if the daemons actually disassociated from the running terminal. > > I notice yum install the group "yum groupinstall Insight-Manager" and > yum exits, but the rest of the kickstart %post doesn't run. If I look > at the hpasmd processes, they're tied to tty1 and owned by init :( If > I hit Alt-F2 in kickstart and chroot /mnt/sysimage, service hpasm stop, > the rest of the kickstart %post finishes. I also strace'd the /tmp/ks- > script process and it's stuck at waitpid(-1, ) > > I cannot wait on a vendor escalation at this time, so I see one method > of a workaround. Move the yum installs to a firstboot scenario instead > of %post. I'd really like to find another way around this. > > Any ideas would be appreciated. Thanks. Hi, more than 3 years later, is there a solution to this? We have a bunch of HP servers and I'm working to make the installs as automated as possible and I'd like to end up with a system with all the HP packages installed after anaconda finishes... How are you guys now proceeding? Regards, Ugo From skvidal at fedoraproject.org Fri Nov 7 15:08:32 2008 From: skvidal at fedoraproject.org (Seth Vidal) Date: Fri, 7 Nov 2008 10:08:32 -0500 (EST) Subject: yum in kickstart %post and HP Insight Mgr In-Reply-To: References: <1117568770.3743.32.camel@localhost.localdomain> Message-ID: On Fri, 7 Nov 2008, Ugo Bellavance wrote: > Brian Long a ?crit : >> I'm trying to use yum in the %post section of kickstart. With well- >> written RPMs, this works great. However, I've come across the lovely HP >> Insight Manager RPMS (hpasm, hprsm, etc) which seem to run "service >> hpasm start" during the RPM postinstall (yuck). This wouldn't be too >> bad if the daemons actually disassociated from the running terminal. >> I notice yum install the group "yum groupinstall Insight-Manager" and >> yum exits, but the rest of the kickstart %post doesn't run. If I look >> at the hpasmd processes, they're tied to tty1 and owned by init :( If >> I hit Alt-F2 in kickstart and chroot /mnt/sysimage, service hpasm stop, >> the rest of the kickstart %post finishes. I also strace'd the /tmp/ks- >> script process and it's stuck at waitpid(-1, ) >> >> I cannot wait on a vendor escalation at this time, so I see one method >> of a workaround. Move the yum installs to a firstboot scenario instead >> of %post. I'd really like to find another way around this. >> >> Any ideas would be appreciated. Thanks. > > Hi, > > more than 3 years later, is there a solution to this? We have a bunch of HP > servers and I'm working to make the installs as automated as possible and I'd > like to end up with a system with all the HP packages installed after > anaconda finishes... How are you guys now proceeding? > 1. did anyone tell hp about this and get them to fix their packages? 2. if you know which packages it is you're worried about and you know there's nothing else running in their %scriptlets then you could install those pkgs w/yum using a yum.conf with tsflags=noscripts under [main]. -sv From Chip.Shabazian at bankofamerica.com Fri Nov 7 15:28:16 2008 From: Chip.Shabazian at bankofamerica.com (Shabazian, Chip) Date: Fri, 07 Nov 2008 07:28:16 -0800 Subject: Using kickstart to resize existing LVM partitions during a rebuild In-Reply-To: <49141ADD.7080709@orodor.org.uk> References: <49141ADD.7080709@orodor.org.uk> Message-ID: Why not resize your LVM before kickstarting? Seems to me it would be a lot easier to run a resize script on all the machines while they are up and running, THEN Kickstart them with the properly sized LVM's. -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Barry O'Rourke Sent: Friday, November 07, 2008 2:39 AM To: kickstart-list at redhat.com Subject: Using kickstart to resize existing LVM partitions during a rebuild Hi, We have around 300 RHEL4 boxes that we are looking to upgrade to Scientific Linux 5, as part of the rebuild we would like to double the size of our current root partition to allow for future expansion and because they are getting a bit short on free space. Our current LVM set up looks like; /dev/mapper/ifp-slash 15Gb /dev/mapper/ifp-var 2Gb /dev/mapper/ifp-log 2Gb /dev/mapper/ifp-scratch fills the rest of the disk. We intend to script the shrinking of the ifp-scratch LV before starting the rebuild to free up 15Gb free space within the volume group which we would then like to add to the ifp-scratch LV during the rebuild. During the rebuild the ifp-scratch LV must stay intact as our users dump a lot of there data there. The problem that I've been having is that I can't figure out the syntax required for the logvol line in the kickstart file that will resize the ifp-slash to 30Gb. Every attempt that I've made so far seems to resize it to 15Gb as it was before. If you need any more information please let me know, and thanks in advance for any help that is offered. Barry _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From barry at orodor.org.uk Fri Nov 7 16:25:43 2008 From: barry at orodor.org.uk (Barry O'Rourke) Date: Fri, 07 Nov 2008 16:25:43 +0000 Subject: Using kickstart to resize existing LVM partitions during a rebuild In-Reply-To: References: <49141ADD.7080709@orodor.org.uk> Message-ID: <49146C07.7030006@orodor.org.uk> Hi, My initial plan was to do it the way you suggest but it seems that doing an online resize of the root partition doesn't agree with fsck when it reboots, the majority of test hosts forced us in to single user mode to run fsck by hand. In theory that would work out perfectly if we were building the machines ourselves as we could run the script and the rebuild without rebooting into the OS, but we are planning to allow our users to rebuild at a time that suits them using a grub rebuild option. Thats how we came up with the idea of resizing during kickstart. Barry Shabazian, Chip wrote: > Why not resize your LVM before kickstarting? Seems to me it would be a > lot easier to run a resize script on all the machines while they are up > and running, THEN Kickstart them with the properly sized LVM's. > > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Barry O'Rourke > Sent: Friday, November 07, 2008 2:39 AM > To: kickstart-list at redhat.com > Subject: Using kickstart to resize existing LVM partitions during a > rebuild > > Hi, > > We have around 300 RHEL4 boxes that we are