From chandan-dutta.chowdhury at hp.com Fri Jul 1 04:34:33 2005 From: chandan-dutta.chowdhury at hp.com (Chowdhury, Chandan Dutta) Date: Fri, 1 Jul 2005 10:04:33 +0530 Subject: Console based kickstart file generator Message-ID: Hello all, Thanks to all for you for the replies, what I really want to control is the partitioning sizes and the package selection part of kickstart, I wan to give the user the ability to choose the package and partition sizes all other kickstart features are hidden from them. Now how do I get the package list for the distribution and the hard disk specs. I would prefer to do it by shell script/dialog , but I am not restricting the options only to this if need be . Thanks in advance for all help Regards -CDC -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Christian.Rohrmeier at SCHERING.DE Sent: Thursday, June 30, 2005 8:09 PM To: Discussion list about Kickstart Cc: Discussion list about Kickstart; kickstart-list-bounces at redhat.com Subject: Re: Console based kickstart file generator Hi, I did something similar. I put tags in a template ks.cfg that I then replace dynamically with the apropriate text via a perl script. The original starting point was a standard kickstart script. The documentation for KS is actually quite good, other than a few omissions. Most important for me was to see others' KS files to get ideas, and then ofcourse to spend quality time writing and testing my own to learn what can and can't be done with KS. Cheers, Christian _________________ Christian Rohrmeier Schering AG Corporate IT - Infrastructure and Services Computer Systems and Operations System Administration - Research and Development Tel +49 30 468 15794 Fax +49 30 468 95794 Klaus Steden To Sent by: Discussion list about Kickstart kickstart-list-bo unces at redhat.com cc Subject 30.06.2005 16:09 Re: Console based kickstart file generator Please respond to Discussion list about Kickstart > Hello Klaus, > > Can you share the same with me ? Where can I find thepackage. > > Thanks in advance for all the help > Sorry, no such luck, I put this together by hand. The approach I've taken was to inject code in between the /sbin/loader starting (the kickstart init program) and anaconda being launched - something I wouldn't advise unless you're brave and possibly foolish. Realistically, though, the kickstart.cfg is just plain text - I grabbed the docs, and a copy of a ks file I had used on another machine, and I used that as my reference point for generating things. Make one using the GUI, and start experimenting with it, would be my best advice. hth, Klaus _______________________________________________ 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 Christian.Rohrmeier at SCHERING.DE Fri Jul 1 09:11:56 2005 From: Christian.Rohrmeier at SCHERING.DE (Christian.Rohrmeier at SCHERING.DE) Date: Fri, 1 Jul 2005 11:11:56 +0200 Subject: Console based kickstart file generator In-Reply-To: Message-ID: Hi , RHEL 3 and 4 (NOT 2.1, atleast not as delivered - you'd need to modify it) are able to use the %include tag in their kickstartscripts. By including external files that contain kickstart directives, you can dynamically create parts of the kickstart script on-the-fly. Anaconda accomplishes this by running the kickstart script twice. In your %pre script you can ask questions which you can then use to generate other parts of the kickstart script (via %include). You can use this to make a list of packages and define the disk partitioning dynamically. Since your %pre script can run sh or python, you can do tests on the system, including looking in /proc, running fdisk or parted, etc. to determine your system configuration and take action accordingly. Cheers, Christian _________________ Christian Rohrmeier Schering AG Corporate IT - Infrastructure and Services Computer Systems and Operations System Administration - Research and Development Tel +49 30 468 15794 Fax +49 30 468 95794 "Chowdhury, Chandan Dutta" "Discussion list about Kickstart" Sent by: kickstart-list-bo cc unces at redhat.com Subject RE: Console based kickstart file 01.07.2005 06:34 generator Please respond to Discussion list about Kickstart Hello all, Thanks to all for you for the replies, what I really want to control is the partitioning sizes and the package selection part of kickstart, I wan to give the user the ability to choose the package and partition sizes all other kickstart features are hidden from them. Now how do I get the package list for the distribution and the hard disk specs. I would prefer to do it by shell script/dialog , but I am not restricting the options only to this if need be . Thanks in advance for all help Regards -CDC -----Original Message----- From: kickstart-list-bounces at redhat.com [mailto:kickstart-list-bounces at redhat.com] On Behalf Of Christian.Rohrmeier at SCHERING.DE Sent: Thursday, June 30, 2005 8:09 PM To: Discussion list about Kickstart Cc: Discussion list about Kickstart; kickstart-list-bounces at redhat.com Subject: Re: Console based kickstart file generator Hi, I did something similar. I put tags in a template ks.cfg that I then replace dynamically with the apropriate text via a perl script. The original starting point was a standard kickstart script. The documentation for KS is actually quite good, other than a few omissions. Most important for me was to see others' KS files to get ideas, and then ofcourse to spend quality time writing and testing my own to learn what can and can't be done with KS. Cheers, Christian _________________ Christian Rohrmeier Schering AG Corporate IT - Infrastructure and Services Computer Systems and Operations System Administration - Research and Development Tel +49 30 468 15794 Fax +49 30 468 95794 Klaus Steden To Sent by: Discussion list about Kickstart kickstart-list-bo unces at redhat.com cc Subject 30.06.2005 16:09 Re: Console based kickstart file generator Please respond to Discussion list about Kickstart > Hello Klaus, > > Can you share the same with me ? Where can I find thepackage. > > Thanks in advance for all the help > Sorry, no such luck, I put this together by hand. The approach I've taken was to inject code in between the /sbin/loader starting (the kickstart init program) and anaconda being launched - something I wouldn't advise unless you're brave and possibly foolish. Realistically, though, the kickstart.cfg is just plain text - I grabbed the docs, and a copy of a ks file I had used on another machine, and I used that as my reference point for generating things. Make one using the GUI, and start experimenting with it, would be my best advice. hth, Klaus _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From info at hostinthebox.net Fri Jul 1 23:33:59 2005 From: info at hostinthebox.net (Dan Trainor) Date: Fri, 01 Jul 2005 16:33:59 -0700 Subject: Questions re: GRUB installation and setup in Anaconda Message-ID: <42C5D2E7.3070201@hostinthebox.net> Hello, all - I just tried to send a similar message to the rpm-list@ list to try to find an answer, but no such luck. I'm trying to figure out how GRUB works in an Anaconda installation. I need to modify some kernel lines of a grub.conf file for my installation. I've already taken apart a grub src RPM and checked through the .spec to see if the actual RPM is responsible for establishing the installation of GRUB and it's related files, but I couldn't find anything in there. Like I told those guys, I'm somewhat convinced that Anaconda actually does the setup and installation of grub, and that's where I need to look to solve this one. I guess what I'm asking here is how one might go about modifying GRUB kernel paramater lines during an Anaconda installation, be it kickstart or otherwise. Should I just make some awk hackery in %post to accomplish this task manually, or is there a more, uhm, "nicer" approach? Thanks! -dant From error27 at gmail.com Sat Jul 2 01:39:57 2005 From: error27 at gmail.com (Dan Carpenter) Date: Fri, 1 Jul 2005 18:39:57 -0700 Subject: Questions re: GRUB installation and setup in Anaconda In-Reply-To: <42C5D2E7.3070201@hostinthebox.net> References: <42C5D2E7.3070201@hostinthebox.net> Message-ID: You can easily append stuff in the kickstart file. bootloader --location=mbr --append="your_ad_here" The whole point of %post scripts is for scripting stuff though... :/ perl -pi -e 's/^(\W*kernel.*)/$1 your_ad_here/' /boot/grub/grub.conf regards, dan carpenter From info at hostinthebox.net Sat Jul 2 01:43:16 2005 From: info at hostinthebox.net (Dan Trainor) Date: Fri, 01 Jul 2005 18:43:16 -0700 Subject: Questions re: GRUB installation and setup in Anaconda In-Reply-To: References: <42C5D2E7.3070201@hostinthebox.net> Message-ID: <42C5F134.1000806@hostinthebox.net> Dan Carpenter wrote: > You can easily append stuff in the kickstart file. > bootloader --location=mbr --append="your_ad_here" > > The whole point of %post scripts is for scripting stuff though... :/ > perl -pi -e 's/^(\W*kernel.*)/$1 your_ad_here/' /boot/grub/grub.conf > > regards, > dan carpenter > Dan - Alright, I'll lay it down straight - I'm retarded. I didn't even think that the manual would have this option in there, but it does. Thanks for the time that you've taken to ask this stupid question. Thanks! -dant From hahaha_30k at yahoo.com Mon Jul 4 07:24:11 2005 From: hahaha_30k at yahoo.com (ha haha) Date: Mon, 4 Jul 2005 00:24:11 -0700 (PDT) Subject: which package contains "/usr/bin/runroot" program ?? argment for "--comp" Message-ID: <20050704072411.45938.qmail@web30209.mail.mud.yahoo.com> Hi, I'm a newbie to the customization of Fedora Core 4 installation CDs. I googled to find documents about the exact steps but failed. So I'm trying to follow the steps for FC3 installation CDs' customization at: http://64.233.179.104/search?q=cache:sf1BD5xFMBwJ:bipolar.longbros.com/%3Fp%3D10+genhdlist+productpath&hl=en But unfortunately, at step #9 the buildinstall command always failed (the #1 --#8 worked great): [root at myTestBox fedora]# buildinstall --pkgorder /0/FC4_custom/fedora/pkgfile.2005-07-03 --version 4 --product 'Fedora Core' --release 'Fedora Core 4' --prodpath Fedora /0/FC4_custom/fedora/i386 Running buildinstall... /0/FC4_custom/fedora/i386/buildinstall.tree.27028 /0/FC4_custom/fedora /0/FC4_custom/fedora Going to run buildinstall again Usage: buildinstall [--comp ] [--pkgorder ] [--version ] [--product ] [--release ] [--prodpath ] [--discs ] [root at myTestBox fedora]# I've debugged the problem a little bit, and found there were two problems I don't know how to fix: 1, where is the "/usr/bin/runroot" program? it does't locate in the anaconda* packages at all, where is it from? and what is its purposes? 2, what's the option argument(s) for option "--comp"? is there a default value for it? From the /usr/lib/anaconda-runtime/buildinstall script, I can not find a default value for the option. this problem led the script to failure and output usage information. Please help......... ____________________________________________________ Yahoo! Sports Rekindle the Rivalries. Sign up for Fantasy Football http://football.fantasysports.yahoo.com ____________________________________________________ Yahoo! Sports Rekindle the Rivalries. Sign up for Fantasy Football http://football.fantasysports.yahoo.com From Paul.Miles at quadriga.com Wed Jul 6 14:11:24 2005 From: Paul.Miles at quadriga.com (Miles, Paul) Date: Wed, 6 Jul 2005 15:11:24 +0100 Subject: Spanning a kickstart across mulitple isos Message-ID: <5D339AE687B0C8438185AE62ED06697D7A8092@ukchmail01.mail.quadriga.com> Hello, I am successfully building single iso kickstarts on Fedora core 3 - however I now need to split my kickstart build across multiple CD's. Can anyone offer some hints as to how this might be possible. Many thanks in advance Paul This e-mail is the property of Quadriga Worldwide Ltd The message (and any associated files) is intended only for the use of the individual or entity to which it is addressed and may contain information that is confidential, subject to copyright or constitutes a trade secret. If you are not the intended recipient you are hereby notified that any dissemination, copying or distribution of this message, or files associated with this message, is strictly prohibited. If you have received this message in error, please notify us immediately by replying to the message and deleting it from your computer. Messages sent to and from us may be monitored. Internet communications cannot be guaranteed to be secure or error-free as information could be intercepted, corrupted, lost, destroyed, arrive late or incomplete, or contain viruses. Therefore, we do not accept responsibility for any errors or omissions that are present in this message, or any attachment, that have arisen as a result of e-mail transmission. If verification is required, please request a hard-copy version. Any views or opinions presented are solely those of the author and do not necessarily represent those of the company. From klaus.steden at thomson.net Wed Jul 6 19:35:07 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Wed, 6 Jul 2005 15:35:07 -0400 Subject: support for jumbo frames? In-Reply-To: <20050630140945.GA842@thomson.net> References: <20050630140945.GA842@thomson.net> Message-ID: <20050706193507.GB842@thomson.net> This might be a stupid question, but is it possible to set an MTU for a network device in kickstart? I have a mix of jumbo/non-jumbo GigE machines and it would be ideal if I could kickstart them appropriately. thanks, Klaus From brilong at cisco.com Wed Jul 6 20:04:45 2005 From: brilong at cisco.com (Brian Long) Date: Wed, 06 Jul 2005 16:04:45 -0400 Subject: support for jumbo frames? In-Reply-To: <20050706193507.GB842@thomson.net> References: <20050630140945.GA842@thomson.net> <20050706193507.GB842@thomson.net> Message-ID: <1120680285.3605.7.camel@localhost.localdomain> On Wed, 2005-07-06 at 15:35 -0400, Klaus Steden wrote: > This might be a stupid question, but is it possible to set an MTU for a > network device in kickstart? This is not currently possible, but we submitted an issue tracker and it's supposed to be added in RHEL 3 U6 and RHEL 4 U2. Bugzilla is 151789. /Brian/ -- Brian Long | | | IT Data Center Systems | .|||. .|||. Cisco Linux Developer | ..:|||||||:...:|||||||:.. Phone: (919) 392-7363 | C i s c o S y s t e m s From info at hostinthebox.net Wed Jul 6 20:33:36 2005 From: info at hostinthebox.net (Dan Trainor) Date: Wed, 06 Jul 2005 13:33:36 -0700 Subject: pkgorder, genhdlist Message-ID: <42CC4020.6070701@hostinthebox.net> Hello, all - I have a few questions here regarding the use of genhdlist and pkgorder. My understanding is that something such as the following must be run, after editing comps.xml to suit one's needs: ...edit comps.xml... ...if any RPMs are included in comps.xml, throw 'em in RedHat/RPMS... pkgorder /tmp/rhel-1 i386 CentOS|tee /tmp/rhel-1/pkgorder.txt genhdlist --withnumbers --fileorder /tmp/rhel-1/pkgorder.txt --hdlist /tmp/rhel-1/RedHat/base/hdlist /tmp/rhel-1/ Now, I almost had it working up until the point where I ran genhdlist a second time. I get a Python error: AttributeError: 'dict' object has no attribute 'order' I know nothing of python, I can only imagine that this has something to do with the order in which my pkgorder.txt file was created. So then I got to thinking. Shouldn't genhdlist create the base/hdlist{2} files from scratch? I removed my existing base/hdlist{2} files, and genhdlist crapped out on me. I'm not sure why. If anyone wouldn't mind giving me some pointers, once again, I would be very appreciative. Thanks! -dant From info at hostinthebox.net Wed Jul 6 20:38:21 2005 From: info at hostinthebox.net (Dan Trainor) Date: Wed, 06 Jul 2005 13:38:21 -0700 Subject: pkgorder, genhdlist In-Reply-To: <42CC4020.6070701@hostinthebox.net> References: <42CC4020.6070701@hostinthebox.net> Message-ID: <42CC413D.70601@hostinthebox.net> Dan Trainor wrote: > Hello, all - > > I have a few questions here regarding the use of genhdlist and pkgorder. > My understanding is that something such as the following must be run, > after editing comps.xml to suit one's needs: > > ...edit comps.xml... > ...if any RPMs are included in comps.xml, throw 'em in RedHat/RPMS... > > pkgorder /tmp/rhel-1 i386 CentOS|tee /tmp/rhel-1/pkgorder.txt > > genhdlist --withnumbers --fileorder /tmp/rhel-1/pkgorder.txt --hdlist > /tmp/rhel-1/RedHat/base/hdlist /tmp/rhel-1/ > > > Now, I almost had it working up until the point where I ran genhdlist a > second time. I get a Python error: > > AttributeError: 'dict' object has no attribute 'order' > > I know nothing of python, I can only imagine that this has something to > do with the order in which my pkgorder.txt file was created. > > So then I got to thinking. Shouldn't genhdlist create the > base/hdlist{2} files from scratch? I removed my existing base/hdlist{2} > files, and genhdlist crapped out on me. I'm not sure why. > > If anyone wouldn't mind giving me some pointers, once again, I would be > very appreciative. > > Thanks! > -dant > Ah, one quick follow-up here; When the base/hdlist{2} files are removed, and I attempt to run pkgorder, I am presented with the following error: Failed to read header list /tmp/rhel-1/RedHat/base/hdlist ...which makes me conclude that for some reason the base/hdlist{2} files must exist already in order to run genhdlist. So I guess I'm not understanding something here since I had assumed that genhdlist creates the base/hdlist{2} files