looking to upgrade to > Scientific Linux 5, as part of the rebuild we would like to double the > size of our current root partition to allow for future expansion and > because they are getting a bit short on free space. > > Our current LVM set up looks like; > > /dev/mapper/ifp-slash 15Gb > /dev/mapper/ifp-var 2Gb > /dev/mapper/ifp-log 2Gb > /dev/mapper/ifp-scratch fills the rest of the disk. > > We intend to script the shrinking of the ifp-scratch LV before starting > the rebuild to free up 15Gb free space within the volume group which we > would then like to add to the ifp-scratch LV during the rebuild. > > During the rebuild the ifp-scratch LV must stay intact as our users dump > > a lot of there data there. > > The problem that I've been having is that I can't figure out the syntax > required for the logvol line in the kickstart file that will resize the > ifp-slash to 30Gb. Every attempt that I've made so far seems to resize > it to 15Gb as it was before. > > If you need any more information please let me know, and thanks in > advance for any help that is offered. > > Barry > > _______________________________________________ > 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 jason at rampaginggeek.com Fri Nov 7 20:24:43 2008 From: jason at rampaginggeek.com (Jason Edgecombe) Date: Fri, 07 Nov 2008 15:24:43 -0500 Subject: kickstart using ksdevice=link doesn't find the right NIC Message-ID: <4914A40B.50402@rampaginggeek.com> Hi Everyone, I'm trying to kickstart a Dell PowerEdge R200 with 4 NIC's (2 onboard & 2 cards). The NIC's are the following: 2 x Broadcom Corporation NetXtreme BCM5722 Gigabit Ethernet PCI Express 2 x Broadcom Corporation NetXtreme BCM5721 Gigabit Ethernet PCI Express the onboard NIC labeled "1" (eth2) is the only one plugged in. I can kickstart and get an IP if I use "ksdevice=eth2 or ksdevice=MAC_ADDRESS, but ksdevice=link just attempts to use eth0. Has anyone has any issues with ksdevice=link failing like this? The OS is RHEL5.2 x86_64. I boot from a CD or grub on the hard drive., The RPMs are on a NFS share and the kickstart file comes from a web server. Any ideas? Thanks, Jason From tru at pasteur.fr Fri Nov 7 20:49:37 2008 From: tru at pasteur.fr (Tru Huynh) Date: Fri, 7 Nov 2008 21:49:37 +0100 Subject: Using kickstart to resize existing LVM partitions during a rebuild In-Reply-To: <49146C07.7030006@orodor.org.uk> References: <49141ADD.7080709@orodor.org.uk> <49146C07.7030006@orodor.org.uk> Message-ID: <20081107204937.GF8665@sillage.bis.pasteur.fr> On Fri, Nov 07, 2008 at 04:25:43PM +0000, Barry O'Rourke wrote: > Hi, > > My initial plan was to do it the way you suggest but it seems that doing > an online resize of the root partition doesn't agree with fsck when it > reboots, the majority of test hosts forced us in to single user mode to > run fsck by hand. why don't you use %pre to do you resizing and %include the partitions information? Tru From tony.chamberlain at lemko.com Sat Nov 8 01:01:14 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Sat, 08 Nov 2008 01:01:14 +0000 Subject: miscellaneous Message-ID: Hey I found the answers to some questions through trial/error and internet. Thought I would share them. 1. If you add a label (say mylabel) to isolinux.cfg then to access that label during install, you don't say "linux mylabel" but simply "mylabel". For instance here if you add label mylabel kernel vmlinuz append initrd=initrd.img ramdisk_size=8192 ks=cdrom:myks.cfg 2. For %post (and %pre ?) there is a way to pass parameters. Say you have the above (from #1). If you say "mylabel install=cdma parm" from your %post (and maybe %pre though I did not try it) there will be a file available called /proc/cmdline. The contents will be something like initrd=initrd.img ramdisk_size=8192 ks=cdrom:myks.cfg install=cdma parm which you will have access to. Again sorry if you knew this already but no one seemed to when I asked. From barry at orodor.org.uk Sat Nov 8 09:48:25 2008 From: barry at orodor.org.uk (Barry O'Rourke) Date: Sat, 08 Nov 2008 09:48:25 +0000 Subject: Using kickstart to resize existing LVM partitions during a rebuild In-Reply-To: <20081107204937.GF8665@sillage.bis.pasteur.fr> References: <49141ADD.7080709@orodor.org.uk> <49146C07.7030006@orodor.org.uk> <20081107204937.GF8665@sillage.bis.pasteur.fr> Message-ID: <49156069.9020503@orodor.org.uk> Hi, > why don't you use %pre to do you resizing and %include the partitions > information? That's what I've been trying to do but I can't figure out how to. Thanks, Barry From mdehaan at redhat.com Mon Nov 10 18:19:43 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Nov 2008 13:19:43 -0500 Subject: miscellaneous In-Reply-To: References: Message-ID: <49187B3F.9040306@redhat.com> tony.chamberlain at lemko.com wrote: > > 2. For %post (and %pre ?) there is a way to pass parameters. Say you have > the above (from #1). If you say "mylabel install=cdma parm" > from your %post (and maybe %pre though I did not try it) there will be > a file available called /proc/cmdline. The contents will be something like > initrd=initrd.img ramdisk_size=8192 ks=cdrom:myks.cfg install=cdma parm > which you will have access to. > > Again sorry if you knew this already but no one seemed to when I asked. > Rather than passing parameters through the kernel (which can be messy in case the kernel cares about those values or you run out of room on teh command line), you can also use a kickstart templating system to define parameters. This is much nicer IMHO. Cobbler (http://fedorahosted.org/cobbler) contains a system to do this. Install it and read the manpage on --ks-meta and the Wiki pages on KickstartTemplating for more details. > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From sukhuran at cisco.com Tue Nov 11 05:02:15 2008 From: sukhuran at cisco.com (Sudhir Khurana (sukhuran)) Date: Tue, 11 Nov 2008 10:32:15 +0530 Subject: Adding Custom packages to RHEL Kickstart process In-Reply-To: References: Message-ID: Hi , I am looking for some help in adding custom packages to kickstart process for RHEL 5.1. Can some point me to relevant information / or has done similar task before who can be of help in solving the following issues 1. Adding custom RPM's to %package, I am trying to add custom rpm packages in the Kickstart process and