from scratch. Thanks again! -dant From info at hostinthebox.net Thu Jul 7 18:06:30 2005 From: info at hostinthebox.net (Dan Trainor) Date: Thu, 07 Jul 2005 11:06:30 -0700 Subject: pkgorder, genhdlist In-Reply-To: <42CC4020.6070701@hostinthebox.net> References: <42CC4020.6070701@hostinthebox.net> Message-ID: <42CD6F26.6010606@hostinthebox.net> Dan Trainor wrote: > Hello, all - > > I have a few questions here regarding the use of genhdlist and pkgorder. > My understanding is that something such as the following must be run, > after editing comps.xml to suit one's needs: > > ...edit comps.xml... > ...if any RPMs are included in comps.xml, throw 'em in RedHat/RPMS... > > pkgorder /tmp/rhel-1 i386 CentOS|tee /tmp/rhel-1/pkgorder.txt > > genhdlist --withnumbers --fileorder /tmp/rhel-1/pkgorder.txt --hdlist > /tmp/rhel-1/RedHat/base/hdlist /tmp/rhel-1/ > > > Now, I almost had it working up until the point where I ran genhdlist a > second time. I get a Python error: > > AttributeError: 'dict' object has no attribute 'order' > > I know nothing of python, I can only imagine that this has something to > do with the order in which my pkgorder.txt file was created. > > So then I got to thinking. Shouldn't genhdlist create the > base/hdlist{2} files from scratch? I removed my existing base/hdlist{2} > files, and genhdlist crapped out on me. I'm not sure why. > > If anyone wouldn't mind giving me some pointers, once again, I would be > very appreciative. > > Thanks! > -dant > Good morning - If anyone has any questions as to what kind of data I'm working with, I have attached some resources that may help: comps.xml: http://pastebin.com/309018 VC1 of Anaconda: http://img113.imageshack.us/img113/4018/screenshot5sd.png VC3 of Anaconda: http://img298.imageshack.us/img298/816/screenshot10dx.png Oddly enough, you'll notice that VC3 complains about not being able to find comps.xml, but at a later date says that it's being used from a remote server. Furthermore, the webserver's logs indicate that this file, along with a few others, were retreived properly. Again, any and all help with this would be greatly appreciated. Thanks! -dant From dky at utcc.utoronto.ca Thu Jul 7 19:14:46 2005 From: dky at utcc.utoronto.ca (Derek Yeung) Date: Thu, 7 Jul 2005 15:14:46 -0400 (EDT) Subject: serial console output Message-ID: Hi! Does anyone know how I can have all the messages during boot time, after the "Welcome to RedHat Linux ..." line output to *both* the monitor and ttyS0? Right now, with the grub.conf changes below, seem to restrict the output of messages after the welcome line to either the monitor or ttyS0 ... I'd like the output to go to both places. I've made the changes in grub.conf, specifically: -- /etc/grub.conf -- #boot=/dev/sda default=3 timeout=10 serial --unit=0 --speed=9600 --word=8 --parity=no --stop=1 terminal --timeout=2 console serial #splashimage=(hd0,0)/grub/splash.xpm.gz title Red Hat Enterprise Linux AS (2.4.21-32.0.1.ELsmp) root (hd0,0) kernel /vmlinuz-2.4.21-32.0.1.ELsmp ro root=/dev/vg00/root console=tty0 console=ttyS0,9600n8 ------------------- I recall there was another line, somewhere in /etc/sysconfig that i needed to change? Suggestions? Very much appreciated, /dky From ian at smallworld.cx Fri Jul 8 19:54:15 2005 From: ian at smallworld.cx (Ian Leonard) Date: Fri, 08 Jul 2005 20:54:15 +0100 Subject: Can't find cdrom on boot Message-ID: <42CED9E7.6000207@smallworld.cx> Greetings, I have built a custom install disk using FC2 on cd. When booting, all goes well, the ks.cfg file is found but it fails when it tries to mount the cd. I get a dialog box saying that the cd was not found in any of the CDROM drives. I see that on one of the VT's there is a message "Unable to load NLS charset utf8". I don't know if that's anything to do with it. I realise this is not much to go on but any advice would be appreciated. I guess the next step will be to recompile anaconda with more diagnostics. -- Ian Leonard Please ignore spelling and punctuation - I did. From error27 at gmail.com Fri Jul 8 20:57:26 2005 From: error27 at gmail.com (Dan Carpenter) Date: Fri, 8 Jul 2005 22:57:26 +0200 Subject: Can't find cdrom on boot In-Reply-To: <42CED9E7.6000207@smallworld.cx> References: <42CED9E7.6000207@smallworld.cx> Message-ID: On 7/8/05, Ian Leonard wrote: > Greetings, > > I have built a custom install disk using FC2 on cd. When booting, all > goes well, the ks.cfg file is found but it fails when it tries to mount > the cd. I get a dialog box saying that the cd was not found in any of > the CDROM drives. > What does the .discinfo file say on your customized CDROM? (shot in the dark) regards, dan From ian at smallworld.cx Fri Jul 8 22:15:44 2005 From: ian at smallworld.cx (Ian Leonard) Date: Fri, 08 Jul 2005 23:15:44 +0100 Subject: Can't find cdrom on boot In-Reply-To: References: <42CED9E7.6000207@smallworld.cx> Message-ID: <42CEFB10.3050109@smallworld.cx> Dan Carpenter wrote: > On 7/8/05, Ian Leonard wrote: > >>Greetings, >> >>I have built a custom install disk using FC2 on cd. When booting, all >>goes well, the ks.cfg file is found but it fails when it tries to mount >>the cd. I get a dialog box saying that the cd was not found in any of >>the CDROM drives. >> > > > What does the .discinfo file say on your customized CDROM? 1120841598.592341 Fedora Core 2 i386 1 RedHat/base RedHat/RPMS RedHat/pixmaps I have no idea what the first line is about. > (shot in the dark) Thanks. Keep shooting. > regards, > dan -- Ian Leonard Please ignore spelling and punctuation - I did. From rakesh.chaudhari at gmail.com Sat Jul 9 06:00:32 2005 From: rakesh.chaudhari at gmail.com (rakesh chaudhari) Date: Sat, 9 Jul 2005 11:30:32 +0530 Subject: where is getGroupPkgs.py and syncRpms.py? Message-ID: <2e5b29bc0507082300ee52191@mail.gmail.com> hi, i have gone thru' "Hacking Redhat Kickstart article" and i'm following same steps for Redhat 9 but unfoutunately unable to get above scripts..... are those steps applicable to RH9? can anyone help me??? thanks in advance. From brett_schwarz at yahoo.com Sat Jul 9 07:04:46 2005 From: brett_schwarz at yahoo.com (Brett Schwarz) Date: Sat, 9 Jul 2005 00:04:46 -0700 (PDT) Subject: where is getGroupPkgs.py and syncRpms.py? In-Reply-To: <2e5b29bc0507082300ee52191@mail.gmail.com> Message-ID: <20050709070446.41156.qmail@web40602.mail.yahoo.com> you can get the scripts from http://www.bschwarz.com/rh/ I believe people have gotten the procedure to work with RH9, although I believe with RH9 you need to make sure the .discinfo file was copied over. There is also a modified syncRpms.py script for RH9 at the bottom of that page... HTH, --brett --- rakesh chaudhari wrote: > hi, > > i have gone thru' "Hacking Redhat Kickstart > article" and i'm > following same steps for Redhat 9 > but unfoutunately unable to get above scripts..... > > are those steps applicable to RH9? > > can anyone help me??? > > thanks in advance. > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > --brett __________________________________ Yahoo! Mail Stay connected, organized, and protected. Take the tour: http://tour.mail.yahoo.com/mailtour.html From susanta at vfortress.com Mon Jul 11 13:19:26 2005 From: susanta at vfortress.com (Susanta Kumar Padhi) Date: Mon, 11 Jul 2005 18:49:26 +0530 Subject: Create Custom Boot CD. Message-ID: <1121087966.9271.85.camel@susant.padhi.com> Hi All, I am Susanta Padhi and having Fedora Core 2 Linux box. I am interest in create a bootable custom CD. I don't need all the RPMS of the Fedora Core2.I need some of them and also my own packages.I am unknown to this CD creation. Please help me to how to take approach. Regards Susant From dky at utcc.utoronto.ca Mon Jul 11 16:56:50 2005 From: dky at utcc.utoronto.ca (Derek Yeung) Date: Mon, 11 Jul 2005 12:56:50 -0400 (EDT) Subject: serial console output In-Reply-To: <942A778F73A72141889C1F53A8EAA1C52B9A1C@MSGMMKCLM2WIN.DMN1.FMR.COM> References: <942A778F73A72141889C1F53A8EAA1C52B9A1C@MSGMMKCLM2WIN.DMN1.FMR.COM> Message-ID: I've discovered if we remove the terminal --timeout=2 console serial line in grub.conf, all *grub* messages will go to both ttyS0 and tty0. But still don't have ideas for how to get the messages after the "Welcome to RedHat Linux..." line to show up. /dky On Mon, 11 Jul 2005, Keegan, Gordon wrote: > I've had the same issue under Lilo as well as grub and haven't found a > solution for either one. Theoretically, specifying tty0 and ttyS0 > should setup both devices to receive console output, but it doesn't seem > to work that way. Here's hoping that someone else has had better luck. > > Gordon Keegan > > > -----Original Message----- > From: Derek Yeung [mailto:dky at utcc.utoronto.ca] > Sent: Thursday, July 07, 2005 3:15 PM > To: kickstart-list at redhat.com > Subject: serial console output > > > > Hi! > > Does anyone know how I can have all the messages during boot time, after > > the > "Welcome to RedHat Linux ..." > > line output to *both* the monitor and ttyS0? Right now, with the > grub.conf changes below, seem to restrict the output of messages after > the > welcome line to either the monitor or ttyS0 ... I'd like the output to > go > to both places. > > > I've made the changes in grub.conf, specifically: > > > -- /etc/grub.conf -- > #boot=/dev/sda > default=3 > timeout=10 > serial --unit=0 --speed=9600 --word=8 --parity=no --stop=1 > terminal --timeout=2 console serial > #splashimage=(hd0,0)/grub/splash.xpm.gz > title Red Hat Enterprise Linux AS (2.4.21-32.0.1.ELsmp) > root (hd0,0) > kernel /vmlinuz-2.4.21-32.0.1.ELsmp ro root=/dev/vg00/root > console=tty0 console=ttyS0,9600n8 > > ------------------- > > I recall there was another line, somewhere in /etc/sysconfig that i > needed > to change? > > Suggestions? > > > Very much appreciated, > /dky > > > From info at hostinthebox.net Mon Jul 11 17:27:01 2005 From: info at hostinthebox.net (Dan Trainor) Date: Mon, 11 Jul 2005 10:27:01 -0700 Subject: where is getGroupPkgs.py and syncRpms.py? In-Reply-To: <2e5b29bc0507082300ee52191@mail.gmail.com> References: <2e5b29bc0507082300ee52191@mail.gmail.com> Message-ID: <42D2ABE5.5040706@hostinthebox.net> rakesh chaudhari wrote: > hi, > > i have gone thru' "Hacking Redhat Kickstart article" and i'm > following same steps for Redhat 9 > but unfoutunately unable to get above scripts..... > > are those steps applicable to RH9? > > can anyone help me??? > > thanks in advance. > http://www.bschwarz.com/rh/ Thanks -dant From info at hostinthebox.net Mon Jul 11 17:27:54 2005 From: info at hostinthebox.net (Dan Trainor) Date: Mon, 11 Jul 2005 10:27:54 -0700 Subject: Create Custom Boot CD. In-Reply-To: <1121087966.9271.85.camel@susant.padhi.com> References: <1121087966.9271.85.camel@susant.padhi.com> Message-ID: <42D2AC1A.2040709@hostinthebox.net> Susanta Kumar Padhi wrote: > Hi All, > > I am Susanta Padhi and having Fedora Core 2 Linux box. I am interest in > create a bootable custom CD. I don't need all the RPMS of the Fedora > Core2.I need some of them and also my own packages.I am unknown to this > CD creation. Please help me to how to take approach. > > Regards > Susant > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > Susant - This has been an uphill battle for me in most part due to Anaconda not being very well documented, and the process of actually making a CD in the form that you speak of, next to impossible. I have been researching this for a few months now to varying degrees of success because information on the subject is very limited, regardless of what anyone else on this list says. Information is only plentiful on the Internet if you're interested in making a custom RH6 or RH7.2 CD. When I'm all finished, I plan on writing a pretty complete guide to what I had done, hopefully so that others can learn from it and use it and go on about their lives and not have to waste months on this. Anaconda has an amazing amount of functionality, and purpose. The only problem is, seems as if no one knows how to really use it and prepare media for it to use. I'll be posting here when I'm all done. Thanks -dant From avery at port25.com Mon Jul 11 17:34:53 2005 From: avery at port25.com (Avery Fay) Date: Mon, 11 Jul 2005 13:34:53 -0400 Subject: Create Custom Boot CD. In-Reply-To: <42D2AC1A.2040709@hostinthebox.net> References: <1121087966.9271.85.camel@susant.padhi.com> <42D2AC1A.2040709@hostinthebox.net> Message-ID: <1121103293.7148.3.camel@localhost.localdomain> On Mon, 2005-07-11 at 10:27 -0700, Dan Trainor wrote: > This has been an uphill battle for me in most part due to Anaconda not > being very well documented, and the process of actually making a CD in > the form that you speak of, next to impossible. I just finished doing this using RHEL4 as a base. I got the packages all on one CD including custom RPMs and conf files. It was a PITA. > When I'm all finished, I plan on writing a pretty complete guide to what > I had done, hopefully so that others can learn from it and use it and go > on about their lives and not have to waste months on this. Anaconda has > an amazing amount of functionality, and purpose. The only problem is, > seems as if no one knows how to really use it and prepare media for it > to use. I've been meaning to write up something too, although my time right now is limited. I'll be able to answer any questions anyone has though (hopefully). As a start, here's some sites I found useful (most are outdated): http://sysadmin.cs.caltech.edu/docs/help/Linux/rh7.2-ks-cd http://tldp.org/HOWTO/RedHat-CD-HOWTO/ http://www.harkness.co.uk/other/kickstart.html https://engineering.purdue.edu/ECN/Resources/KnowledgeBase/Docs/20050217134106 Hope these help, -- Avery Fay From klaus.steden at thomson.net Mon Jul 11 17:57:14 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Mon, 11 Jul 2005 13:57:14 -0400 Subject: Create Custom Boot CD. In-Reply-To: <1121103293.7148.3.camel@localhost.localdomain> References: <1121087966.9271.85.camel@susant.padhi.com> <42D2AC1A.2040709@hostinthebox.net> <1121103293.7148.3.camel@localhost.localdomain> Message-ID: <20050711175714.GG73653@thomson.net> > > This has been an uphill battle for me in most part due to Anaconda not > > being very well documented, and the process of actually making a CD in > > the form that you speak of, next to impossible. > > I just finished doing this using RHEL4 as a base. I got the packages all > on one CD including custom RPMs and conf files. It was a PITA. > It's less of a PITA than doing network-based installs. If you can swing it, I would suggest that as an alternative to custom boot CDs since it can involve a lot less misery when generating hdlist files. Dan - good to hear about the documentation project - looking forward to seeing it, and would be happy to help out. Klaus From info at hostinthebox.net Mon Jul 11 18:02:24 2005 From: info at hostinthebox.net (Dan Trainor) Date: Mon, 11 Jul 2005 11:02:24 -0700 Subject: Create Custom Boot CD. In-Reply-To: <20050711175714.GG73653@thomson.net> References: <1121087966.9271.85.camel@susant.padhi.com> <42D2AC1A.2040709@hostinthebox.net> <1121103293.7148.3.camel@localhost.localdomain> <20050711175714.GG73653@thomson.net> Message-ID: <42D2B430.1020708@hostinthebox.net> Klaus Steden wrote: >>>This has been an uphill battle for me in most part due to Anaconda not >>>being very well documented, and the process of actually making a CD in >>>the form that you speak of, next to impossible. >> >>I just finished doing this using RHEL4 as a base. I got the packages all >>on one CD including custom RPMs and conf files. It was a PITA. >> > > It's less of a PITA than doing network-based installs. > > If you can swing it, I would suggest that as an alternative to custom boot CDs > since it can involve a lot less misery when generating hdlist files. > > Dan - good to hear about the documentation project - looking forward to seeing > it, and would be happy to help out. > > Klaus > Klaus - Yeah, I hope it turns out well. However, I'm still stumped here. I managed to generate my hdlist files, which seem alright. However, the installation still dies. VC1 immediately drops to "install exited abnormally" after comps.rpm is retrieved from the HTTP server. VC3 is a little less than helpful by saying "WARNING: not all packages in hdlist had order tag". I'm not quite sure what this means, however, I'm researching it. How am I supposed to rebuild comps.rpm? I'm going to rip it apart and take a look, but I'll also wait for your response in the meantime, if you've got any ideas. Thanks again! -dant From avery at port25.com Mon Jul 11 18:12:23 2005 From: avery at port25.com (Avery Fay) Date: Mon, 11 Jul 2005 14:12:23 -0400 Subject: Create Custom Boot CD. In-Reply-To: <42D2B430.1020708@hostinthebox.net> References: <1121087966.9271.85.camel@susant.padhi.com> <42D2AC1A.2040709@hostinthebox.net> <1121103293.7148.3.camel@localhost.localdomain> <20050711175714.GG73653@thomson.net> <42D2B430.1020708@hostinthebox.net> Message-ID: <1121105543.7832.6.camel@localhost.localdomain> On Mon, 2005-07-11 at 11:02 -0700, Dan Trainor wrote: > VC1 immediately drops to "install exited abnormally" after comps.rpm is > retrieved from the HTTP server. VC3 is a little less than helpful by > saying "WARNING: not all packages in hdlist had order tag". I'm not > quite sure what this means, however, I'm researching it. What version of RH are you basing your install off of? I never had to touch comps.rpm only comps.xml. > How am I supposed to rebuild comps.rpm? I'm going to rip it apart and > take a look, but I'll also wait for your response in the meantime, if > you've got any ideas. After changes to comps.xml and deleting RPMS/adding RPMS, here's what I do: /usr/lib/anaconda-runtime/genhdlist /usr/lib/anaconda-runtime/pkgorder i386 > \ pkgorder.txt /usr/lib/anaconda-runtime/genhdlist --withnumbers \ --fileorder=pkgorder.txt Yeah, you have to run genhdlist twice. No idea why, but it's required. Also, obviously is just the path to the RH install hierarchy if you are doing a network install (this path is the one with the RedHat directory in it). -- Avery Fay From info at hostinthebox.net Mon Jul 11 18:21:10 2005 From: info at hostinthebox.net (Dan Trainor) Date: Mon, 11 Jul 2005 11:21:10 -0700 Subject: Create Custom Boot CD. In-Reply-To: <1121105543.7832.6.camel@localhost.localdomain> References: <1121087966.9271.85.camel@susant.padhi.com> <42D2AC1A.2040709@hostinthebox.net> <1121103293.7148.3.camel@localhost.localdomain> <20050711175714.GG73653@thomson.net> <42D2B430.1020708@hostinthebox.net> <1121105543.7832.6.camel@localhost.localdomain> Message-ID: <42D2B896.2030803@hostinthebox.net> Avery Fay wrote: > On Mon, 2005-07-11 at 11:02 -0700, Dan Trainor wrote: > >>VC1 immediately drops to "install exited abnormally" after comps.rpm is >>retrieved from the HTTP server. VC3 is a little less than helpful by >>saying "WARNING: not all packages in hdlist had order tag". I'm not >>quite sure what this means, however, I'm researching it. > > > What version of RH are you basing your install off of? I never had to > touch comps.rpm only comps.xml. > > >>How am I supposed to rebuild comps.rpm? I'm going to rip it apart and >>take a look, but I'll also wait for your response in the meantime, if >>you've got any ideas. > > > After changes to comps.xml and deleting RPMS/adding RPMS, here's what I > do: > > /usr/lib/anaconda-runtime/genhdlist > /usr/lib/anaconda-runtime/pkgorder i386 > \ > pkgorder.txt > /usr/lib/anaconda-runtime/genhdlist --withnumbers \ > --fileorder=pkgorder.txt > > Yeah, you have to run genhdlist twice. No idea why, but it's required. > Also, obviously is just the path to the RH > install hierarchy if you are doing a network install (this path is the > one with the RedHat directory in it). > Thanks for the response, Avery - I think the problem is, I'm actually editing comps.xml and adding my own packages to @base. I think this conflicts with comps.rpm, which does not have my packages listed in it's own comps.xml. However, since Anaconda is a bit less than friendly with it's debugging information, I cannot verify this as a fact. Thanks -dant From alexander_rau at yahoo.com Mon Jul 11 18:26:48 2005 From: alexander_rau at yahoo.com (Alexander Rau) Date: Mon, 11 Jul 2005 11:26:48 -0700 (PDT) Subject: Create Custom Boot CD. In-Reply-To: <42D2B896.2030803@hostinthebox.net> Message-ID: <20050711182648.85458.qmail@web52107.mail.yahoo.com> A while back a couple of guys and I started the "Anaconda Documentation project" http://rau.homedns.org/twiki/bin/view/Anaconda/WebHome It is a bit outdated but it might be a good starting point and you are more than welcome to make updates as you go along.... Thanks AR From info at hostinthebox.net Mon Jul 11 18:32:08 2005 From: info at hostinthebox.net (Dan Trainor) Date: Mon, 11 Jul 2005 11:32:08 -0700 Subject: Create Custom Boot CD. In-Reply-To: <20050711182648.85458.qmail@web52107.mail.yahoo.com> References: <20050711182648.85458.qmail@web52107.mail.yahoo.com> Message-ID: <42D2BB28.6090809@hostinthebox.net> Alexander Rau wrote: > A while back a couple of guys and I started the > "Anaconda Documentation project" > > http://rau.homedns.org/twiki/bin/view/Anaconda/WebHome > > It is a bit outdated but it might be a good starting > point and you are more than welcome to make updates as > you go along.... > > Thanks > > AR > Might not be a bad idea. I've seen this page advertised around the 'net while I've been looking for information, but this is the first time I've actually been able to connect ;) I'll take a look over it. Thanks -dant From alexander_rau at yahoo.com Mon Jul 11 19:00:12 2005 From: alexander_rau at yahoo.com (Alexander Rau) Date: Mon, 11 Jul 2005 12:00:12 -0700 (PDT) Subject: Create Custom Boot CD. In-Reply-To: <42D2BB28.6090809@hostinthebox.net> Message-ID: <20050711190012.53021.qmail@web52103.mail.yahoo.com> > while I've been looking for information, but this is > the first time I've > actually been able to connect ;) > > I'll take a look over it. > > Thanks > -dant > Sorry about the downtime, I moved my server and forgot to update the DNS for that URL. I am making the best effort to keep this site up. Sorry again AR From info at hostinthebox.net Mon Jul 11 19:02:00 2005 From: info at hostinthebox.net (Dan Trainor) Date: Mon, 11 Jul 2005 12:02:00 -0700 Subject: Create Custom Boot CD. In-Reply-To: <20050711190012.53021.qmail@web52103.mail.yahoo.com> References: <20050711190012.53021.qmail@web52103.mail.yahoo.com> Message-ID: <42D2C228.1050909@hostinthebox.net> Alexander Rau wrote: > >>while I've been looking for information, but this is >>the first time I've >>actually been able to connect ;) >> >>I'll take a look over it. >> >>Thanks >>-dant >> > > Sorry about the downtime, I moved my server and forgot > to update the DNS for that URL. I am making the best > effort to keep this site up. > > Sorry again > > AR > Alex - No worries, I appreciate the effort. Thanks -dant From teng at dataway.com Mon Jul 11 19:13:19 2005 From: teng at dataway.com (Tedman Eng) Date: Mon, 11 Jul 2005 12:13:19 -0700 Subject: serial console output Message-ID: <37ED92F9890FAF4BB947613C66FF8B1AFC7574@dw-mail.dataway.com> Pick one as your primary. Bootloader and kernel and messages can echo to both consoles, but init scripts will only echo to one console (your primary console). The order of the "console=" in the kernel line in grub.conf determines which is the primary. You can have multiple "console=", ie "console=ttyS0,9600n8 console=ttyS1,9600n8 console=tty0", but only the primary display will show init script output. See: http://www.tldp.org/HOWTO/Remote-Serial-Console-HOWTO/bugs-monitor.html -----Original Message----- From: Derek Yeung [mailto:dky at utcc.utoronto.ca] Sent: Monday, July 11, 2005 9:57 AM To: Keegan, Gordon Cc: kickstart-list at redhat.com Subject: RE: serial console output I've discovered if we remove the terminal --timeout=2 console serial line in grub.conf, all *grub* messages will go to both ttyS0 and tty0. But still don't have ideas for how to get the messages after the "Welcome to RedHat Linux..." line to show up. /dky On Mon, 11 Jul 2005, Keegan, Gordon wrote: > I've had the same issue under Lilo as well as grub and haven't found a > solution for either one. Theoretically, specifying tty0 and ttyS0 > should setup both devices to receive console output, but it doesn't seem > to work that way. Here's hoping that someone else has had better luck. > > Gordon Keegan > > > -----Original Message----- > From: Derek Yeung [mailto:dky at utcc.utoronto.ca] > Sent: Thursday, July 07, 2005 3:15 PM > To: kickstart-list at redhat.com > Subject: serial console output > > > > Hi! > > Does anyone know how I can have all the messages during boot time, after > > the > "Welcome to RedHat Linux ..." > > line output to *both* the monitor and ttyS0? Right now, with the > grub.conf changes below, seem to restrict the output of messages after > the > welcome line to either the monitor or ttyS0 ... I'd like the output to > go > to both places. > > > I've made the changes in grub.conf, specifically: > > > -- /etc/grub.conf -- > #boot=/dev/sda > default=3 > timeout=10 > serial --unit=0 --speed=9600 --word=8 --parity=no --stop=1 > terminal --timeout=2 console serial > #splashimage=(hd0,0)/grub/splash.xpm.gz > title Red Hat Enterprise Linux AS (2.4.21-32.0.1.ELsmp) > root (hd0,0) > kernel /vmlinuz-2.4.21-32.0.1.ELsmp ro root=/dev/vg00/root > console=tty0 console=ttyS0,9600n8 > > ------------------- > > I recall there was another line, somewhere in /etc/sysconfig that i > needed > to change? > > Suggestions? > > > Very much appreciated, > /dky > > > _______________________________________________ Kickstart-list mailing list Kickstart-list at redhat.com https://www.redhat.com/mailman/listinfo/kickstart-list From info at hostinthebox.net Mon Jul 11 23:06:38 2005 From: info at hostinthebox.net (Dan Trainor) Date: Mon, 11 Jul 2005 16:06:38 -0700 Subject: Using ks.cfg off of cd Message-ID: <42D2FB7E.6090607@hostinthebox.net> Hello, all - I've done it a while ago, but I since cannot remember. I'm trying to get a machine to boot off of a CD, modifying isolinux/isolinux.cfg to take a "ks=cdrom:/ks.cfg" option. I'm quite certain that the isolinux/isolinux.cfg file is the correct one to modify. Can I get someone to verify this? Thanks! -dant From klaus.steden at thomson.net Mon Jul 11 23:10:53 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Mon, 11 Jul 2005 19:10:53 -0400 Subject: Using ks.cfg off of cd In-Reply-To: <42D2FB7E.6090607@hostinthebox.net> References: <42D2FB7E.6090607@hostinthebox.net> Message-ID: <20050711231053.GJ73653@thomson.net> > Hello, all - > > I've done it a while ago, but I since cannot remember. > > I'm trying to get a machine to boot off of a CD, modifying > isolinux/isolinux.cfg to take a "ks=cdrom:/ks.cfg" option. > > I'm quite certain that the isolinux/isolinux.cfg file is the correct one > to modify. Can I get someone to verify this? > Yup, that's the file. From info at hostinthebox.net Tue Jul 12 18:09:57 2005 From: info at hostinthebox.net (Dan Trainor) Date: Tue, 12 Jul 2005 11:09:57 -0700 Subject: Anaconda Debugging Message-ID: <42D40775.2040301@hostinthebox.net> Hello, all - I have an RPM that gets installed during the whole Anaconda process. It's a bit odd, in the fact that it starts a daemon during install, then runs a client program, and feeds some data to the client program, and then kills the daemon. The RPM was working perfectly prior to whatever change I made yesterday, which I can't find. I don't have a backup, so I'm up crap creek here. The RPM works properly when installed at a later time, i.e. on a system that has already been installed. But the RPM does not work when installed using Anaconda. Instead, it craps out and returns an error that I don't understand, since it's proprietary software - and we all know how convoluted that can be. So what I'd like to do at this point, even if it does not help fix the problem, is to see some more verbose Anaconda output. Is there any way that I can set my kickstart up to tell Anaconda to have pretty much debug output? This would be great. Thanks -dant From ltbraswell at yahoo.com Wed Jul 13 18:53:33 2005 From: ltbraswell at yahoo.com (Lance Braswell) Date: Wed, 13 Jul 2005 11:53:33 -0700 (PDT) Subject: probing a system for drives in the pre phase Message-ID: <20050713185333.22994.qmail@web40914.mail.yahoo.com> Found a reference to this in the archives I believe. Kudos to Jesse Keating: http://geek.j2solutions.net/stuff/autokickfc2.py It's an example for doing some %pre probing to dynamically generate a ks.cfg based on probed for disk drives. I have a requirement to support automated installations (from CD) of RHEL 4 on somewhat arbitrary PC hardware. I want to have complete control over most of the install process but I need the customer to basically make some decisions about the boot/OS drives to be used before the rest of the kickstart process begins. So what I was planning on doing was using the above autokickfc2.py program as a starting point for my script. I would run my program as a %pre step. I want the system to boot up and present a simple text menu that presents the users with a list of found drives. The user would select one or more of these drives. From those selected drives, I would create a single root volume group and then generate my logical volumes accordingly (based on size mainly). I would be using the abstraction provided by the python isys module because I don't think I need to worry about whether it's ide, scsi, or hardware raid. In other words I just want "harddrives" as anaconda would see them and deal with them in an abstract way for presentation to the customer and for partitioning. So here are my questions: Does this seem feasible? Haven't been able to find any external documentation for the parted and isys python modules. Anyone have some pointers? Will I have access the the console for user input during the %pre phase when my script is running? What is a good python module for handling user input? I'm pretty new to PC hardware, given a list of harddrives that anaconda (and my script) would see, how can I determine if it _would be_ bootable? Sorry to ask such noob questions. I can hack perl but only grok python a little. ____________________________________________________ Start your day with Yahoo! - make it your home page http://www.yahoo.com/r/hs From vanallsburg at hope.edu Wed Jul 13 19:49:30 2005 From: vanallsburg at hope.edu (Paul VanAllsburg) Date: Wed, 13 Jul 2005 15:49:30 -0400 Subject: NFS using eth1 installation Message-ID: <42D69D63@hope.edu> How do I create a kickstart file that will do a NFS using eth1 installation? I can successfully do a manual boot: askmethod, specify NFS with eth1, but it does not show up in the anaconda-ks.cfg file. I end up with: nfs --server=192.168.2.100 --dir=/data/iso/fc3/i386 When I run a automated install using the anaconda generated script I get 'That directory could not be mounted from the server'. I've been able to move the NFS share over to a system on eth0 and get it to work but not one on eth1. The network/share on eth1 is functional... Thanks, Paul Van Allsburg From klaus.steden at thomson.net Wed Jul 13 20:08:27 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Wed, 13 Jul 2005 16:08:27 -0400 Subject: NFS using eth1 installation In-Reply-To: <42D69D63@hope.edu> References: <42D69D63@hope.edu> Message-ID: <20050713200827.GR73653@thomson.net> Paul, The kickstart option you're looking for is 'ksdevice=', i.e. ksdevice=eth1 should do the trick. Klaus > How do I create a kickstart file that will do a NFS using eth1 installation? > > I can successfully do a manual boot: askmethod, specify NFS with eth1, but it > does not show up in the anaconda-ks.cfg file. I end up with: > nfs --server=192.168.2.100 --dir=/data/iso/fc3/i386 > > When I run a automated install using the anaconda generated script I get 'That > directory could not be mounted from the server'. I've been able to move the > NFS share over to a system on eth0 and get it to work but not one on eth1. > The network/share on eth1 is functional... > > Thanks, > Paul Van Allsburg > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Klaus Steden | Senior Systems Administrator | Technicolor Creative Services | TODO: Toronto | 1) Learn to use my new Unix account. klaus.steden at thomson.net | 2) Learn how to change this list. Phone: (416) 585-9995 | Fax: (416) 364-1585 | From info at hostinthebox.net Wed Jul 13 22:29:36 2005 From: info at hostinthebox.net (Dan Trainor) Date: Wed, 13 Jul 2005 15:29:36 -0700 Subject: Installing an RPM later on in pkgorder Message-ID: <42D595D0.2020307@hostinthebox.net> Hello, all - I've got another one for ya. I have an RPM that I made which I've integrated into comps.xml, and it works fine that way. However, it does start a TCP-based server, which itself as a client connects to, to set up some default configuration options. A few days ago, this RPM was installed next to last in Anaconda, and it worked fine. However, I have since added some more RPMs to comps.xml, and now this particular RPM is being installed right in the middle of Anaconda. I believe that this is dictated by using pkgorder. This RPM is no longer installing properly during Anaconda, and some of the errors that this RPM is giving upon install, indicate that something regarding the TCP/IP subsystem is either not installed yet, or not configured properly, as if waiting for another RPM to install later via ANaconda and set up the since "broken" service. I guess what I'm asking is, is there any way to specify in which order RPMs are installed, inside of Anaconda? I understand that pkgorder does it's best to guess to correct order in which to line up packages for install with Anaconda, but there's got to be a way to make a few exceptions to this, and install the packages in the order that I see fit. If anyone has any experience with this, or if they can advise me a bit, I would greatly appreciate it. Thanks -dant From info at hostinthebox.net Wed Jul 13 22:41:30 2005 From: info at hostinthebox.net (Dan Trainor) Date: Wed, 13 Jul 2005 15:41:30 -0700 Subject: Installing an RPM later on in pkgorder In-Reply-To: <42D595D0.2020307@hostinthebox.net> References: <42D595D0.2020307@hostinthebox.net> Message-ID: <42D5989A.5090000@hostinthebox.net> Dan Trainor wrote: > Hello, all - > > I've got another one for ya. > > I have an RPM that I made which I've integrated into comps.xml, and it > works fine that way. However, it does start a TCP-based server, which > itself as a client connects to, to set up some default configuration > options. > > A few days ago, this RPM was installed next to last in Anaconda, and it > worked fine. However, I have since added some more RPMs to comps.xml, > and now this particular RPM is being installed right in the middle of > Anaconda. I believe that this is dictated by using