has specified the same in ks.cfg file, but still it doesn't get installed. Also I am unable to locate comps.xml file to add the package information for Red Hat Server Installation process to register the rpm. Not aware of the process, can some one please help in this. 2. Where can I see logs / output of %pre and %post scripts, we are using python as interpreter. Thanks in advance, Sudhir From barry at orodor.org.uk Tue Nov 11 11:09:18 2008 From: barry at orodor.org.uk (Barry O'Rourke) Date: Tue, 11 Nov 2008 11:09:18 +0000 Subject: Using kickstart to resize existing LVM partitions during a rebuild In-Reply-To: <20081107204937.GF8665@sillage.bis.pasteur.fr> References: <49141ADD.7080709@orodor.org.uk> <49146C07.7030006@orodor.org.uk> <20081107204937.GF8665@sillage.bis.pasteur.fr> Message-ID: <491967DE.60507@orodor.org.uk> Hi, > why don't you use %pre to do you resizing and %include the partitions > information? I'm back at work today so have had more of a chance to play around with this. I have added the following to the beginning of the %pre section; /usr/sbin/lvm lvmresize -L30G /dev/mapper/ifp-slash /usr/sbin/resize2fs -f /dev/ifp/slash 30G and I'm including the following partition information; echo "# Create Logical Volumes with a growing /scratch partition generated in %pre " > ${LOGVOL} echo "logvol / --fstype ext3 --name=slash --vgname=ifp --size=${SLASH} --useexisting" >> ${LOGVOL} echo "logvol /var --fstype ext3 --name=var --vgname=ifp --size=${VAR} --useexisting" >> ${LOGVOL} echo "logvol /var/log --fstype ext3 --name=log --vgname=ifp --size=${LOG} --useexisting" >> ${LOGVOL} echo "logvol swap --fstype swap --name=swap --vgname=ifp --recommended --useexisting" >> ${LOGVOL} echo "logvol /scratch --fstype ext3 --name=scratch --vgname=ifp --size=10 --grow --noformat" >> ${LOGVOL} where the disk sizes are; SLASH="30720" VAR="2048" LOG="${VAR}" SCRATCH="1024" Thanks, Barry From tom at ng23.net Tue Nov 11 15:13:04 2008 From: tom at ng23.net (Tom Brown) Date: Tue, 11 Nov 2008 15:13:04 +0000 Subject: yum in kickstart %post and HP Insight Mgr In-Reply-To: References: <1117568770.3743.32.camel@localhost.localdomain> Message-ID: <4919A100.7050204@ng23.net> > > > 1. did anyone tell hp about this and get them to fix their packages? > > 2. if you know which packages it is you're worried about and you know > there's nothing else running in their %scriptlets then you could > install those pkgs w/yum using a yum.conf with tsflags=noscripts under > [main]. i have not come across this issue before and apart from my case being rubbish this is how i automate hardware monitoring rpm installation at kickstart time and it works fine for me, although yes, not using yum.... # Install the distro specific monitoring VENDOR=`dmidecode | grep Vendor | awk '{print$2}'` DISTRO=`cat /etc/redhat-release | awk {'print$7'}` ARCH=`uname -i` OSTYPE=`echo $VENDOR$DISTRO$ARCH` case "$OSTYPE" in Dell4i386) wget http://@@server@@/cobbler/ks_mirror/posts/openmanage-5.2.tar.gz -O /tmp/openmanage.tar.gz && tar zxfv /t mp/openmanage.tar.gz -C /tmp && /tmp/5.2/linux/supportscripts/srvadmin-install.sh -x ; rm -rf /tmp/5.2 /tmp/openmanage.tar.gz ; ;; Dell5.2i386) wget http://@@server@@/cobbler/ks_mirror/posts/openmanage-5.2.tar.gz -O /tmp/openmanage.tar.gz && tar zxfv /t mp/openmanage.tar.gz -C /tmp && /tmp/5.2/linux/supportscripts/srvadmin-install.sh -x ; rm -rf /tmp/5.2 /tmp/openmanage.tar.gz ; ;; Dell4x86_64) wget http://@@server@@/cobbler/ks_mirror/posts/openmanage-5.2.tar.gz -O /tmp/openmanage.tar.gz && tar zxfv /t mp/openmanage.tar.gz -C /tmp && /tmp/5.2/linux/supportscripts/srvadmin-install.sh -x ; rm -rf /tmp/5.2 /tmp/openmanage.tar.gz ; ;; Dell5.2x86_64) wget http://@@server@@/cobbler/ks_mirror/posts/openmanage-5.2.tar.gz -O /tmp/openmanage.tar.gz && tar zxfv /t mp/openmanage.tar.gz -C /tmp && /tmp/5.2/linux/supportscripts/srvadmin-install.sh -x ; rm -rf /tmp/5.2 /tmp/openmanage.tar.gz ; ;; HP4i386) rpm -i http://@@server@@/cobbler/ks_mirror/posts/hpacucli-8.10-2.noarch.rpm http://@@server@@/cobbler/ks_mirr or/posts/hpasm-8.0.0-173.rhel4.i386.rpm && chkconfig hpasm on; ;; HP5.2i386) rpm -i http://@@server@@/cobbler/ks_mirror/posts/hpacucli-8.10-2.noarch.rpm http://@@server@@/cobbler/ks_mirr or/posts/hpasm-8.0.0-173.rhel5.i386.rpm && chkconfig hpasm on; ;; HP4x86_64) rpm -i http://@@server@@/cobbler/ks_mirror/posts/hpacucli-8.10-2.noarch.rpm http://@@server@@/cobbler/ks_mirr or/posts/hpasm-8.0.0-173.rhel4.x86_64.rpm && chkconfig hpasm on; ;; HP5x86_64) rpm -i http://@@server@@/cobbler/ks_mirror/posts/hpacucli-8.10-2.noarch.rpm http://@@server@@/cobbler/ks_mirr or/posts/hpasm-8.0.0-173.rhel5.x86_64.rpm && chkconfig hpasm on; ;; esac From william_dennison at sra.com Wed Nov 12 21:11:51 2008 From: william_dennison at sra.com (William J. Dennison) Date: Wed, 12 Nov 2008 16:11:51 -0500 Subject: Issue With Packages Section Message-ID: <200811121611.51229.william_dennison@sra.com> I am using Cobbler with Kickstart. Here is my problem. I am attempting to install the following packages (see below) in my kickstart file (using Cobbler as operator on RHEL 5.2). When Kickstart installs the jdk, tomcat, and VMWare packaged they are pulled to the server and then installed in the right areas, however the /etc/init.d parameters are not installed correctly (as if kickstart did not utilize the configuration parameters in the rpms). I manually ran the yum install command after the server was booted for these rpms, and they indeed installed correctly, therefore I surmise it is something with kickstart. I looked at the Anaconda log file and it seems that the rpms are not installed in the correct order, could this be the problem? If yes, how can I fix it. If now, any other ideas would help. -------------------------------------------------------- ### Package Installation Specifications %packages # Install the Base System and vim-enhanced by default. @ Base vim-enhanced # The following packages are required for a Multi-Node # VMWare server build. gcc gcc-c++ kernel-headers openssl-devel xinetd libXt libXtst libXrender mysql-server jdk tomcat VMware-server -- William J. Dennison Principal Consultant SRA International Inc. - RABA Center 8830 Stanford Blvd, Suite 205 Columbia, Maryland 21045 Phone: (410) 715-9399 x2351 Mobile: (410) 218-5866 Email: william_dennison at sra.com Email2: wjdennison at gmail.com Email3: wjdennison at cablespeed.com Jabber: wjdennison at jabber.org AIM: dennisonwjonly GTalk: wjdennison at gmail.com From terry.mcintyre at gmail.com Wed Nov 12 21:48:42 2008 From: terry.mcintyre at gmail.com (Terry McIntyre) Date: Wed, 12 Nov 2008 13:48:42 -0800 Subject: RHEL 3 kickstart network retrieval issue In-Reply-To: References: <48199B69.1050903@ng23.net> Message-ID: <67f096ea0811121348g28f0a414u1cf96542ebbc4476@mail.gmail.com> I had a similar issue; the solution which worked for me was to specify the ip address, gateway, and netmask: append initrd=/images/centos-5.2-i386/initrd.img ksdevice=eth0 lang=en_US netmask=255.255.254.0 ip=192.168.62.55 text kssendmac gateway=192.168.62.40 ks=http:// The key bits are "ksdevice=eth0 netmask=255.255.254.0 ip=192.168.62.55 gateway=192.168.62.40 " This will probably be needed if your environment uses something other than 255.255.255.0 for a netmask. -- Terry McIntyre UNIX for hire Software Development, Systems Administration, Security terry.mcintyre at gmail.com From kpowell at redhat.com Wed Nov 12 22:18:31 2008 From: kpowell at redhat.com (Kyle Powell) Date: Wed, 12 Nov 2008 17:18:31 -0500 Subject: Issue With Packages Section In-Reply-To: <200811121611.51229.william_dennison@sra.com> References: <200811121611.51229.william_dennison@sra.com> Message-ID: <491B5637.5060209@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 William J. Dennison wrote: > When Kickstart installs the jdk, tomcat, and VMWare packaged they are pulled > to the server and then installed in the right areas, however the /etc/init.d > parameters are not installed correctly (as if kickstart did not utilize the > configuration parameters in the rpms). What exactly do you mean by this? Are the init scripts not being placed in /etc/rc.d/init.d or are the symlinks not being created in the appropriate /etc/rc?.d directory? "Parameters" are not "installed", especially in the /etc/init.d directory. > I looked at the Anaconda log file and it seems that the rpms are not installed > in the correct order, could this be the problem? There's no way that I'm aware of to control the order in which Anaconda installs packages. They're all installed with --nodeps (and --replacefiles, and a few others, or maybe even just --force) to eliminate any failures due to packages being installed out of order. If you have packages that *must* be installed in a certain order, then your best bet is to install them via yum in a %post script. > jdk > tomcat > VMware-server I can't find these packages for RHEL 5.2. I see "jdkgcj" and "tomcat5" available, but nothing for VMware-server. Looks like these aren't actual RHEL 5 packages. Depending on your answer to my question above, we may need to look at the postinstall scripts for them to determine why things are working correctly during installation. - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJG1Y37pTtanQdBU4RArA4AJ4h/TEyVIVb5Z+u0Osfgwby8V6fRACcCGEc iDzc9ljuFJP3fkRWSB95WL4= =xZ2y -----END PGP SIGNATURE----- From william_dennison at sra.com Wed Nov 12 23:35:55 2008 From: william_dennison at sra.com (William J. Dennison) Date: Wed, 12 Nov 2008 18:35:55 -0500 Subject: Issue With Packages Section In-Reply-To: <491B5637.5060209@redhat.com> References: <200811121611.51229.william_dennison@sra.com> <491B5637.5060209@redhat.com> Message-ID: <200811121835.56120.william_dennison@sra.com> Let me be more specific The 3 packages (rpms) exist in an "extra" "local" repository that Cobbler offers to Kickstart. If the command "yum install tomcat" for example is given, the rpm for tomcat is accessed and installed correctly. If tomcat is placed in the Kickstart file, Tomcat is installed, but the rc.d directory is not populated with a chkconfig'ed link and the start file inside /etc/init.d is created but empty. Again if I run the yum command after the machine is completed it's load, all works correctly. Kickstart is somehow not running the installation script correctly or maybe there is a permission problem, I am at a loss. On Wednesday 12 November 2008 05:18:31 pm Kyle Powell wrote: > William J. Dennison wrote: > > When Kickstart installs the jdk, tomcat, and VMWare packaged they are > > pulled to the server and then installed in the right areas, however the > > /etc/init.d parameters are not installed correctly (as if kickstart did > > not utilize the configuration parameters in the rpms). > > What exactly do you mean by this? Are the init scripts not being placed in > /etc/rc.d/init.d or are the symlinks not being created in the appropriate > /etc/rc?.d directory? "Parameters" are not "installed", especially in the > /etc/init.d directory. > > > I looked at the Anaconda log file and it seems that the rpms are not > > installed in the correct order, could this be the problem? > > There's no way that I'm aware of to control the order in which Anaconda > installs packages. They're all installed with --nodeps (and --replacefiles, > and a few others, or maybe even just --force) to eliminate any failures due > to packages being installed out of order. If you have packages that *must* > be installed in a certain order, then your best bet is to install them via > yum in a %post script. > > > jdk > > tomcat > > VMware-server > > I can't find these packages for RHEL 5.2. I see "jdkgcj" and "tomcat5" > available, but nothing for VMware-server. Looks like these aren't actual > RHEL 5 packages. Depending on your answer to my question above, we may need > to look at the postinstall scripts for them to determine why things are > working correctly during installation. -- William J. Dennison Principal Consultant SRA International Inc. - RABA Center 8830 Stanford Blvd, Suite 205 Columbia, Maryland 21045 Phone: (410) 715-9399 x2351 Mobile: (410) 218-5866 Email: william_dennison at sra.com Email2: wjdennison at gmail.com Email3: wjdennison at cablespeed.com Jabber: wjdennison at jabber.org AIM: dennisonwjonly GTalk: wjdennison at gmail.com From stan.hearn at nscorp.com Thu Nov 13 13:30:32 2008 From: stan.hearn at nscorp.com (Hearn, Stan J.) Date: Thu, 13 Nov 2008 08:30:32 -0500 Subject: How to specify archetecture for package in ks.cfg? In-Reply-To: <200811121611.51229.william_dennison@sra.com> Message-ID: In my package list for