pkgorder. This RPM > is no longer installing properly during Anaconda, and some of the errors > that this RPM is giving upon install, indicate that something regarding > the TCP/IP subsystem is either not installed yet, or not configured > properly, as if waiting for another RPM to install later via ANaconda > and set up the since "broken" service. > > I guess what I'm asking is, is there any way to specify in which order > RPMs are installed, inside of Anaconda? I understand that pkgorder does > it's best to guess to correct order in which to line up packages for > install with Anaconda, but there's got to be a way to make a few > exceptions to this, and install the packages in the order that I see fit. > > If anyone has any experience with this, or if they can advise me a bit, > I would greatly appreciate it. > > Thanks > -dant > Hello - I tried placing my RPM at the "end" of my pkginfo file that's used on the second genhdlist pass. The RPM was installed last, but unfortunately I ran into the same problem. I suspect now that my problem is elsewhere. Thanks for the time -dant From error27 at gmail.com Wed Jul 13 23:02:34 2005 From: error27 at gmail.com (Dan Carpenter) Date: Wed, 13 Jul 2005 16:02:34 -0700 Subject: probing a system for drives in the pre phase In-Reply-To: <20050713185333.22994.qmail@web40914.mail.yahoo.com> References: <20050713185333.22994.qmail@web40914.mail.yahoo.com> Message-ID: I haven't been able to figure out how to get access to the console for a user friendly UI... I do everything in a different console and that's not as good. Has anyone done this? regards, dan carpenter From klaus.steden at thomson.net Wed Jul 13 23:07:01 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Wed, 13 Jul 2005 19:07:01 -0400 Subject: probing a system for drives in the pre phase In-Reply-To: References: <20050713185333.22994.qmail@web40914.mail.yahoo.com> Message-ID: <20050713230701.GS73653@thomson.net> > I haven't been able to figure out how to get access to the console for > a user friendly UI... I do everything in a different console and > that's not as good. > > Has anyone done this? > Yes. I customized the netstg2.img, inserting my own script called 'anaconda'. /sbin/loader invokes /usr/bin/anaconda when bootstrapping begins (after loading some drivers and chewing on some of /proc/cmdline); I put a script in this stage which does what I want, then I have it exec the real anaconda script once it's done running. I suppose it's a little dodgy, but it was the way to get what I wanted accomplished. It should be possible to insert such a mechanism in any of the boot images (hdstg2.img, netstg2.img, etc.). If you want more detail, email me off list, or point me at a Wiki and I'll be happy to fill you in! hth, Klaus From info at hostinthebox.net Thu Jul 14 00:12:00 2005 From: info at hostinthebox.net (Dan Trainor) Date: Wed, 13 Jul 2005 17:12:00 -0700 Subject: Installing an RPM later on in pkgorder In-Reply-To: <42D5989A.5090000@hostinthebox.net> References: <42D595D0.2020307@hostinthebox.net> <42D5989A.5090000@hostinthebox.net> Message-ID: <42D5ADD0.8070502@hostinthebox.net> Dan Trainor wrote: > Dan Trainor wrote: > >>Hello, all - >> >>I've got another one for ya. >> >>I have an RPM that I made which I've integrated into comps.xml, and it >>works fine that way. However, it does start a TCP-based server, which >>itself as a client connects to, to set up some default configuration >>options. >> >>A few days ago, this RPM was installed next to last in Anaconda, and it >>worked fine. However, I have since added some more RPMs to comps.xml, >>and now this particular RPM is being installed right in the middle of >>Anaconda. I believe that this is dictated by using pkgorder. This RPM >>is no longer installing properly during Anaconda, and some of the errors >>that this RPM is giving upon install, indicate that something regarding >>the TCP/IP subsystem is either not installed yet, or not configured >>properly, as if waiting for another RPM to install later via ANaconda >>and set up the since "broken" service. >> >>I guess what I'm asking is, is there any way to specify in which order >>RPMs are installed, inside of Anaconda? I understand that pkgorder does >>it's best to guess to correct order in which to line up packages for >>install with Anaconda, but there's got to be a way to make a few >>exceptions to this, and install the packages in the order that I see fit. >> >>If anyone has any experience with this, or if they can advise me a bit, >>I would greatly appreciate it. >> >>Thanks >>-dant >> > > > Hello - > > I tried placing my RPM at the "end" of my pkginfo file that's used on > the second genhdlist pass. The RPM was installed last, but > unfortunately I ran into the same problem. I suspect now that my > problem is elsewhere. > > Thanks for the time > -dant > Aaaaaaaalright, I think I figured it out here. The only thing that's changed, was the manner in which the kickstart file was loading. And here's the deal: Anaconda now loads the kickstart off of a CD, which is how I want it to work. However, when Anaconda boots off of an HTTP network install, it somehow, someway, or somewhat start some kind of network system, devices activate, TCP/IP "works", yada yada. Since my fancy RPM uses a client/server method to establish a certain setup, and booting from CDROM does not load this networking component (?), my RPM fails to install properly. How might one go about fully enabling network functionality in the manner that I am looking for, during a cd-based kickstart installation? If I can get this answered, I'm set. Thanks! -dant From info at hostinthebox.net Thu Jul 14 00:51:56 2005 From: info at hostinthebox.net (Dan Trainor) Date: Wed, 13 Jul 2005 17:51:56 -0700 Subject: Installing an RPM later on in pkgorder In-Reply-To: <42D5ADD0.8070502@hostinthebox.net> References: <42D595D0.2020307@hostinthebox.net> <42D5989A.5090000@hostinthebox.net> <42D5ADD0.8070502@hostinthebox.net> Message-ID: <42D5B72C.4030002@hostinthebox.net> Dan Trainor wrote: > Dan Trainor wrote: > >>Dan Trainor wrote: >> >> >>>Hello, all - >>> >>>I've got another one for ya. >>> >>>I have an RPM that I made which I've integrated into comps.xml, and it >>>works fine that way. However, it does start a TCP-based server, which >>>itself as a client connects to, to set up some default configuration >>>options. >>> >>>A few days ago, this RPM was installed next to last in Anaconda, and it >>>worked fine. However, I have since added some more RPMs to comps.xml, >>>and now this particular RPM is being installed right in the middle of >>>Anaconda. I believe that this is dictated by using pkgorder. This RPM >>>is no longer installing properly during Anaconda, and some of the errors >>>that this RPM is giving upon install, indicate that something regarding >>>the TCP/IP subsystem is either not installed yet, or not configured >>>properly, as if waiting for another RPM to install later via ANaconda >>>and set up the since "broken" service. >>> >>>I guess what I'm asking is, is there any way to specify in which order >>>RPMs are installed, inside of Anaconda? I understand that pkgorder does >>>it's best to guess to correct order in which to line up packages for >>>install with Anaconda, but there's got to be a way to make a few >>>exceptions to this, and install the packages in the order that I see fit. >>> >>>If anyone has any experience with this, or if they can advise me a bit, >>>I would greatly appreciate it. >>> >>>Thanks >>>-dant >>> >> >> >>Hello - >> >>I tried placing my RPM at the "end" of my pkginfo file that's used on >>the second genhdlist pass. The RPM was installed last, but >>unfortunately I ran into the same problem. I suspect now that my >>problem is elsewhere. >> >>Thanks for the time >>-dant >> > > > Aaaaaaaalright, I think I figured it out here. > > The only thing that's changed, was the manner in which the kickstart > file was loading. And here's the deal: > > Anaconda now loads the kickstart off of a CD, which is how I want it to > work. However, when Anaconda boots off of an HTTP network install, it > somehow, someway, or somewhat start some kind of network system, devices > activate, TCP/IP "works", yada yada. > > Since my fancy RPM uses a client/server method to establish a certain > setup, and booting from CDROM does not load this networking component > (?), my RPM fails to install properly. > > How might one go about fully enabling network functionality in the > manner that I am looking for, during a cd-based kickstart installation? > If I can get this answered, I'm set. > > Thanks! > -dant > Hey, one more update for you all - It is indeed a network issue of some sort. I set the IP addressing to be of type DHCP, and I did not run into my problem. I'm hoping someone can answer my question here, as this would be an incredible help. Thanks -dant From mats.ohlander at saabsystems.se Thu Jul 14 04:01:18 2005 From: mats.ohlander at saabsystems.se (mats.ohlander at saabsystems.se) Date: Thu, 14 Jul 2005 06:01:18 +0200 Subject: Mats Ohlander/SAAB Systems =?iso-8859-1?q?=E4r_p=E5_semester?= Message-ID: I will be out of the office starting 2005-07-14 and will not return until 2005-08-15. Har semester. Tillbaka 15 Augusti. N?s p? telefon 073-437 50 96. Ha en trevlig sommar. From error27 at gmail.com Thu Jul 14 07:14:25 2005 From: error27 at gmail.com (Dan Carpenter) Date: Thu, 14 Jul 2005 09:14:25 +0200 Subject: Installing an RPM later on in pkgorder In-Reply-To: <42D5B72C.4030002@hostinthebox.net> References: <42D595D0.2020307@hostinthebox.net> <42D5989A.5090000@hostinthebox.net> <42D5ADD0.8070502@hostinthebox.net> <42D5B72C.4030002@hostinthebox.net> Message-ID: On 7/14/05, Dan Trainor wrote: > Hey, one more update for you all - > > It is indeed a network issue of some sort. I set the IP addressing to > be of type DHCP, and I did not run into my problem. > > I'm hoping someone can answer my question here, as this would be an > incredible help. > > Thanks > -dant That's far to vague... AND YOU KNOW IT. Based on the fact that you have said it works with dhcp and not with a static IP address my complete guess is that it's because of the `hostname`. Unfortunately, I have no idea what hostname anaconda uses with a dhcp or when you set a static IP. But I like to play guessing games. :) regards, dan carpenter From info at hostinthebox.net Thu Jul 14 08:06:20 2005 From: info at hostinthebox.net (Dan Trainor) Date: Thu, 14 Jul 2005 01:06:20 -0700 Subject: Installing an RPM later on in pkgorder In-Reply-To: References: <42D595D0.2020307@hostinthebox.net> <42D5989A.5090000@hostinthebox.net> <42D5ADD0.8070502@hostinthebox.net> <42D5B72C.4030002@hostinthebox.net> Message-ID: <42D61CFC.40507@hostinthebox.net> Dan Carpenter wrote: > On 7/14/05, Dan Trainor wrote: > >>Hey, one more update for you all - >> >>It is indeed a network issue of some sort. I set the IP addressing to >>be of type DHCP, and I did not run into my problem. >> >>I'm hoping someone can answer my question here, as this would be an >>incredible help. >> >>Thanks >>-dant > > > That's far to vague... AND YOU KNOW IT. > > Based on the fact that you have said it works with dhcp and not with a > static IP address my complete guess is that it's because of the > `hostname`. Unfortunately, I have no idea what hostname anaconda uses > with a dhcp or when you set a static IP. But I like to play guessing > games. :) > > regards, > dan carpenter > > Vague, perhaps, if you didn't read the rest of the thread ;) This really bums me out. Like I said, I imagine that there's some networking component that I'm missing here, causing this operation to fail. I'll hack at it some more tomorrow. Thanks! -dant From ian at smallworld.cx Thu Jul 14 10:34:28 2005 From: ian at smallworld.cx (Ian Leonard) Date: Thu, 14 Jul 2005 11:34:28 +0100 Subject: Create Custom Boot CD. In-Reply-To: <20050711182648.85458.qmail@web52107.mail.yahoo.com> References: <20050711182648.85458.qmail@web52107.mail.yahoo.com> Message-ID: <42D63FB4.1040402@smallworld.cx> Alexander Rau wrote: > A while back a couple of guys and I started the > "Anaconda Documentation project" > > http://rau.homedns.org/twiki/bin/view/Anaconda/WebHome > > It is a bit outdated but it might be a good starting > point and you are more than welcome to make updates as > you go along.... A useful page, thanks. I tried to login to make a comment but was unable to do so. pkgorder failed with some meaningless error. A quick look revealed that it doesn't like any non rpm files in the RPMS directory. The solution was to do something along the lines of: find $BASEDIR -name TRANS.TBL -exec rm -f {} \; -- Ian Leonard Please ignore spelling and punctuation - I did. From jgao at optonline.net Thu Jul 14 13:25:11 2005 From: jgao at optonline.net (Jian Gao) Date: Thu, 14 Jul 2005 09:25:11 -0400 Subject: FC4 Kickstart Message-ID: <0IJM006L2DAEV8QG@mta7.srv.hcvlny.cv.net> I need help on FC4 KIckstart with updated rpms. I tried to include the most recent updated rpms in the Kickstart, but I can't find headers directory in Fedora Core 4 updates. There're only repodata directory and xml files. What I used to do for FC2 is merging the header files of updated rpms into base line headers directory, generate new list file, and Kickstart works well installing new rpms. How could I merge the them together for FC4? Or is there different way to handle the situation? Thanks, Jian -------------- next part -------------- An HTML attachment was scrubbed... URL: From vanallsburg at hope.edu Thu Jul 14 14:09:09 2005 From: vanallsburg at hope.edu (Paul VanAllsburg) Date: Thu, 14 Jul 2005 10:09:09 -0400 Subject: NFS using eth1 installation Message-ID: <42D81333@hope.edu> Thanks Klaus, It works now... Booted to DVD, ran ks.cfg and iso sourced on eth1, I'll set this up with PXE next... for the record, boot: linux ks=nfs:192.168.2.100:/data/ks/fc3/i386/computenode/ks.cfg ksdevice=eth1 and ks.cfg had: nfs --server=192.168.2.100 --dir=/data/iso/fc3/i386 and, network --device eth0 --bootproto dhcp # 192.168.1.101 network --device eth1 --bootproto dhcp # 192.168.2.101 Paul >===== Original Message From Klaus Steden ===== >Paul, > >The kickstart option you're looking for is 'ksdevice=', i.e. ksdevice=eth1 >should do the trick. > >Klaus > >> How do I create a kickstart file that will do a NFS using eth1 installation? >> >> I can successfully do a manual boot: askmethod, specify NFS with eth1, but it >> does not show up in the anaconda-ks.cfg file. I end up with: >> nfs --server=192.168.2.100 --dir=/data/iso/fc3/i386 >> >> When I run a automated install using the anaconda generated script I get 'That >> directory could not be mounted from the server'. I've been able to move the >> NFS share over to a system on eth0 and get it to work but not one on eth1. >> The network/share on eth1 is functional... >> >> Thanks, >> Paul Van Allsburg >> >> _______________________________________________ >> Kickstart-list mailing list >> Kickstart-list at redhat.com >> https://www.redhat.com/mailman/listinfo/kickstart-list > >-- >Klaus Steden | >Senior Systems Administrator | >Technicolor Creative Services | TODO: >Toronto | 1) Learn to use my new Unix account. >klaus.steden at thomson.net | 2) Learn how to change this list. >Phone: (416) 585-9995 | >Fax: (416) 364-1585 | > >_______________________________________________ >Kickstart-list mailing list >Kickstart-list at redhat.com >https://www.redhat.com/mailman/listinfo/kickstart-list From mats.ohlander at saabsystems.se Thu Jul 14 17:07:58 2005 From: mats.ohlander at saabsystems.se (mats.ohlander at saabsystems.se) Date: Thu, 14 Jul 2005 19:07:58 +0200 Subject: Mats Ohlander/SAAB Systems =?iso-8859-1?q?=E4r_p=E5_semester?= Message-ID: I will be out of the office starting 2005-07-14 and will not return until 2005-08-15. Har semester. Tillbaka 15 Augusti. N?s p? telefon 073-437 50 96. Ha en trevlig sommar. From phr at doc.ic.ac.uk Thu Jul 14 18:11:44 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Thu, 14 Jul 2005 19:11:44 +0100 (BST) Subject: probing a system for drives in the pre phase In-Reply-To: References: <20050713185333.22994.qmail@web40914.mail.yahoo.com> Message-ID: On Wed, 13 Jul 2005, Dan Carpenter wrote: >I haven't been able to figure out how to get access to the console for >a user friendly UI... I do everything in a different console and >that's not as good. It's not really a problem on a VGA console, as chvt(1) can bring any VT to the "foreground". Not sure what would happen with serial. For ananconda-like interaction, I'd suggest scripting with dialog. This is all assuming text-mode interaction, which is all that's available in %pre (before the second stage loads). Cheers, Phil From ian at smallworld.cx Thu Jul 14 18:53:51 2005 From: ian at smallworld.cx (Ian Leonard) Date: Thu, 14 Jul 2005 19:53:51 +0100 Subject: Failure to run %post in rpms Message-ID: <42D6B4BF.3070608@smallworld.cx> Hi, I have produced a single FC4 disk. It appears to install okay but checking the install.log I see it has failed to run the %post scripts on several rpms. The reason is that the script couldn't find utilities like cp and grep. I tried putting in absolute paths to the files but it still doesn't work. I don't know when these are installed. Any suggestions on what to do next would be appreciated. Thanks. -- Ian Leonard Please ignore spelling and punctuation - I did. From error27 at gmail.com Thu Jul 14 19:25:07 2005 From: error27 at gmail.com (Dan Carpenter) Date: Thu, 14 Jul 2005 12:25:07 -0700 Subject: Failure to run %post in rpms In-Reply-To: <42D6B4BF.3070608@smallworld.cx> References: <42D6B4BF.3070608@smallworld.cx> Message-ID: %post is run in a chroot on the system. What did the error messages look like? thanks, dan On 7/14/05, Ian Leonard wrote: > Hi, > > I have produced a single FC4 disk. It appears to install okay but > checking the install.log I see it has failed to run the %post scripts on > several rpms. The reason is that the script couldn't find utilities like > cp and grep. I tried putting in absolute paths to the files but it still > doesn't work. > > I don't know when these are installed. Any suggestions on what to do > next would be appreciated. Thanks. > From info at hostinthebox.net Thu Jul 14 23:05:32 2005 From: info at hostinthebox.net (Dan Trainor) Date: Thu, 14 Jul 2005 16:05:32 -0700 Subject: Dan you're so awesome, yes you are! - WAS: Installing an RPM later on in pkgorder In-Reply-To: <42D5B72C.4030002@hostinthebox.net> References: <42D595D0.2020307@hostinthebox.net> <42D5989A.5090000@hostinthebox.net> <42D5ADD0.8070502@hostinthebox.net> <42D5B72C.4030002@hostinthebox.net> Message-ID: <42D6EFBC.1060800@hostinthebox.net> Dan Trainor wrote: > Dan Trainor wrote: > >>Dan Trainor wrote: >> >> >>>Dan Trainor wrote: >>> >>> >>> >>>>Hello, all - >>>> >>>>I've got another one for ya. >>>> >>>>I have an RPM that I made which I've integrated into comps.xml, and it >>>>works fine that way. However, it does start a TCP-based server, which >>>>itself as a client connects to, to set up some default configuration >>>>options. >>>> >>>>A few days ago, this RPM was installed next to last in Anaconda, and it >>>>worked fine. However, I have since added some more RPMs to comps.xml, >>>>and now this particular RPM is being installed right in the middle of >>>>Anaconda. I believe that this is dictated by using pkgorder. This RPM >>>>is no longer installing properly during Anaconda, and some of the errors >>>>that this RPM is giving upon install, indicate that something regarding >>>>the TCP/IP subsystem is either not installed yet, or not configured >>>>properly, as if waiting for another RPM to install later via ANaconda >>>>and set up the since "broken" service. >>>> >>>>I guess what I'm asking is, is there any way to specify in which order >>>>RPMs are installed, inside of Anaconda? I understand that pkgorder does >>>>it's best to guess to correct order in which to line up packages for >>>>install with Anaconda, but there's got to be a way to make a few >>>>exceptions to this, and install the packages in the order that I see fit. >>>> >>>>If anyone has any experience with this, or if they can advise me a bit, >>>>I would greatly appreciate it. >>>> >>>>Thanks >>>>-dant >>>> >>> >>> >>>Hello - >>> >>>I tried placing my RPM at the "end" of my pkginfo file that's used on >>>the second genhdlist pass. The RPM was installed last, but >>>unfortunately I ran into the same problem. I suspect now that my >>>problem is elsewhere. >>> >>>Thanks for the time >>>-dant >>> >> >> >>Aaaaaaaalright, I think I figured it out here. >> >>The only thing that's changed, was the manner in which the kickstart >>file was loading. And here's the deal: >> >>Anaconda now loads the kickstart off of a CD, which is how I want it to >>work. However, when Anaconda boots off of an HTTP network install, it >>somehow, someway, or somewhat start some kind of network system, devices >>activate, TCP/IP "works", yada yada. >> >>Since my fancy RPM uses a client/server method to establish a certain >>setup, and booting from CDROM does not load this networking component >>(?), my RPM fails to install properly. >> >>How might one go about fully enabling network functionality in the >>manner that I am looking for, during a cd-based kickstart installation? >> If I can get this answered, I'm set. >> >>Thanks! >>-dant >> > > > Hey, one more update for you all - > > It is indeed a network issue of some sort. I set the IP addressing to > be of type DHCP, and I did not run into my problem. > > I'm hoping someone can answer my question here, as this would be an > incredible help. > > Thanks > -dant > Alright, well, I just wanted to update you guys again here. No network adapters were up, which was causing my problems. THe solution was to bring up lo in %pre, and stuff started to work. The only reason why I replied was in the hopes that others might learn from this in the future, if they run into this problem, as well. With that being said, I am now complete with my custom RHEL4.0 Kickstart install, including custom menus, comps.xml, base/hdlist{2} files, and so on. I will be documenting all of this shortly, as I would hate to have others go through the same amount of work that I've been through over the past month. Thanks -dant From ashley.hodder at otago.ac.nz Fri Jul 15 01:53:42 2005 From: ashley.hodder at otago.ac.nz (Ashley Hodder) Date: Fri, 15 Jul 2005 13:53:42 +1200 Subject: to the moderator please remove me from the list Message-ID: <1121392422.1919.302.camel@malibu.otago.ac.nz> HI have forgotten my actual username and are about to leave this employer, so would like to remove myself from the kickstart discussions group before the mail starts bouncing back at the list. Thanks Ashley hodder From ian at smallworld.cx Fri Jul 15 11:21:29 2005 From: ian at smallworld.cx (Ian Leonard) Date: Fri, 15 Jul 2005 12:21:29 +0100 Subject: Failure to run %post in rpms In-Reply-To: References: <42D6B4BF.3070608@smallworld.cx> Message-ID: <42D79C39.90500@smallworld.cx> Dan Carpenter wrote: > %post is run in a chroot on the system. Just to be sure, this is the %post script in the rpm, not the kickstart file. > > What did the error messages look like? I see lines like this: /var/tmp/rpm-tmp.34500:line 4: /bin/cp: No such file or directory /var/tmp/rpm-tmp.34500:line 18: /bin/cat: No such file or directory error: %post(microplay-1.0.0.31.i386) scriplet failed, exit 127 Thanks. > thanks, > dan > > On 7/14/05, Ian Leonard wrote: > >>Hi, >> >>I have produced a single FC4 disk. It appears to install okay but >>checking the install.log I see it has failed to run the %post scripts on >>several rpms. The reason is that the script couldn't find utilities like >>cp and grep. I tried putting in absolute paths to the files but it still >>doesn't work. >> >>I don't know when these are installed. Any suggestions on what to do >>next would be appreciated. Thanks. >> > > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list -- Ian Leonard Please ignore spelling and punctuation - I did. From phr at doc.ic.ac.uk Fri Jul 15 13:34:49 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Fri, 15 Jul 2005 14:34:49 +0100 (BST) Subject: Failure to run %post in rpms In-Reply-To: <42D79C39.90500@smallworld.cx> References: <42D6B4BF.3070608@smallworld.cx> <42D79C39.90500@smallworld.cx> Message-ID: On Fri, 15 Jul 2005, Ian Leonard wrote: >Dan Carpenter wrote: >> %post is run in a chroot on the system. > >Just to be sure, this is the %post script in the rpm, not the kickstart >file. > >> >> What did the error messages look like? > >I see lines like this: > >/var/tmp/rpm-tmp.34500:line 4: /bin/cp: No such file or directory >/var/tmp/rpm-tmp.34500:line 18: /bin/cat: No such file or directory >error: %post(microplay-1.0.0.31.i386) scriplet failed, exit 127 Much badness; does your package list those as dependencies? (or just coreutils as a whole?) I really ought to check up on the current status of this, but there is/was a bug in anaconda which didn't realise when package orders were messed up, and the workaround was to call "genhdlist; pkgorder; genhdlist" with some flags. How did you build the custom CD (genhdlist specifically)? Cheers, Phil From mats.ohlander at saabsystems.se Fri Jul 15 17:07:56 2005 From: mats.ohlander at saabsystems.se (mats.ohlander at saabsystems.se) Date: Fri, 15 Jul 2005 19:07:56 +0200 Subject: Mats Ohlander/SAAB Systems =?iso-8859-1?q?=E4r_p=E5_semester?= Message-ID: I will be out of the office starting 2005-07-14 and will not return until 2005-08-15. Har semester. Tillbaka 15 Augusti. N?s p? telefon 073-437 50 96. Ha en trevlig sommar. From kernel at linuxace.com Fri Jul 15 17:12:43 2005 From: kernel at linuxace.com (Phil Oester) Date: Fri, 15 Jul 2005 10:12:43 -0700 Subject: Odd partition labeling Message-ID: <20050715171243.GA17940@linuxace.com> Working on a FC4 kickstart, and seeing some odd partition labels which have a '1' appended. For example, on my first kickstart I see: / = /1 var = /var1 usr = /usr1 etc... I then re-kick, and end up with proper labels for all but boot, which ends up as /boot1. I then re-kick again, and boot is /boot, but now all the other partition labels end with a 1. Any ideas why this is happening? Phil From klaus.steden at thomson.net Fri Jul 15 17:37:32 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Fri, 15 Jul 2005 13:37:32 -0400 Subject: Odd partition labeling In-Reply-To: <20050715171243.GA17940@linuxace.com> References: <20050715171243.GA17940@linuxace.com> Message-ID: <20050715173732.GZ73653@thomson.net> > Working on a FC4 kickstart, and seeing some odd partition labels which > have a '1' appended. For example, on my first kickstart I see: > > / = /1 > var = /var1 > usr = /usr1 > etc... > > I then re-kick, and end up with proper labels for all but boot, > which ends up as /boot1. I then re-kick again, and boot is /boot, > but now all the other partition labels end with a 1. > > Any ideas why this is happening? > Hrm, Me too. I thought it was due to a bug in something I was doing, but if others are seeing it ... I've also seen "/12", although not as common as "/" or "/1". Klaus From gressett at noao.edu Fri Jul 15 17:44:13 2005 From: gressett at noao.edu (Katrina Gressett) Date: Fri, 15 Jul 2005 10:44:13 -0700 (MST) Subject: Odd partition labeling In-Reply-To: <20050715173732.GZ73653@thomson.net> Message-ID: I've had the same problem with FC3 and my root partitions. Instead of / which I prefer, I get /1. It doesn't seem to make a difference since it is just the partion label and every thing mounts the way you expect, but I don't like it. You could try a little post script to check and then change the disk label with e2label. Just make sure that you also change the labels in your boot loader and fstab. Katrina On Fri, 15 Jul 2005, Klaus Steden wrote: > > Working on a FC4 kickstart, and seeing some odd partition labels which > > have a '1' appended. For example, on my first kickstart I see: > > > > / = /1 > > var = /var1 > > usr = /usr1 > > etc... > > > > I then re-kick, and end up with proper labels for all but boot, > > which ends up as /boot1. I then re-kick again, and boot is /boot, > > but now all the other partition labels end with a 1. > > > > Any ideas why this is happening? > > > Hrm, > > Me too. I thought it was due to a bug in something I was doing, but if others > are seeing it ... I've also seen "/12", although not as common as "/" or "/1". > > Klaus > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > -- -- ___________________________________________________________ Katrina Gressett NSO/GONG -- Data Reduction Specialist gressett at noao.edu 520.318.8568 ___________________________________________________________ From adkinss at ohio.edu Fri Jul 15 17:50:11 2005 From: adkinss at ohio.edu (Scott Adkins) Date: Fri, 15 Jul 2005 13:50:11 -0400 Subject: kickstart and network configuration Message-ID: <9FBBB8160DFFD8B31908296A@Callisto> I am having an issue with using a single kickstart for two different models of similar hardware (IBM Blade Center model 8678 and model 8832). Apparently, the newer model flips its concept of what is the primary ethernet, causing eth1 to come up instead of eth0. Since both interfaces actually have link state on them, I can't just "not specify" and expect the right one to be chosen. So, I am trying to find a better way to adapt to the hardware. First off, the line "network --bootproto dhcp --device eth1" is required to get the newer blades to boot. However, it will break the older blades. I wondered if there was a way to not specify the network and maybe set it up in the %PRE section or something. When exactly does the %PRE section get executed anyways? Does it get executed with every reboot until after the packages are installed? Is it only once? What I am looking for is a more generic way to set up the network to the way I want it so that the kickstart will work. What I do know is the serial numbers of the old blades (we only have a handful of them), so I could have the network set up one way by default, and a different way if one of the serial numbers show up... I have access to the serial number in the kickstart (already verified that). Thoughts? Thanks! Scott -- +-----------------------------------------------------------------------+ Scott W. Adkins http://www.cns.ohiou.edu/~sadkins/ UNIX Systems Engineer mailto:adkinss at ohio.edu ICQ 7626282 Work (740)593-9478 Fax (740)593-1944 +-----------------------------------------------------------------------+ PGP Public Key available at http://www.cns.ohiou.edu/~sadkins/pgp/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 231 bytes Desc: not available URL: From kernel at linuxace.com Fri Jul 15 17:55:27 2005 From: kernel at linuxace.com (Phil Oester) Date: Fri, 15 Jul 2005 10:55:27 -0700 Subject: kickstart and network configuration In-Reply-To: <9FBBB8160DFFD8B31908296A@Callisto> References: <9FBBB8160DFFD8B31908296A@Callisto> Message-ID: <20050715175527.GA18022@linuxace.com> On Fri, Jul 15, 2005 at 01:50:11PM -0400, Scott Adkins wrote: > I am having an issue with using a single kickstart for two different models > of similar hardware (IBM Blade Center model 8678 and model 8832). > Apparently, > the newer model flips its concept of what is the primary ethernet, causing > eth1 to come up instead of eth0. Since both interfaces actually have link > state on them, I can't just "not specify" and expect the right one to be > chosen. So, I am trying to find a better way to adapt to the hardware. > > First off, the line "network --bootproto dhcp --device eth1" is required to > get the newer blades to boot. However, it will break the older blades. How about something like this in PRE: if [ -n "dmidecode | grep 8678" ] ; then echo "network --bootproto dhcp --device eth1" > /tmp/ksnet else echo "network --bootproto dhcp --device eth0" > /tmp/ksnet fi Then %include /tmp/ksnet where you would normally put the network line. Since you mention serial numbers, you could also use that in your tests. Phil From kernel at linuxace.com Fri Jul 15 17:59:02 2005 From: kernel at linuxace.com (Phil Oester) Date: Fri, 15 Jul 2005 10:59:02 -0700 Subject: kickstart and network configuration In-Reply-To: <20050715175527.GA18022@linuxace.com> References: <9FBBB8160DFFD8B31908296A@Callisto> <20050715175527.GA18022@linuxace.com> Message-ID: <20050715175902.GB18022@linuxace.com> On Fri, Jul 15, 2005 at 10:55:27AM -0700, Phil Oester wrote: > if [ -n "dmidecode | grep 8678" ] ; then make that if [ -n "`dmidecode | grep 8678`" ] ; then From adkinss at ohio.edu Fri Jul 15 17:59:48 2005 From: adkinss at ohio.edu (Scott Adkins) Date: Fri, 15 Jul 2005 13:59:48 -0400 Subject: kickstart and network configuration In-Reply-To: <20050715175527.GA18022@linuxace.com> References: <9FBBB8160DFFD8B31908296A@Callisto> <20050715175527.GA18022@linuxace.com> Message-ID: --On Friday, July 15, 2005 10:55 AM -0700 Phil Oester wrote: > On Fri, Jul 15, 2005 at 01:50:11PM -0400, Scott Adkins wrote: >> I am having an issue with using a single kickstart for two different >> models of similar hardware (IBM Blade Center model 8678 and model 8832). >> Apparently, >> the newer model flips its concept of what is the primary ethernet, >> causing eth1 to come up instead of eth0. Since both interfaces actually >> have link state on them, I can't just "not specify" and expect the right >> one to be chosen. So, I am trying to find a better way to adapt to the >> hardware. >> >> First off, the line "network --bootproto dhcp --device eth1" is required >> to get the newer blades to boot. However, it will break the older >> blades. > > How about something like this in PRE: > > if [ -n "dmidecode | grep 8678" ] ; then > echo "network --bootproto dhcp --device eth1" > /tmp/ksnet > else > echo "network --bootproto dhcp --device eth0" > /tmp/ksnet > fi > > Then %include /tmp/ksnet where you would normally put the network > line. Since you mention serial numbers, you could also use > that in your tests. > > Phil Wow! I have been looking for all kinds of commands that would give the model type, and you came up with one right away! :) I will give that a try, and I won't need to do anything with serial numbers, hopefully :) So, I am still confused about %PRE. Does it execute after every reboot? Or does it execute only once, after something specific occurs? Thanks! Scott -- +-----------------------------------------------------------------------+ Scott W. Adkins http://www.cns.ohiou.edu/~sadkins/ UNIX Systems Engineer mailto:adkinss at ohio.edu ICQ 7626282 Work (740)593-9478 Fax (740)593-1944 +-----------------------------------------------------------------------+ PGP Public Key available at http://www.cns.ohiou.edu/~sadkins/pgp/ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 231 bytes Desc: not available URL: From klaus.steden at thomson.net Fri Jul 15 18:01:05 