and X86_64 server I want to specify several packages with i386 architectures. But using the version number doesn't work. How do I specify the architecture? Or do I just need to do that in the %post section using rpm command? %packages compat-libstdc++-33-3.2.3-61.i386 libaio-0.3.105-2.i386 expat-1.95.7-4.i386 Thanks, Stan From kpowell at redhat.com Thu Nov 13 16:15:19 2008 From: kpowell at redhat.com (Kyle Powell) Date: Thu, 13 Nov 2008 11:15:19 -0500 Subject: How to specify archetecture for package in ks.cfg? In-Reply-To: References: Message-ID: <491C5297.9020506@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hearn, Stan J. wrote: > How do I specify the architecture? Or do I just need to do that in the > %post section using rpm command? > > %packages > compat-libstdc++-33-3.2.3-61.i386 > libaio-0.3.105-2.i386 > expat-1.95.7-4.i386 Take the version info out. Just use package-name.arch: %packages compat-libstdc++-33.i386 libaio.i386 expat.i386 - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJHFKX7pTtanQdBU4RAkTiAJ9kzSAZRWmi8dn1tqB5I3O60yZgLQCgjYSV ALQABPffoAOgLhU1uHQSq/0= =ESlu -----END PGP SIGNATURE----- From kpowell at redhat.com Thu Nov 13 16:22:38 2008 From: kpowell at redhat.com (Kyle Powell) Date: Thu, 13 Nov 2008 11:22:38 -0500 Subject: Issue With Packages Section In-Reply-To: <200811121835.56120.william_dennison@sra.com> References: <200811121611.51229.william_dennison@sra.com> <491B5637.5060209@redhat.com> <200811121835.56120.william_dennison@sra.com> Message-ID: <491C544E.6050807@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 William J. Dennison wrote: > Let me be more specific > > The 3 packages (rpms) exist in an "extra" "local" repository that Cobbler > offers to Kickstart. If the command "yum install tomcat" for example is > given, the rpm for tomcat is accessed and installed correctly. If tomcat is > placed in the Kickstart file, Tomcat is installed, but the rc.d directory is > not populated with a chkconfig'ed link and the start file inside /etc/init.d > is created but empty. > > Again if I run the yum command after the machine is completed it's load, all > works correctly. Kickstart is somehow not running the installation script > correctly or maybe there is a permission problem, I am at a loss. OK, thanks. That clears up what the problem is. Now to determine the cause. I suspect a problem with the %postinstall scripts of the packages since other package installations are correctly placing files in /etc/rc.d/init.d and setting up the proper symlinks in the /etc/rc?.d dirs. So lets start by looking at the scripts in those packages. Can you show us the output of: rpm -q --scripts jdk rpm -q --scripts tomcat rpm -q --scripts VMware-server - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJHFRO7pTtanQdBU4RAne9AJ9+NOEqqy2hfJCyrKJ6ejSw4FBs1QCdFCYA iQh+7HOVMii4RvqgA2JolRs= =y3Gu -----END PGP SIGNATURE----- From Moray.Henderson at ict.om.org Fri Nov 14 10:00:50 2008 From: Moray.Henderson at ict.om.org (Moray Henderson (ICT)) Date: Fri, 14 Nov 2008 10:00:50 -0000 Subject: Issue With Packages Section In-Reply-To: <200811121835.56120.william_dennison@sra.com> References: <200811121611.51229.william_dennison@sra.com><491B5637.5060209@redhat.com> <200811121835.56120.william_dennison@sra.com> Message-ID: >From: William J. Dennison [mailto:william_dennison at sra.com] >The 3 packages (rpms) exist in an "extra" "local" repository that Cobbler >offers to Kickstart. If the command "yum install tomcat" for example is >given, the rpm for tomcat is accessed and installed correctly. If tomcat >is placed in the Kickstart file, Tomcat is installed, but the rc.d >directory is not populated with a chkconfig'ed link and the start file >inside /etc/init.d is created but empty. It feels as it Tomcat is being installed before some of the packages needed by its postinstall script, such as chkconfig (you can check the anaconda log to make sure). If you can get the source rpms, and modify them so that the .spec file lists those packages as PreReq (not Requires), you may get it to work. Otherwise, install them from %post. Moray. "To err is human.? To purr, feline" From Moray.Henderson at ict.om.org Fri Nov 14 10:22:14 2008 From: Moray.Henderson at ict.om.org (Moray Henderson (ICT)) Date: Fri, 14 Nov 2008 10:22:14 -0000 Subject: Adding Custom packages to RHEL Kickstart process In-Reply-To: References: Message-ID: >From: Sudhir Khurana (sukhuran) [mailto:sukhuran at cisco.com] >Hi , > > I am looking for some help in adding custom packages to kickstart >process for RHEL 5.1. > > Can some point me to relevant information / or has done similar task >before who can be of help in solving the following issues > >1. Adding custom RPM's to %package, I am trying to add custom rpm >packages in the Kickstart process and has specified the same in ks.cfg >file, but still it doesn't get installed. Also I am unable to locate >comps.xml file to add the package information for Red Hat Server >Installation process to register the rpm. Not aware of the process, can >some one please help in this. > >2. Where can I see logs / output of %pre and %post scripts, we are using >python as interpreter. > >Thanks in advance, > >Sudhir Search the list archives - they already contain the relevant information. In particular, there are several links to http://www.shabazian.com/lw2007.pdf, which should help you. Moray. "To err is human.? To purr, feline" From gbailey at terremark.com Thu Nov 20 01:12:12 2008 From: gbailey at terremark.com (Glenn Bailey) Date: Wed, 19 Nov 2008 19:12:12 -0600 Subject: simple pre script for user input Message-ID: <005AB96F733E674B89550DDAE730CB4A4BF895E3E5@daa20725exc001.drtn.corp> Playing around with a %pre script for user input, and while this works fine in a normal shell, after I enter an IP it just hangs there and I have to hit alt-f2 and kill the process before it moves forward .. Anyway ideas? Like I said this works fine in a normal shell: %pre chvt 3 clear CONSOLEIP="" YES=n while [ $YES != "y" ] do echo -n "Enter the console IP: " read CONSOLEIP echo "You entered: $CONSOLEIP" echo -n "Is this correct? [y/n]: " read YES done echo $CONSOLEIP > /tmp/CONSOLEIP chvt 1 %end Glenn E. Bailey III, Operational Support Services gbailey at terremark.com 972-827-4292 terremark worldwide 222 W Las Colinas Blvd Ste 350E Irving, TX 75039 From kpowell at redhat.com Thu Nov 20 16:40:44 2008 From: kpowell at redhat.com (Kyle Powell) Date: Thu, 20 Nov 2008 11:40:44 -0500 Subject: simple pre script for user input In-Reply-To: <005AB96F733E674B89550DDAE730CB4A4BF895E3E5@daa20725exc001.drtn.corp> References: <005AB96F733E674B89550DDAE730CB4A4BF895E3E5@daa20725exc001.drtn.corp> Message-ID: <4925930C.4030302@redhat.