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Fri, 15 Jul 2005 14:01:05 -0400 Subject: kickstart and network configuration In-Reply-To: <9FBBB8160DFFD8B31908296A@Callisto> References: <9FBBB8160DFFD8B31908296A@Callisto> Message-ID: <20050715180105.GA73653@thomson.net> > I am having an issue with using a single kickstart for two different models > of similar hardware (IBM Blade Center model 8678 and model 8832). > Apparently, > the newer model flips its concept of what is the primary ethernet, causing > eth1 to come up instead of eth0. Since both interfaces actually have link > state on them, I can't just "not specify" and expect the right one to be > chosen. So, I am trying to find a better way to adapt to the hardware. > > First off, the line "network --bootproto dhcp --device eth1" is required to > get the newer blades to boot. However, it will break the older blades. > > I wondered if there was a way to not specify the network and maybe set it > up in the %PRE section or something. When exactly does the %PRE section > get executed anyways? Does it get executed with every reboot until after > the packages are installed? Is it only once? > > What I am looking for is a more generic way to set up the network to the > way I want it so that the kickstart will work. What I do know is the > serial numbers of the old blades (we only have a handful of them), so I > could have the network set up one way by default, and a different way if > one of the serial numbers show up... I have access to the serial number > in the kickstart (already verified that). > Ah yeah, I've run into that before with the Blade Center. IIRC, there was actually a firmware patch from IBM to fix that, but I am not 100%. I think there is a setting for anaconda to choose the NIC that is linked automagically - but I can't remember the exact parameter setting. Does someone else out there remember, or am I blowing smoke on this one? You could simply specify 'ksdevice=eth' in your boot command line and it will figure out the rest ... to further automate it, adjust your boot media's selection list to have one for the 8667 (ksdevice=eth1) and one for 8832 (ksdevice=eth0). hth, Klaus From kernel at linuxace.com Fri Jul 15 18:02:15 2005 From: kernel at linuxace.com (Phil Oester) Date: Fri, 15 Jul 2005 11:02:15 -0700 Subject: kickstart and network configuration In-Reply-To: References: <9FBBB8160DFFD8B31908296A@Callisto> <20050715175527.GA18022@linuxace.com> Message-ID: <20050715180215.GC18022@linuxace.com> On Fri, Jul 15, 2005 at 01:59:48PM -0400, Scott Adkins wrote: > So, I am still confused about %PRE. Does it execute after every reboot? > Or does it execute only once, after something specific occurs? It runs only at the beginning of a kickstart. Phil From email at jasonkohles.com Fri Jul 15 18:12:55 2005 From: email at jasonkohles.com (Jason Kohles) Date: Fri, 15 Jul 2005 14:12:55 -0400 Subject: kickstart and network configuration In-Reply-To: <20050715180105.GA73653@thomson.net> References: <9FBBB8160DFFD8B31908296A@Callisto> <20050715180105.GA73653@thomson.net> Message-ID: <20050715181255.GB23887@mail.jasonkohles.com> On Fri, Jul 15, 2005 at 02:01:05PM -0400, Klaus Steden wrote: > > I think there is a setting for anaconda to choose the NIC that is linked > automagically - but I can't remember the exact parameter setting. Does someone > else out there remember, or am I blowing smoke on this one? > ksdevice=link, although the original poster specified that part of the problem was that both interface had link so he had to specify. -- Jason Kohles A witty saying proves nothing. email at jasonkohles.com -- Voltaire (1694 - 1778) http://www.jasonkohles.com/ From kernel at linuxace.com Fri Jul 15 18:14:10 2005 From: kernel at linuxace.com (Phil Oester) Date: Fri, 15 Jul 2005 11:14:10 -0700 Subject: Odd partition labeling In-Reply-To: References: <20050715173732.GZ73653@thomson.net> Message-ID: <20050715181410.GD18022@linuxace.com> On Fri, Jul 15, 2005 at 01:37:32PM -0400, Klaus Steden wrote: > Me too. I thought it was due to a bug in something I was doing, but if others > are seeing it ... I've also seen "/12", although not as common as "/" or "/1". > > Klaus On Fri, Jul 15, 2005 at 10:44:13AM -0700, Katrina Gressett wrote: > I've had the same problem with FC3 and my root partitions. Instead of / > which I prefer, I get /1. It doesn't seem to make a difference since it > is just the partion label and every thing mounts the way you expect, > but I don't like it. You could try a little post script to check and then > change the disk label with e2label. Just make sure that you also change > the labels in your boot loader and fstab. > > Katrina Well, at least this confirms that this is probably a bug as opposed to something I'm screwing up. (BTW - this worked fine in FC1) What seems to be happening is that anaconda sees an existing label with the same name, so it appends a 1 (see fsset.py). Since I mount the old root in my PRE, think I'll now add the additional step of changing the labels so they won't conflict with the new ones anaconda wants to use. Something like the below: mkdir -p /mnt/oldroot for dev in `fdisk -l | grep ^/dev | grep Linux$ | cut -d\ -f 1` ; do part=`e2label $dev` if [ "$part" == "/" ] ; then mount $dev /mnt/oldroot e2label $dev old_/ else e2label $dev old_`e2label $dev` fi done Phil From klaus.steden at thomson.net Fri Jul 15 18:21:10 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Fri, 15 Jul 2005 14:21:10 -0400 Subject: kickstart and network configuration In-Reply-To: <20050715181255.GB23887@mail.jasonkohles.com> References: <9FBBB8160DFFD8B31908296A@Callisto> <20050715180105.GA73653@thomson.net> <20050715181255.GB23887@mail.jasonkohles.com> Message-ID: <20050715182110.GB73653@thomson.net> > ksdevice=link, although the original poster specified that part of the > problem was that both interface had link so he had to specify. > D'oh. Reading was never my strong suit. ;> Klaus From klaus.steden at thomson.net Fri Jul 15 18:29:20 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Fri, 15 Jul 2005 14:29:20 -0400 Subject: Odd partition labeling In-Reply-To: References: <20050715173732.GZ73653@thomson.net> Message-ID: <20050715182920.GC73653@thomson.net> > I've had the same problem with FC3 and my root partitions. Instead of / > which I prefer, I get /1. It doesn't seem to make a difference since it > is just the partion label and every thing mounts the way you expect, > but I don't like it. You could try a little post script to check and then > change the disk label with e2label. Just make sure that you also change > the labels in your boot loader and fstab. > Hrm, that's good to know ... but where does that original labelling get done? It would be easier to change it immediately after it's set incorrectly - i.e. before the bootloader and /etc/fstab configurations are written ... otherwise, that's a lot of dodgy post-scripting, imo. Any ideas? Klaus From ian at smallworld.cx Fri Jul 15 19:00:16 2005 From: ian at smallworld.cx (Ian Leonard) Date: Fri, 15 Jul 2005 20:00:16 +0100 Subject: Failure to run %post in rpms In-Reply-To: References: <42D6B4BF.3070608@smallworld.cx> <42D79C39.90500@smallworld.cx> Message-ID: <42D807C0.9090207@smallworld.cx> Philip Rowlands wrote: > On Fri, 15 Jul 2005, Ian Leonard wrote: > > >>Dan Carpenter wrote: >> >>>%post is run in a chroot on the system. >> >>Just to be sure, this is the %post script in the rpm, not the kickstart >>file. >> >> >>>What did the error messages look like? >> >>I see lines like this: >> >>/var/tmp/rpm-tmp.34500:line 4: /bin/cp: No such file or directory >>/var/tmp/rpm-tmp.34500:line 18: /bin/cat: No such file or directory >>error: %post(microplay-1.0.0.31.i386) scriplet failed, exit 127 > > > Much badness; does your package list those as dependencies? (or just > coreutils as a whole?) My RPM does not list anything as dependencies. This may be wrong but it is not the only one to fail, a few more do. > I really ought to check up on the current status of this, but there > is/was a bug in anaconda which didn't realise when package orders were > messed up, and the workaround was to call "genhdlist; pkgorder; > genhdlist" with some flags. How did you build the custom CD (genhdlist > specifically)? A big question. I am trying to make a general purpose script for handle several cd builds. I have a list of files which my script reads, deletes from the comps.xml file and deletes the rpms. This leaves me with a distribution small enough for a single cd. I also add in my own rpm. The I run: /usr/lib/anaconda-runtime/genhdlist --productpath Fedora $distro/$arch Next: /usr/lib/anaconda-runtime/getfullcomps.py comps.xml $distro i386 >/tmp/comps.out I grep this for the word CRITICAL and bomb out if there is a problem. If all is well: /usr/lib/anaconda-runtime/pkgorder $distro i386 i386/Fedora > $distro/pkgorder.txt and /usr/lib/anaconda-runtime/genhdlist --withnumbers --fileorder $distro/pkgorder.txt --productpath Fedora $distro That should be the main distribution made, so the next step is to split it up: /usr/lib/anaconda-runtime/splittree.py --arch=i386 \ --total-discs=$max_disks --bin-discs=$discs --src-discs=$source_disks \ --release-string="$product" \ --pkgorderfile=$distro/pkgorder.txt \ --distdir=$distro/i386 \ --srcdir=$distro/SRPMS --productpath=Fedora and finally run genhdlist on the tree for each disc: /usr/lib/anaconda-runtime/genhdlist --withnumbers --productpath Fedora $distro/i386-disc1 And that's it. I used kickstart with a RH 7.3 distribution several years ago but basically I am new to this. Please let me know if I have got any of it wrong. Thanks. -- Ian Leonard Please ignore spelling and punctuation - I did. From mshuler at rackspace.com Fri Jul 15 19:17:49 2005 From: mshuler at rackspace.com (Michael Shuler) Date: Fri, 15 Jul 2005 14:17:49 -0500 Subject: Odd partition labeling In-Reply-To: <20050715182920.GC73653@thomson.net> References: <20050715173732.GZ73653@thomson.net> <20050715182920.GC73653@thomson.net> Message-ID: <42D80BDD.5080600@rackspace.com> Klaus Steden wrote: > Hrm, that's good to know ... but where does that original labelling get done? On a previous partitioning run. > It would be easier to change it immediately after it's set incorrectly - i.e. > before the bootloader and /etc/fstab configurations are written ... otherwise, > that's a lot of dodgy post-scripting, imo. Agreed - get rid of LABEL= in fstab, altogether: https://www.redhat.com/archives/kickstart-list/2005-March/msg00096.html Kind Regards, Michael From kernel at linuxace.com Fri Jul 15 19:41:21 2005 From: kernel at linuxace.com (Phil Oester) Date: Fri, 15 Jul 2005 12:41:21 -0700 Subject: Odd partition labeling In-Reply-To: <42D80BDD.5080600@rackspace.com> References: <20050715173732.GZ73653@thomson.net> <20050715182920.GC73653@thomson.net> <42D80BDD.5080600@rackspace.com> Message-ID: <20050715194121.GA18452@linuxace.com> On Fri, Jul 15, 2005 at 02:17:49PM -0500, Michael Shuler wrote: > Agreed - get rid of LABEL= in fstab, altogether: > > https://www.redhat.com/archives/kickstart-list/2005-March/msg00096.html This leads to bad things down the line if you add a new drive and it magically becomes sda and pushes current drive to sdb. There are pluses and minuses to each solution... Phil From klaus.steden at thomson.net Fri Jul 15 19:52:07 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Fri, 15 Jul 2005 15:52:07 -0400 Subject: Odd partition labeling In-Reply-To: <20050715194121.GA18452@linuxace.com> References: <20050715173732.GZ73653@thomson.net> <20050715182920.GC73653@thomson.net> <42D80BDD.5080600@rackspace.com> <20050715194121.GA18452@linuxace.com> Message-ID: <20050715195206.GE73653@thomson.net> > > Agreed - get rid of LABEL= in fstab, altogether: > > > > https://www.redhat.com/archives/kickstart-list/2005-March/msg00096.html > > This leads to bad things down the line if you add a new drive > and it magically becomes sda and pushes current drive to sdb. > > There are pluses and minuses to each solution... > Personally, I find them annoying, but you're right, there are advantages to both approaches. Would it make sense to add a switch to the disk partitioning section of anaconda to allow for the use of labels or device names as desired? Seems to me the most reasonable approach. Klaus From ian at smallworld.cx Sat Jul 16 11:18:56 2005 From: ian at smallworld.cx (Ian Leonard) Date: Sat, 16 Jul 2005 12:18:56 +0100 Subject: Failure to run %post in rpms In-Reply-To: <42D807C0.9090207@smallworld.cx> References: <42D6B4BF.3070608@smallworld.cx> <42D79C39.90500@smallworld.cx> <42D807C0.9090207@smallworld.cx> Message-ID: <42D8ED20.2040406@smallworld.cx> Ian Leonard wrote: > Philip Rowlands wrote: > [ snip ] > > That should be the main distribution made, so the next step is to split > it up: > > > /usr/lib/anaconda-runtime/splittree.py --arch=i386 \ > --total-discs=$max_disks --bin-discs=$discs > --src-discs=$source_disks \ > --release-string="$product" \ > --pkgorderfile=$distro/pkgorder.txt \ > --distdir=$distro/i386 \ > --srcdir=$distro/SRPMS --productpath=Fedora > > > and finally run genhdlist on the tree for each disc: > > /usr/lib/anaconda-runtime/genhdlist --withnumbers --productpath Fedora > $distro/i386-disc1 > I found the problem. I wasn't giving the --fileorder option on the last genhdlist. Sorry about that. -- Ian Leonard Please ignore spelling and punctuation - I did. From kimmo.koivisto at surfeu.fi Sat Jul 16 17:14:39 2005 From: kimmo.koivisto at surfeu.fi (Kimmo Koivisto) Date: Sat, 16 Jul 2005 20:14:39 +0300 Subject: Kickstart: correct way to disable SELinux Message-ID: <200507162014.45464.kimmo.koivisto@surfeu.fi> Hello What is the correct way to disable SELinux from kickstart installation? I'm using Fedora Core 4. Regards Kimmo Koivisto From ian at smallworld.cx Sat Jul 16 17:28:06 2005 From: ian at smallworld.cx (Ian Leonard) Date: Sat, 16 Jul 2005 18:28:06 +0100 Subject: Kickstart: correct way to disable SELinux In-Reply-To: <200507162014.45464.kimmo.koivisto@surfeu.fi> References: <200507162014.45464.kimmo.koivisto@surfeu.fi> Message-ID: <42D943A6.7030600@smallworld.cx> Kimmo Koivisto wrote: > Hello > > What is the correct way to disable SELinux from kickstart installation? > I'm using Fedora Core 4. > > Regards > Kimmo Koivisto Try adding this to ks.cfg: selinux --disabled -- Ian Leonard Please ignore spelling and punctuation - I did. From info at hostinthebox.net Sat Jul 16 18:53:24 2005 From: info at hostinthebox.net (Dan Trainor) Date: Sat, 16 Jul 2005 11:53:24 -0700 Subject: Kickstart: correct way to disable SELinux In-Reply-To: <42D943A6.7030600@smallworld.cx> References: <200507162014.45464.kimmo.koivisto@surfeu.fi> <42D943A6.7030600@smallworld.cx> Message-ID: <42D957A4.7020603@hostinthebox.net> Ian Leonard wrote: > Kimmo Koivisto wrote: > >> Hello >> >> What is the correct way to disable SELinux from kickstart installation? >> I'm using Fedora Core 4. >> >> Regards >> Kimmo Koivisto > > > Try adding this to ks.cfg: > > selinux --disabled > > Hi - You know, this never did work for me, I'm not sure why. I've tried this on dozens of kickstarts, with dozens of different versions, to no avail. I add an --append="selinux=0" to my grub setup in kickstart, which does work. Thanks -dant From dan at half-asleep.com Sun Jul 17 01:23:41 2005 From: dan at half-asleep.com (Daniel Segall) Date: Sat, 16 Jul 2005 21:23:41 -0400 Subject: Kickstart: correct way to disable SELinux In-Reply-To: <42D957A4.7020603@hostinthebox.net> References: <200507162014.45464.kimmo.koivisto@surfeu.fi> <42D943A6.7030600@smallworld.cx> <42D957A4.7020603@hostinthebox.net> Message-ID: <42D9B31D.5000103@half-asleep.com> set /etc/sysconfig/selinux to: SELINUX=disabled -Dan Dan Trainor wrote: > Ian Leonard wrote: > >> Kimmo Koivisto wrote: >> >>> Hello >>> >>> What is the correct way to disable SELinux from kickstart installation? >>> I'm using Fedora Core 4. >>> >>> Regards >>> Kimmo Koivisto >> >> >> >> Try adding this to ks.cfg: >> >> selinux --disabled >> >> > > Hi - > > You know, this never did work for me, I'm not sure why. I've tried this > on dozens of kickstarts, with dozens of different versions, to no avail. > > I add an --append="selinux=0" to my grub setup in kickstart, which does > work. > > Thanks > -dant > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > From info at hostinthebox.net Sun Jul 17 03:05:48 2005 From: info at hostinthebox.net (Dan Trainor) Date: Sat, 16 Jul 2005 20:05:48 -0700 Subject: Kickstart: correct way to disable SELinux In-Reply-To: <42D9B31D.5000103@half-asleep.com> References: <200507162014.45464.kimmo.koivisto@surfeu.fi> <42D943A6.7030600@smallworld.cx> <42D957A4.7020603@hostinthebox.net> <42D9B31D.5000103@half-asleep.com> Message-ID: <42D9CB0C.5070704@hostinthebox.net> Daniel Segall wrote: > set /etc/sysconfig/selinux to: > SELINUX=disabled > > -Dan > > Dan Trainor wrote: > >> Ian Leonard wrote: >> >>> Kimmo Koivisto wrote: >>> >>>> Hello >>>> >>>> What is the correct way to disable SELinux from kickstart installation? >>>> I'm using Fedora Core 4. >>>> >>>> Regards >>>> Kimmo Koivisto >>> >>> >>> >>> >>> Try adding this to ks.cfg: >>> >>> selinux --disabled >>> >>> >> >> Hi - >> >> You know, this never did work for me, I'm not sure why. I've tried >> this on dozens of kickstarts, with dozens of different versions, to no >> avail. >> >> I add an --append="selinux=0" to my grub setup in kickstart, which >> does