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Glenn Bailey wrote: > Playing around with a %pre script for user input, and while this works fine > in a normal shell, after I enter an IP it just hangs there and I have to hit > alt-f2 and kill the process before it moves forward .. Anyway ideas? Like I > said this works fine in a normal shell: Your script works fine for me. When you say "after I enter an IP it just hangs there" I assume that means you don't see "You entered: $CONSOLEIP" after you press enter? I would think this means the "read CONSOLEIP" is not using tty3 for STDIN. What's the name of the process you're killing? I'm guessing it's something like "ks-script-KLscIP". What's the state of the process when you kill it? S+? Get the process' PID and check what STDIN (fd 0) is pointed at: sh-3.1# ps a | grep ks-script 397 tty1 S+ 0:00 /bin/sh /tmp/ks-script-KLscIP sh-3.1# ls -l /proc/397/fd/0 lr-x------ 1 root 0 64 Nov 20 16:26 /proc/397/fd/0 -> /dev/tty3 Does it say /dev/tty3? What distro/ver/release/arch are you kickstarting? I tested your script with a RHEL 5.1 x86_64 kickstart. - -- Kyle Powell | Red Hat | Senior Consultant, RHCE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.7 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFJJZML7pTtanQdBU4RAs42AJ9II6Te4uQJkJJgAbNFOj7B2/mMdwCeI4Sf OrXhrkYNvBXxQoKm1eM+igo= =Qb4Z -----END PGP SIGNATURE----- From tony.chamberlain at lemko.com Thu Nov 20 20:23:55 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Thu, 20 Nov 2008 20:23:55 +0000 Subject: simple pre script for user input Message-ID: I'm betting the RPM program does not have its stdin from your terminal, so it is waiting on the read for the IP, which, even though you enter something, does not get read since your tty is not its stdin. -----Original Message----- From: Glenn Bailey [mailto:gbailey at terremark.com] Sent: Wednesday, November 19, 2008 07:12 PM To: kickstart-list at redhat.com Subject: simple pre script for user input Playing around with a %pre script for user input, and while this works fine in a normal shell, after I enter an IP it just hangs there and I have to hit alt-f2 and kill the process before it moves forward .. Anyway ideas? Like I said this works fine in a normal shell: %pre chvt 3 clear CONSOLEIP="" YES=n while [ $YES != "y" ] do echo -n "Enter the console IP: " read CONSOLEIP echo "You entered: $CONSOLEIP" echo -n "Is this correct? [y/n]: " read YES done echo $CONSOLEIP > /tmp/CONSOLEIP chvt 1 %end Glenn E. Bailey III, Operational Support Services gbailey at terremark.com 972-827-4292 terremark worldwide 222 W Las Colinas Blvd Ste 350E Irving, TX 75039 _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From ebrown at lanl.gov Thu Nov 20 22:09:21 2008 From: ebrown at lanl.gov (Ed Brown) Date: Thu, 20 Nov 2008 15:09:21 -0700 Subject: simple pre script for user input In-Reply-To: References: Message-ID: <4925E011.9060500@lanl.gov> Try adding this after 'chvt 3': exec /dev/tty3 2>/dev/tty3 -Ed tony.chamberlain at lemko.com wrote: > I'm betting the RPM program does not have its stdin from your terminal, > so it is waiting on the read for the IP, which, even though you enter > something, does not get read since your tty is not its stdin. > > -----Original Message----- > From: Glenn Bailey [mailto:gbailey at terremark.com] > Sent: Wednesday, November 19, 2008 07:12 PM > To: kickstart-list at redhat.com > Subject: simple pre script for user input > > Playing around with a %pre script for user input, and while this works fine > in a normal shell, after I enter an IP it just hangs there and I have to hit > alt-f2 and kill the process before it moves forward .. Anyway ideas? Like I > said this works fine in a normal shell: > > %pre > chvt 3 From tony.chamberlain at lemko.com Thu Nov 20 22:24:32 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Thu, 20 Nov 2008 22:24:32 +0000 Subject: simple pre script for user input Message-ID: How will you know it is tty3 though? maybe exec < $(tty) > $(tty) 2> /dev/$(tty) -----Original Message----- From: Ed Brown [mailto:ebrown at lanl.gov] Sent: Thursday, November 20, 2008 04:09 PM To: 'Discussion list about Kickstart' Subject: Re: simple pre script for user input Try adding this after 'chvt 3': exec /dev/tty3 2>/dev/tty3 -Ed tony.chamberlain at lemko.com wrote: > I'm betting the RPM program does not have its stdin from your terminal, > so it is waiting on the read for the IP, which, even though you enter > something, does not get read since your tty is not its stdin. > > -----Original Message----- > From: Glenn Bailey [mailto:gbailey at terremark.com] > Sent: Wednesday, November 19, 2008 07:12 PM > To: kickstart-list at redhat.com > Subject: simple pre script for user input > > Playing around with a %pre script for user input, and while this works fine > in a normal shell, after I enter an IP it just hangs there and I have to hit > alt-f2 and kill the process before it moves forward .. Anyway ideas? Like I > said this works fine in a normal shell: > > %pre > chvt 3 _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From ebrown at lanl.gov Thu Nov 20 22:38:34 2008 From: ebrown at lanl.gov (Ed Brown) Date: Thu, 20 Nov 2008 15:38:34 -0700 Subject: simple pre script for user input In-Reply-To: References: Message-ID: <4925E6EA.3070507@lanl.gov> tony.chamberlain at lemko.com wrote: > How will you know it is tty3 though? Because of the 'chvt 3' already in his script. From the man page for chvt: The command chvt N makes /dev/ttyN the foreground terminal. -Ed From tony.chamberlain at lemko.com Fri Nov 21 12:53:13 2008 From: tony.chamberlain at lemko.com (tony.chamberlain at lemko.com) Date: Fri, 21 Nov 2008 12:53:13 +0000 Subject: simple pre script for user input Message-ID: I hadn't known about chvt before. Thanks. -----Original Message----- From: Ed Brown [mailto:ebrown at lanl.gov] Sent: Thursday, November 20, 2008 04:38 PM To: 'Discussion list about Kickstart' Subject: Re: simple pre script for user input tony.chamberlain at lemko.com wrote: > How will you know it is tty3 though? Because of the 'chvt 3' already in his script. From the man page for chvt: The command chvt N makes /dev/ttyN the foreground terminal. -Ed _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From gbailey at terremark.com Fri Nov 21 17:57:28 2008 From: gbailey at terremark.com (Glenn Bailey) Date: Fri, 21 Nov 2008 11:57:28 -0600 Subject: simple pre script for user input In-Reply-To: <4925E011.9060500@lanl.gov> References: <4925E011.9060500@lanl.gov> Message-ID: <005AB96F733E674B89550DDAE730CB4A4BF895E4C4@daa20725exc001.drtn.corp> After looking at what Kyle suggested it was grabbing the input from tty1 and I was able to flip back and forth between consoles to make it work. Your suggestion below fixed the issue, thnx! FYI, this is on ESX 3.5 Update 2 .. > -----Original Message----- > From: kickstart-list-bounces at redhat.com > [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Ed Brown > Sent: Thursday, November 20, 2008 4:09 PM > To: Discussion list about Kickstart > Subject: Re: simple pre script for user input > > Try adding this after 'chvt 3': > > exec /dev/tty3 2>/dev/tty3 > > -Ed > > tony.chamberlain at lemko.com wrote: > > I'm betting the RPM program does not have its stdin from your > > terminal, so it is waiting on the read for the IP, which, > even though > > you enter something, does not get read since your tty is > not its stdin. > > > > -----Original Message----- > > From: Glenn Bailey [mailto:gbailey at terremark.com] > > Sent: Wednesday, November 19, 2008 07:12 PM > > To: kickstart-list at redhat.com > > Subject: simple pre script for user input > > > > Playing around with a %pre script for user input, and while > this works > > fine in a normal shell, after I enter an IP it just hangs > there and I > > have to hit > > alt-f2 and kill the process before it moves forward .. > Anyway ideas? > > Like I said this works fine in a normal shell: > > > > %pre > > chvt 3 > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From bryan.gartner at hp.com Mon Nov 24 17:17:15 2008 From: bryan.gartner at hp.com (Bryan Gartner) Date: Mon, 24 Nov 2008 10:17:15 -0700 Subject: yum in kickstart %post and HP Insight Mgr In-Reply-To: References: <1117568770.3743.32.camel@localhost.localdomain> Message-ID: <20081124171715.GJ6268@postino.fc.hp.com> Ugo, Haven't been on the list that long, so haven't reviewed the 3 year old history archives. On Fri, Nov 07, 2008 at 03:01:30PM +0000, Ugo Bellavance wrote: > Brian Long a ?crit : > > I'm trying to use yum in the %post section of kickstart. With well- > > written RPMs, this works great. However, I've come across the lovely HP > > Insight Manager RPMS (hpasm, hprsm, etc) which seem to run "service > > hpasm start" during the RPM postinstall (yuck). This wouldn't be too > > bad if the daemons actually disassociated from the running terminal. > > > > I notice yum install the group "yum groupinstall Insight-Manager" and > > yum exits, but the rest of the kickstart %post doesn't run. If I look > > at the hpasmd processes, they're tied to tty1 and owned by init :( If > > I hit Alt-F2 in kickstart and chroot /mnt/sysimage, service hpasm stop, > > the rest of the kickstart %post finishes. I also strace'd the /tmp/ks- > > script process and it's stuck at waitpid(-1, ) > > > > I cannot wait on a vendor escalation at this time, so I see one method > > of a workaround. Move the yum installs to a firstboot scenario instead > > of %post. I'd really like to find another way around this. > > > > Any ideas would be appreciated. Thanks. > > Hi, > > more than 3 years later, is there a solution to this? We have a bunch > of HP servers and I'm working to make the installs as automated as > possible and I'd like to end up with a system with all the HP packages > installed after anaconda finishes... How are you guys now proceeding? I am kind of confused though, since current package versions don't seem to try to start the services mentioned (since they require a configuration step). That said, given some version information for the query, there is likely a way to cause the agents not to auto-start, upon package installation, through environment variables at the very least. Also, since a yum groupinstall command was cited and the packages aren't referenced that way, I suspect some other interactions may also be in play with the original post. Feel free to follow up with the versioning info, and I'll have a look at it, bryang From mdehaan at redhat.com Tue Nov 25 16:26:38 2008 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 25 Nov 2008 11:26:38 -0500 Subject: yum in kickstart %post and HP Insight Mgr In-Reply-To: <20081124171715.GJ6268@postino.fc.hp.com> References: <1117568770.3743.32.camel@localhost.localdomain> <20081124171715.GJ6268@postino.fc.hp.com> Message-ID: <492C273E.3070204@redhat.com> Bryan Gartner wrote: > Ugo, > > Haven't been on the list that long, so haven't reviewed the 3 year old > history archives. > WRT the archives, bookmark this page. It's a google custom search engine for kickstart list's archives: http://www.google.com/coop/cse?cx=016811804524159694721:1h7btspnxtu Very useful. --Michael > On Fri, Nov 07, 2008 at 03:01:30PM +0000, Ugo Bellavance wrote: > >> Brian Long a ?crit : >> >>> I'm trying to use yum in the %post section of kickstart. With well- >>> written RPMs, this works great. However, I've come across the lovely HP >>> Insight Manager RPMS (hpasm, hprsm, etc) which seem to run "service >>> hpasm start" during the RPM postinstall (yuck). This wouldn't be too >>> bad if the daemons actually disassociated from the running terminal. >>> >>> I notice yum install the group "yum groupinstall Insight-Manager" and >>> yum exits, but the rest of the kickstart %post doesn't run. If I look >>> at the hpasmd processes, they're tied to tty1 and owned by init :( If >>> I hit Alt-F2 in kickstart and chroot /mnt/sysimage, service hpasm stop, >>> the rest of the kickstart %post finishes. I also strace'd the /tmp/ks- >>> script process and it's stuck at waitpid(-1, ) >>> >>> I cannot wait on a vendor escalation at this time, so I see one method >>> of a workaround. Move the yum installs to a firstboot scenario instead >>> of %post. I'd really like to find another way around this. >>> >>> Any ideas would be appreciated. Thanks. >>> >> Hi, >> >> more than 3 years later, is there a solution to this? We have a bunch >> of HP servers and I'm working to make the installs as automated as >> possible and I'd like to end up with a system with all the HP packages >> installed after anaconda finishes... How are you guys now proceeding? >> > > I am kind of confused though, since current package versions don't seem > to try to start the services mentioned (since they require a > configuration step). That said, given some version information for > the query, there is likely a way to cause the agents not to auto-start, > upon package installation, through environment variables at the very least. > > Also, since a yum groupinstall command was cited and the packages > aren't referenced that way, I suspect some other interactions may > also be in play with the original post. > > Feel free to follow up with the versioning info, and I'll have a look > at it, > > bryang > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From Yishai.Hadas at aladdin.com Wed Nov 26 15:26:41 2008 From: Yishai.Hadas at aladdin.com (Yishai Hadas) Date: Wed, 26 Nov 2008 17:26:41 +0200 Subject: package installation ordering problem Message-ID: Hi All, I'm using Red Hat EL3. I face a problem via kickstart installation that for now I don't find its solution. * The ln utility is part of coreutils-4.5.3-28.1.i386.rpm * The ln utility is used in a post script in other rpm of my ISO installation. * In pkgorder coreutils rpm appears before the second rpm so that ln should exist when the second rpm is installed. * I have used the /usr/lib/anaconda-runtime/genhdlist --withnumbers -fileorder option to create hdlist & hdlist2 based on mentioned pkgorder * The coreutils rpm is part of @Core group mentioned in my comps.xml * The second rpm is part of other group from comps.xml. * In ks.cfg I put explicitly @Core at the %packages section and also the package coreutils itself as follows: o %packages o @Core o coreutils o .. The problem: Looking at the end of installation at the install.log under the root directory: coreutils is installed after the second rpm and ln command is failed as it doesn't exist yet. I have changed in the pkgorder file coreutils to be the first rpm and changed also other rpms order and it seems that the change affects other rpms but not coreutils. My questions: * What is the package installation logic/order that is used by anaconda ? * How can I get/save an extra logging via installation ? o Can someone supply any bash script that will be part of ks.cfg that will help here. * Any other idea ? Any help will be appreciated. Thanks, Yishai. ********************************************************************************************** The contents of this email and any attachments are confidential. It is intended for the named recipient(s) only. If you have received this email in error please notify the system manager or? the sender immediately and do not disclose the contents to anyone or make copies. ** eSafe scanned this email for viruses, vandals and malicious content ** ********************************************************************************************** -------------- next part -------------- An HTML attachment was scrubbed... URL: From Moray.Henderson at ict.om.org Thu Nov 27 09:43:33 2008 From: Moray.Henderson at ict.om.org (Moray Henderson (ICT)) Date: Thu, 27 Nov 2008 09:43:33 -0000 Subject: package installation ordering problem In-Reply-To: References: Message-ID: >From: Yishai Hadas [mailto:Yishai.Hadas at aladdin.com] >? The ln utility is part of coreutils-4.5.3-28.1.i386.rpm >? The ln utility is used in a post script in other rpm of my ISO >installation. >? In pkgorder coreutils rpm appears before the second rpm so that ln >should exist when the second rpm is installed. >? I have used the /usr/lib/anaconda-runtime/genhdlist --withnumbers ? >fileorder option to create hdlist & hdlist2 based on mentioned pkgorder >? The coreutils rpm is part of @Core group mentioned in my comps.xml >? The second rpm is part of other group from comps.xml. > >coreutils is installed after the second rpm and ln command is failed as it >doesn?t exist yet. Hi Yishai Your second rpm needs to specify coreutils as a PreReq, not Requires in its spec file. Moray. "To err is human.? To purr, feline" From Yishai.Hadas at aladdin.com Sun Nov 30 12:14:12 2008 From: Yishai.Hadas at aladdin.com (Yishai Hadas) Date: Sun, 30 Nov 2008 14:14:12 +0200 Subject: package installation ordering problem In-Reply-To: References: Message-ID: Hi Moray, Thanks a lot - it solved the issue. Yishai. -----Original Message----- From: Moray Henderson (ICT) [mailto:Moray.Henderson at ict.om.org] Sent: Thursday, November 27, 2008 11:44 To: Discussion list about Kickstart Subject: RE: package installation ordering problem >From: Yishai Hadas [mailto:Yishai.Hadas at aladdin.com] >* The ln utility is part of coreutils-4.5.3-28.1.i386.rpm >* The ln utility is used in a post script in other rpm of my ISO >installation. >* In pkgorder coreutils rpm appears before the second rpm so that ln >should exist when the second rpm is installed. >* I have used the /usr/lib/anaconda-runtime/genhdlist --withnumbers - >fileorder option to create hdlist & hdlist2 based on mentioned pkgorder >* The coreutils rpm is part of @Core group mentioned in my comps.xml >* The second rpm is part of other group from comps.xml. > >coreutils is installed after the second rpm and ln command is failed as it >doesn't exist yet. Hi Yishai Your second rpm needs to specify coreutils as a PreReq, not Requires in its spec file. Moray. "To err is human.? To purr, feline" ********************************************************************************************** The contents of this email and any attachments are confidential. It is intended for the named recipient(s) only. If you have received this email in error please notify the system manager or? the sender immediately and do not disclose the contents to anyone or make copies. ** eSafe scanned this email for viruses, vandals and malicious content ** **********************************************************************************************