work. >> >> Thanks >> -dant >> Dan - Yeah, that could be done. I'm still disappointed that Anaconda just won't listen to the directives that it's told to do. Ah well. Thanks -dant From phr at doc.ic.ac.uk Sun Jul 17 12:25:27 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Sun, 17 Jul 2005 13:25:27 +0100 (BST) Subject: Failure to run %post in rpms In-Reply-To: <42D807C0.9090207@smallworld.cx> References: <42D6B4BF.3070608@smallworld.cx> <42D79C39.90500@smallworld.cx> <42D807C0.9090207@smallworld.cx> Message-ID: On Fri, 15 Jul 2005, Ian Leonard wrote: >> Much badness; does your package list those as dependencies? (or just >> coreutils as a whole?) > >My RPM does not list anything as dependencies. This may be wrong but it >is not the only one to fail, a few more do. then On Sat, 16 Jul 2005, Ian Leonard wrote: >I found the problem. I wasn't giving the --fileorder option on the last >genhdlist. Sorry about that. Glad it's now fixed. Generally when one sees [RPM] %postinstall scripts failing due to missing binaries, it's because packages have been applied in the wrong order. Cheers, Phil From jkeating at j2solutions.net Sun Jul 17 22:17:48 2005 From: jkeating at j2solutions.net (Jesse Keating) Date: Sun, 17 Jul 2005 15:17:48 -0700 Subject: Kickstart: correct way to disable SELinux In-Reply-To: <42D957A4.7020603@hostinthebox.net> References: <200507162014.45464.kimmo.koivisto@surfeu.fi> <42D943A6.7030600@smallworld.cx> <42D957A4.7020603@hostinthebox.net> Message-ID: <1121638668.2769.54.camel@yoda.loki.me> On Sat, 2005-07-16 at 11:53 -0700, Dan Trainor wrote: > Hi - > > You know, this never did work for me, I'm not sure why. I've tried this > on dozens of kickstarts, with dozens of different versions, to no avail. How does this "not work"? I've done it w/ FC3 and FC4 and it works just fine... Selinux is in disabled mode. -- Jesse Keating RHCE (geek.j2solutions.net) Fedora Legacy Team (www.fedoralegacy.org) GPG Public Key (geek.j2solutions.net/jkeating.j2solutions.pub) Was I helpful? Let others know: http://svcs.affero.net/rm.php?r=jkeating From clumens at redhat.com Mon Jul 18 13:58:32 2005 From: clumens at redhat.com (Chris Lumens) Date: Mon, 18 Jul 2005 09:58:32 -0400 (EDT) Subject: Odd partition labeling In-Reply-To: <20050715181410.GD18022@linuxace.com> References: <20050715173732.GZ73653@thomson.net> <20050715181410.GD18022@linuxace.com> Message-ID: > What seems to be happening is that anaconda sees an existing label with the same > name, so it appends a 1 (see fsset.py). Since I mount the old root in my PRE, > think I'll now add the additional step of changing the labels so they won't > conflict with the new ones anaconda wants to use. Something like the below: This does indeed seem to be what's going on. I've added some code to clobber preexisting labels but that doesn't end up making any difference because labels are read well before the clobber code gets run. I don't yet know what the solution here is but I'll look at it again during the Rawhide cycle. - Chris From error27 at gmail.com Mon Jul 18 15:39:55 2005 From: error27 at gmail.com (Dan Carpenter) Date: Mon, 18 Jul 2005 17:39:55 +0200 Subject: Odd partition labeling In-Reply-To: References: <20050715173732.GZ73653@thomson.net> <20050715181410.GD18022@linuxace.com> Message-ID: On 7/18/05, Chris Lumens wrote: > > What seems to be happening is that anaconda sees an existing label with the same > > name, so it appends a 1 (see fsset.py). The thing is that I always clobber the partition table in the %pre step and I still see the extra 1 added. regards, dan carpneter From ian at smallworld.cx Tue Jul 19 11:37:00 2005 From: ian at smallworld.cx (Ian Leonard) Date: Tue, 19 Jul 2005 12:37:00 +0100 Subject: Kickstart: correct way to disable SELinux In-Reply-To: <1121638668.2769.54.camel@yoda.loki.me> References: <200507162014.45464.kimmo.koivisto@surfeu.fi> <42D943A6.7030600@smallworld.cx> <42D957A4.7020603@hostinthebox.net> <1121638668.2769.54.camel@yoda.loki.me> Message-ID: <42DCE5DC.2060304@smallworld.cx> Jesse Keating wrote: > On Sat, 2005-07-16 at 11:53 -0700, Dan Trainor wrote: > >>Hi - >> >>You know, this never did work for me, I'm not sure why. I've tried this >>on dozens of kickstarts, with dozens of different versions, to no avail. > > > How does this "not work"? I've done it w/ FC3 and FC4 and it works just > fine... Selinux is in disabled mode. Well, it's just failed for me. I have single disk install and found that the SELINUX line was set to enforcing in /etc/sysconfig/selinux. It's always worked in the past so off the top of my head I have the following observations: 1. I may have removed an important rpm 2. This is the first time I have used FC4 3. I may never have noticed (but I don't believe this one) 4. This is the first time I have build CD's as opposed to a http install So what does the selinux --disable line actually do? Does it just edit the sysconfig/selinux file or something more subtle? -- Ian Leonard Please ignore spelling and punctuation - I did. From error27 at gmail.com Tue Jul 19 15:33:56 2005 From: error27 at gmail.com (Dan Carpenter) Date: Tue, 19 Jul 2005 17:33:56 +0200 Subject: Kickstart: correct way to disable SELinux In-Reply-To: <42DCE5DC.2060304@smallworld.cx> References: <200507162014.45464.kimmo.koivisto@surfeu.fi> <42D943A6.7030600@smallworld.cx> <42D957A4.7020603@hostinthebox.net> <1121638668.2769.54.camel@yoda.loki.me> <42DCE5DC.2060304@smallworld.cx> Message-ID: On 7/19/05, Ian Leonard wrote: > So what does the selinux --disable line actually do? I look this up in the source yesterday. It's selinux --disabled w/ a 'd'. regards, dan From ian at smallworld.cx Tue Jul 19 16:15:10 2005 From: ian at smallworld.cx (Ian Leonard) Date: Tue, 19 Jul 2005 17:15:10 +0100 Subject: Kickstart: correct way to disable SELinux In-Reply-To: References: <200507162014.45464.kimmo.koivisto@surfeu.fi> <42D943A6.7030600@smallworld.cx> <42D957A4.7020603@hostinthebox.net> <1121638668.2769.54.camel@yoda.loki.me> <42DCE5DC.2060304@smallworld.cx> Message-ID: <42DD270E.8070407@smallworld.cx> Dan Carpenter wrote: > On 7/19/05, Ian Leonard wrote: > >>So what does the selinux --disable line actually do? > > > I look this up in the source yesterday. It's selinux --disabled w/ a 'd'. > Sorry, that was a typeo on that line only. I do use 'disabled'. -- Ian Leonard Please ignore spelling and punctuation - I did. From tibbs at math.uh.edu Wed Jul 20 16:23:18 2005 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: Wed, 20 Jul 2005 11:23:18 -0500 Subject: NFS over TCP? Message-ID: Ever since I migrated my servers to gigabit Ethernet I've found that kickstart installs of 100mbit clients are terribly slow. I think this has to do with a flow control problem with the switch, but that's another message on another mailing list. Through testing I found that the problem only happens with UDP NFS; TCP NFS and other TCP-based protocols are fine. I switched over to HTTP installs and it works, but due to the separate download step for each package it's not quite as quick so I'd like to try TCP NFS. Is it possible to enable TCP NFS for a kickstart install? - J< From seun.wright at csueastbay.edu Wed Jul 20 17:43:29 2005 From: seun.wright at csueastbay.edu (Seun Wright) Date: Wed, 20 Jul 2005 10:43:29 -0700 Subject: Problems installing Fedora core 4 with kickstart Message-ID: <42DE8D41.3070607@csueastbay.edu> Hello, I am new to kickstart and I am testing it in a vmware environment. What I did was walked to the installation process installing all the packages I wanted installed. After the installation process, I wanted to try out the kickstart with the config file generated by anaconda (/root/anaconda-ks.cfg). Everything works just fine until it gets to the section where it has to install the packages, then it crashes. I found out after some testing that the issue has to do with the packages selected. Could someone please shed some light on my problem. Thanks in advance. install url --url ftp://mirror.stanford.edu//pub/mirrors/fedora/linux/core/4/i386/os lang en_US.UTF-8 langsupport --default=en_US.UTF-8 keyboard us xconfig --card "VMWare" --videoram 16384 --hsync 31.5-37.9 --vsync 50-70 --resolution 800x600 --depth 16 network --device eth0 --bootproto dhcp rootpw --iscrypted XXX firewall --enabled --port=22:tcp selinux --permissive authconfig --enableshadow --enablemd5 timezone America/Los_Angeles bootloader --location=mbr # 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 --all --drives=sda #part /boot --fstype ext3 --size=100 --ondisk=sda #part pv.11 --size=0 --grow --ondisk=sda #volgroup VolGroup00 --pesize=32768 pv.11 #logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 --grow #logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=256 --grow --maxsize=512 %packages @ editors @ system-tools @ text-internet -perl-LDAP -doxygen -gcc-gfortran -perl-XML-Twig -samba-client kernel -mutt -pstack -fetchmail -perl-XML-NamespaceSupport -zsh grub -automake15 -elfutils kernel-devel -ctags -texinfo -xdelta -diffstat -perl-Crypt-SSLeay -cvs -automake14 -perl-XML-Parser -patchutils -cadaver -automake16 -valgrind -nmap e2fsprogs -oprofile -perl-XML-LibXML-Common -cscope -slrn -perl-XML-Encoding -python-ldap -bluez-pin -perl-XML-Grove -openldap-clients lynx lvm2 -perl-XML-SAX -valgrind-callgrind -rcs -ltrace -indent -perl-XML-LibXML -open sysstat -perl-XML-Dumper -byacc -automake17 -ckermit %post From info at hostinthebox.net Wed Jul 20 17:50:13 2005 From: info at hostinthebox.net (Dan Trainor) Date: Wed, 20 Jul 2005 10:50:13 -0700 Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: <42DE8D41.3070607@csueastbay.edu> References: <42DE8D41.3070607@csueastbay.edu> Message-ID: <42DE8ED5.7090004@hostinthebox.net> Seun Wright wrote: > Hello, > > I am new to kickstart and I am testing it in a vmware environment. What > I did was walked to the installation process installing all the packages > I wanted installed. After the installation process, I wanted to try out > the kickstart with the config file generated by anaconda > (/root/anaconda-ks.cfg). > > Everything works just fine until it gets to the section where it has to > install the packages, then it crashes. I found out after some testing > that the issue has to do with the packages selected. Could someone > please shed some light on my problem. > > Thanks in advance. > > > install > url --url > ftp://mirror.stanford.edu//pub/mirrors/fedora/linux/core/4/i386/os > lang en_US.UTF-8 langsupport --default=en_US.UTF-8 > keyboard us > xconfig --card "VMWare" --videoram 16384 --hsync 31.5-37.9 --vsync 50-70 > --resolution 800x600 --depth 16 > network --device eth0 --bootproto dhcp > rootpw --iscrypted XXX > firewall --enabled --port=22:tcp > selinux --permissive > authconfig --enableshadow --enablemd5 > timezone America/Los_Angeles > bootloader --location=mbr > # 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 --all --drives=sda > #part /boot --fstype ext3 --size=100 --ondisk=sda > #part pv.11 --size=0 --grow --ondisk=sda > #volgroup VolGroup00 --pesize=32768 pv.11 > #logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 > --grow > #logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 > --size=256 --grow --maxsize=512 > > %packages > @ editors > @ system-tools > @ text-internet > -perl-LDAP > -doxygen > -gcc-gfortran > -perl-XML-Twig > -samba-client > kernel > -mutt > -pstack > -fetchmail > -perl-XML-NamespaceSupport > -zsh > grub > -automake15 > -elfutils > kernel-devel > -ctags > -texinfo > -xdelta > -diffstat > -perl-Crypt-SSLeay > -cvs > -automake14 > -perl-XML-Parser > -patchutils > -cadaver > -automake16 > -valgrind > -nmap > e2fsprogs > -oprofile > -perl-XML-LibXML-Common > -cscope > -slrn > -perl-XML-Encoding > -python-ldap > -bluez-pin > -perl-XML-Grove > -openldap-clients > lynx > lvm2 > -perl-XML-SAX > -valgrind-callgrind > -rcs > -ltrace > -indent > -perl-XML-LibXML > -open > sysstat > -perl-XML-Dumper > -byacc > -automake17 > -ckermit %post > Hello, Seun - Where's your partition layout specified? This example shows all of that stuff commented out. I don't recall off the top of my head at what point the system tries to create partitions and install stuff, but I think it's right after it makes the partitions. If it can't, then it will die when trying to install packages. Wait around here for a bit to see if anyone else can folow up on this. Thanks -dant From error27 at gmail.com Wed Jul 20 17:49:35 2005 From: error27 at gmail.com (Dan Carpenter) Date: Wed, 20 Jul 2005 10:49:35 -0700 Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: <42DE8D41.3070607@csueastbay.edu> References: <42DE8D41.3070607@csueastbay.edu> Message-ID: On 7/20/05, Seun Wright wrote: > -ckermit %post %post should be on a seperate line. regards, dan carpenter From gressett at noao.edu Wed Jul 20 17:51:06 2005 From: gressett at noao.edu (Katrina Gressett) Date: Wed, 20 Jul 2005 10:51:06 -0700 (MST) Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: <42DE8D41.3070607@csueastbay.edu> Message-ID: I believe the problem is with valgrind. I had a similiar issue trying to install that package from CD a few months after doing a basic install without a lot of stuff we decided we wanted later. Trying to install it from the FC3 would crash the machine. I left it out and installed everything else we wanted without problems. Good Luck, Katrina On Wed, 20 Jul 2005, Seun Wright wrote: > Hello, > > I am new to kickstart and I am testing it in a vmware environment. What > I did was walked to the installation process installing all the packages > I wanted installed. After the installation process, I wanted to try out > the kickstart with the config file generated by anaconda > (/root/anaconda-ks.cfg). > > Everything works just fine until it gets to the section where it has to > install the packages, then it crashes. I found out after some testing > that the issue has to do with the packages selected. Could someone > please shed some light on my problem. > > Thanks in advance. > > > install > url --url > ftp://mirror.stanford.edu//pub/mirrors/fedora/linux/core/4/i386/os > lang en_US.UTF-8 langsupport --default=en_US.UTF-8 > keyboard us > xconfig --card "VMWare" --videoram 16384 --hsync 31.5-37.9 --vsync 50-70 > --resolution 800x600 --depth 16 > network --device eth0 --bootproto dhcp > rootpw --iscrypted XXX > firewall --enabled --port=22:tcp > selinux --permissive > authconfig --enableshadow --enablemd5 > timezone America/Los_Angeles > bootloader --location=mbr > # 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 --all --drives=sda > #part /boot --fstype ext3 --size=100 --ondisk=sda > #part pv.11 --size=0 --grow --ondisk=sda > #volgroup VolGroup00 --pesize=32768 pv.11 > #logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 > --grow > #logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 > --size=256 --grow --maxsize=512 > > %packages > @ editors > @ system-tools > @ text-internet > -perl-LDAP > -doxygen > -gcc-gfortran > -perl-XML-Twig > -samba-client > kernel > -mutt > -pstack > -fetchmail > -perl-XML-NamespaceSupport > -zsh > grub > -automake15 > -elfutils > kernel-devel > -ctags > -texinfo > -xdelta > -diffstat > -perl-Crypt-SSLeay > -cvs > -automake14 > -perl-XML-Parser > -patchutils > -cadaver > -automake16 > -valgrind > -nmap > e2fsprogs > -oprofile > -perl-XML-LibXML-Common > -cscope > -slrn > -perl-XML-Encoding > -python-ldap > -bluez-pin > -perl-XML-Grove > -openldap-clients > lynx > lvm2 > -perl-XML-SAX > -valgrind-callgrind > -rcs > -ltrace > -indent > -perl-XML-LibXML > -open > sysstat > -perl-XML-Dumper > -byacc > -automake17 > -ckermit %post > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > -- -- ___________________________________________________________ Katrina Gressett NSO/GONG -- Data Reduction Specialist gressett at noao.edu 520.318.8568 ___________________________________________________________ From seun.wright at csueastbay.edu Wed Jul 20 17:53:28 2005 From: seun.wright at csueastbay.edu (Seun Wright) Date: Wed, 20 Jul 2005 10:53:28 -0700 Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: References: <42DE8D41.3070607@csueastbay.edu> Message-ID: <42DE8F98.8050301@csueastbay.edu> An HTML attachment was scrubbed... URL: From phr at doc.ic.ac.uk Wed Jul 20 17:56:21 2005 From: phr at doc.ic.ac.uk (Philip Rowlands) Date: Wed, 20 Jul 2005 18:56:21 +0100 (BST) Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: <42DE8D41.3070607@csueastbay.edu> References: <42DE8D41.3070607@csueastbay.edu> Message-ID: On Wed, 20 Jul 2005, Seun Wright wrote: >Everything works just fine until it gets to the section where it has to >install the packages, then it crashes. I found out after some testing >that the issue has to do with the packages selected. Could someone >please shed some light on my problem. What's the error when it crashes? How did you determine it was related to package selection? Cheers, Phil From seun.wright at csueastbay.edu Wed Jul 20 18:01:09 2005 From: seun.wright at csueastbay.edu (Seun Wright) Date: Wed, 20 Jul 2005 11:01:09 -0700 Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: References: <42DE8D41.3070607@csueastbay.edu> Message-ID: <42DE9165.2060603@csueastbay.edu> An HTML attachment was scrubbed... URL: From tdiehl at rogueind.com Wed Jul 20 18:09:31 2005 From: tdiehl at rogueind.com (Tom Diehl) Date: Wed, 20 Jul 2005 14:09:31 -0400 (EDT) Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: <42DE9165.2060603@csueastbay.edu> References: <42DE8D41.3070607@csueastbay.edu> <42DE9165.2060603@csueastbay.edu> Message-ID: On Wed, 20 Jul 2005, Seun Wright wrote: > Philip, > > When I reduce my package selection to the following, it works fine: > > %packages > @ editors > @ system-tools > @ text-internet > @ dialup > kernel > grub > e2fsprogs > lynx > lvm2 > sysstat FC4 kickstart is B0rked. There is a fix here: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=160209 With any luck they might someday issue an update. Of course there is zero chance of getting updated iso's for this. Regards, Tom > > > Philip Rowlands wrote: > > On Wed, 20 Jul 2005, Seun Wright wrote: > > > > Everything works just fine until it gets to the section where it has to > install the packages, then it crashes. I found out after some testing > that the issue has to do with the packages selected. Could someone > please shed some light on my problem. > > > What's the error when it crashes? How did you determine it was related > to package selection? > > > Cheers, > Phil > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > > > > > > From seun.wright at csueastbay.edu Wed Jul 20 18:19:09 2005 From: seun.wright at csueastbay.edu (Seun Wright) Date: Wed, 20 Jul 2005 11:19:09 -0700 Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: References: <42DE8D41.3070607@csueastbay.edu> <42DE9165.2060603@csueastbay.edu> Message-ID: <42DE959D.10504@csueastbay.edu> Tom, Thanks for the link, but could you explain what rawhide is? Tom Diehl wrote: > On Wed, 20 Jul 2005, Seun Wright wrote: > >> Philip, >> >> When I reduce my package selection to the following, it works fine: >> >> %packages >> @ editors >> @ system-tools >> @ text-internet >> @ dialup >> kernel >> grub >> e2fsprogs >> lynx >> lvm2 >> sysstat > > > FC4 kickstart is B0rked. There is a fix here: > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=160209 > > With any luck they might someday issue an update. Of course there is > zero chance of getting updated iso's for this. > > Regards, > > Tom > >> >> >> Philip Rowlands wrote: >> >> On Wed, 20 Jul 2005, Seun Wright wrote: >> >> >> >> Everything works just fine until it gets to the section where it has to >> install the packages, then it crashes. I found out after some testing >> that the issue has to do with the packages selected. Could someone >> please shed some light on my problem. >> >> >> What's the error when it crashes? How did you determine it was related >> to package selection? >> >> >> Cheers, >> Phil >> >> _______________________________________________ >> 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 -- Seun Wright Network Security Analyst Information and Computing Services California State University - East Bay 510-885-2542 From tdiehl at rogueind.com Wed Jul 20 18:32:27 2005 From: tdiehl at rogueind.com (Tom Diehl) Date: Wed, 20 Jul 2005 14:32:27 -0400 (EDT) Subject: Problems installing Fedora core 4 with kickstart In-Reply-To: <42DE959D.10504@csueastbay.edu> References: <42DE8D41.3070607@csueastbay.edu> <42DE9165.2060603@csueastbay.edu> <42DE959D.10504@csueastbay.edu> Message-ID: On Wed, 20 Jul 2005, Seun Wright wrote: > Tom, > > Thanks for the link, but could you explain what rawhide is? Rawhide == development. This is currently what will someday be FC5. If you look on the mirror sites you will find a development tree that is what people refer to as Rawhide. It is a term left over from the old Red Hat linux when there was an actual Rewhide tree. Why they changed it with the introduction of Fedora is beyond me. Everyone still refers to it as Rawhide. Regards, Tom > > Tom Diehl wrote: > >> On Wed, 20 Jul 2005, Seun Wright wrote: >> >>> Philip, >>> >>> When I reduce my package selection to the following, it works fine: >>> >>> %packages >>> @ editors >>> @ system-tools >>> @ text-internet >>> @ dialup >>> kernel >>> grub >>> e2fsprogs >>> lynx >>> lvm2 >>> sysstat >> >> >> FC4 kickstart is B0rked. There is a fix here: >> https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=160209 >> >> With any luck they might someday issue an update. Of course there is zero >> chance of getting updated iso's for this. >> >> Regards, >> >> Tom >> >>> >>> >>> Philip Rowlands wrote: >>> >>> On Wed, 20 Jul 2005, Seun Wright wrote: >>> >>> >>> >>> Everything works just fine until it gets to the section where it has to >>> install the packages, then it crashes. I found out after some testing >>> that the issue has to do with the packages selected. Could someone >>> please shed some light on my problem. >>> >>> >>> What's the error when it crashes? How did you determine it was related >>> to package selection? >>> >>> >>> Cheers, >>> Phil >>> >>> _______________________________________________ >>> 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 seun.wright at csueastbay.edu Thu Jul 21 21:55:16 2005 From: seun.wright at csueastbay.edu (Seun Wright) Date: Thu, 21 Jul 2005 14:55:16 -0700 Subject: installing updated packages Message-ID: <42E019C4.1080004@csueastbay.edu> Hi, Does anyone know of any way to use kickstart to install updated packages? -- Seun Wright Network Security Analyst Information and Computing Services California State University - East Bay 510-885-2542 From joe at swelltech.com Thu Jul 21 22:05:14 2005 From: joe at swelltech.com (Joe Cooper) Date: Thu, 21 Jul 2005 17:05:14 -0500 Subject: installing updated packages In-Reply-To: <42E019C4.1080004@csueastbay.edu> References: <42E019C4.1080004@csueastbay.edu> Message-ID: <42E01C1A.8060702@swelltech.com> Seun Wright wrote: > Hi, > > Does anyone know of any way to use kickstart to install updated packages? %post yum -y update Any old update system will work, including up2date or apt-get. You'll probably also have to add the RPM GPG keys for the repos in question during your post, something like this: # Setup the keys, so yum works without complaining about keys... rpm --import http://www.fedora.us/FEDORA-GPG-KEY rpm --import /usr/share/rhn/RPM-GPG-KEY-fedora rpm --import /usr/share/rhn/RPM-GPG-KEY Though there's probably a command line switch to turn off the GPG check. From ltbraswell at yahoo.com Thu Jul 21 22:54:57 2005 From: ltbraswell at yahoo.com (Lance Braswell) Date: Thu, 21 Jul 2005 15:54:57 -0700 (PDT) Subject: probing for network devices during the pre phase Message-ID: <20050721225457.29991.qmail@web52109.mail.yahoo.com> Basically I have the same question as posed in the thread below: http://www.redhat.com/archives/kickstart-list/2003-December/msg00043.html In the %pre phase, I'm working on a small interactive menu to grab some basic info from which I will write out a complete ks.cfg using %include statements. For the network section, I would like to probe the system for network devices and prompt the user to choose one (and only one). I can't find anything obvious in the isys.py module like I can for probing disk drives. There is /usr/lib/anaconda/kudzu-probe-stub which might be a possibility. I guess I could parse ifconfig -a output but that seems kind of messy. Anyone know a better way to get at this info? regards, Lance __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From klaus.steden at thomson.net Thu Jul 21 23:12:34 2005 From: klaus.steden at thomson.net (Klaus Steden) Date: Thu, 21 Jul 2005 19:12:34 -0400 Subject: probing for network devices during the pre phase In-Reply-To: <20050721225457.29991.qmail@web52109.mail.yahoo.com> References: <20050721225457.29991.qmail@web52109.mail.yahoo.com> Message-ID: <20050721231234.GH2065@thomson.net> > Basically I have the same question as posed in the > thread below: > > http://www.redhat.com/archives/kickstart-list/2003-December/msg00043.html > > In the %pre phase, I'm working on a small interactive > menu to grab some basic info from which I will write > out a complete ks.cfg using %include statements. For > the network section, I would like to probe the system > for network devices and prompt the user to choose one > (and only one). I can't find anything obvious in the > isys.py module like I can for probing disk drives. > There is /usr/lib/anaconda/kudzu-probe-stub which > might be a possibility. I guess I could parse ifconfig > -a output but that seems kind of messy. Anyone know a > better way to get at this info? > I don't know about inside anaconda but you could try to parse /proc/net/dev, like this: list=`cat /proc/net/dev | grep : |awk -F: '{print $1}'` and then: select nic in $list do done ... on a related note, I did some hacking of the anaconda process itself and have developed a method for inserting code in between /sbin/loader (a.k.a init) and anaconda itself during kickstart ... some people have asked me to explain in detail, but I though it would be more useful to put it on the Anaconda Wiki ... but I'm having some difficulty getting in touch with people so far. So if anyone out there reads this and knows the email address of Greg Morgan (who maintains the Anaconda Wiki), please email me off list - it is much appreciated! cheers, Klaus From ltbraswell at yahoo.com Fri Jul 22 14:41:38 2005 From: ltbraswell at yahoo.com (Lance Braswell) Date: Fri, 22 Jul 2005 07:41:38 -0700 (PDT) Subject: probing for network devices during the pre phase In-Reply-To: <20050721231234.GH2065@thomson.net> Message-ID: <20050722144138.16934.qmail@web52109.mail.yahoo.com> Thanks for the /proc/net/dev hint Klaus. Here is the relavant anaconda step for the same: #!/usr/bin/python import sys sys.path.append('/usr/lib/anaconda') import network interfaces = network.Network().available() for interface in interfaces.keys() : print interface #end I am anxiously awaiting the details of your /sbin/loader hack :) Hope you can get in touch with the wiki owner. --- Klaus Steden wrote: > > Basically I have the same question as posed in the > > thread below: > > > > > http://www.redhat.com/archives/kickstart-list/2003-December/msg00043.html > > > > In the %pre phase, I'm working on a small > interactive > > menu to grab some basic info from which I will > write > > out a complete ks.cfg using %include statements. > For > > the network section, I would like to probe the > system > > for network devices and prompt the user to choose > one > > (and only one). I can't find anything obvious in > the > > isys.py module like I can for probing disk drives. > > There is /usr/lib/anaconda/kudzu-probe-stub which > > might be a possibility. I guess I could parse > ifconfig > > -a output but that seems kind of messy. Anyone > know a > > better way to get at this info? > > > I don't know about inside anaconda but you could try > to parse /proc/net/dev, > like this: > > list=`cat /proc/net/dev | grep : |awk -F: '{print > $1}'` > > and then: > > select nic in $list > do > > done > > ... on a related note, I did some hacking of the > anaconda process itself and > have developed a method for inserting code in > between /sbin/loader (a.k.a > init) and anaconda itself during kickstart ... some > people have asked me to > explain in detail, but I though it would be more > useful to put it on the > Anaconda Wiki ... but I'm having some difficulty > getting in touch with people > so far. > > So if anyone out there reads this and knows the > email address of Greg Morgan > (who maintains the Anaconda Wiki), please email me > off list - it is much > appreciated! > > cheers, > Klaus > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list > __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From James_Martin at ao.uscourts.gov Fri Jul 22 15:09:50 2005 From: James_Martin at ao.uscourts.gov (James_Martin at ao.uscourts.gov) Date: Fri, 22 Jul 2005 11:09:50 -0400 Subject: probing for network devices during the pre phase In-Reply-To: <20050722144138.16934.qmail@web52109.mail.yahoo.com> Message-ID: I do something similar to your loader hack, but I've found it safer (and easier) to hack the anaconda python script.. You can insert routines right before the kickstart process. On RHEL3, this is in the initrd.img, at /usr/bin/anaconda. I modify the code at this block: for f in os.listdir("/usr/%s/python2.2/site-packages/rhpl" %(libdir,)): if os.access("/tmp/updates/rhpl/%s" %(f,), os.R_OK): continue elif f.endswith(".pyc") and os.access("/tmp/updates/rhpl/%s" %(f[:-1],), os.R_OK): # dont copy .pyc files we are replacing with updates continue os.symlink("/usr/%s/python2.2/site-packages/rhpl/%s" %(libdir, f), "/tmp/updates/rhpl/%s" %(f,)) ##do array config stuff <---------------------------------Modification begin print "Running Array Config...\n" os.system("/usr/local/ao/ao-stuff.sh") ## <---------------------------------------------------------------Modification End # For anaconda in test mode if (os.path.exists('isys')): sys.path.append('isys') Inside that shell script (ao-stuff.sh), I run HP's array configuration utility. It configures the hardware RAID array. It creates the logical drives. Anaconda will fail if you attempt a kickstart with no logical drives. I'm sure you can apply this to what you are trying to do. James James S. Martin, RHCE Contractor Administrative Office of the United States Courts Washington, DC (202) 502-2394 kickstart-list-bounces at redhat.com wrote on 07/22/2005 10:41:38 AM: > Thanks for the /proc/net/dev hint Klaus. Here is the > relavant anaconda step for the same: > > #!/usr/bin/python > > import sys > sys.path.append('/usr/lib/anaconda') > > import network > > interfaces = network.Network().available() > for interface in interfaces.keys() : > print interface > > > #end > > I am anxiously awaiting the details of your > /sbin/loader hack :) Hope you can get in touch with > the wiki owner. > > > --- Klaus Steden wrote: > > > > Basically I have the same question as posed in the > > > thread below: > > > > > > > > > http://www.redhat.com/archives/kickstart-list/2003-December/msg00043.html > > > > > > In the %pre phase, I'm working on a small > > interactive > > > menu to grab some basic info from which I will > > write > > > out a complete ks.cfg using %include statements. > > For > > > the network section, I would like to probe the > > system > > > for network devices and prompt the user to choose > > one > > > (and only one). I can't find anything obvious in > > the > > > isys.py module like I can for probing disk drives. > > > There is /usr/lib/anaconda/kudzu-probe-stub which > > > might be a possibility. I guess I could parse > > ifconfig > > > -a output but that seems kind of messy. Anyone > > know a > > > better way to get at this info? > > > > > I don't know about inside anaconda but you could try > > to parse /proc/net/dev, > > like this: > > > > list=`cat /proc/net/dev | grep : |awk -F: '{print > > $1}'` > > > > and then: > > > > select nic in $list > > do > > > > done > > > > ... on a related note, I did some hacking of the > > anaconda process itself and > > have developed a method for inserting code in > > between /sbin/loader (a.k.a > > init) and anaconda itself during kickstart ... some > > people have asked me to > > explain in detail, but I though it would be more > > useful to put it on the > > Anaconda Wiki ... but I'm having some difficulty > > getting in touch with people > > so far. > > > > So if anyone out there reads this and knows the > > email address of Greg Morgan > > (who maintains the Anaconda Wiki), please email me > > off list - it is much > > appreciated! > > > > cheers, > > Klaus > > > > _______________________________________________ > > Kickstart-list mailing list > > Kickstart-list at redhat.com > > > https://www.redhat.com/mailman/listinfo/kickstart-list > > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From brilong at cisco.com Fri Jul 22 15:31:17 2005 From: brilong at cisco.com (Brian Long) Date: Fri, 22 Jul 2005 11:31:17 -0400 Subject: probing for network devices during the pre phase In-Reply-To: References: Message-ID: <1122046277.3782.24.camel@localhost.localdomain> > > print "Running Array Config...\n" > os.system("/usr/local/ao/ao-stuff.sh") > > Inside that shell script (ao-stuff.sh), I run HP's array configuration > utility. It configures the hardware RAID array. It creates the logical > drives. Anaconda will fail if you attempt a kickstart with no logical > drives. I'm sure you can apply this to what you are trying to do. This is cool. How does ao-stuff.sh get installed in /usr/local/ao? Also, is the hack to Anaconda really needed? Could you just modify the proper Python .py file and put it in an updates.img file? /Brian/ -- Brian Long | | | IT Data Center Systems | .|||. .|||. Cisco Linux Developer | ..:|||||||:...:|||||||:.. Phone: (919) 392-7363 | C i s c o S y s t e m s From cleve.cornelius at kroger.com Fri Jul 22 20:02:47 2005 From: cleve.cornelius at kroger.com (cleve.cornelius at kroger.com) Date: Fri, 22 Jul 2005 16:02:47 -0400 Subject: Cleve Cornelius/MIS/CORP/KrogerCo is out of the office. Message-ID: I will be out of the office starting 07/22/2005 and will not return until 07/25/2005. From James_Martin at ao.uscourts.gov Fri Jul 22 21:22:36 2005 From: James_Martin at ao.uscourts.gov (James_Martin at ao.uscourts.gov) Date: Fri, 22 Jul 2005 17:22:36 -0400 Subject: probing for network devices during the pre phase In-Reply-To: <1122046277.3782.24.camel@localhost.localdomain> Message-ID: I tried doing that before with updates.img, and never had any luck. You're right though, it should work. I deconstruct the stage2.img and put my own files in there and re-construct it. James S. Martin, RHCE Contractor Administrative Office of the United States Courts Washington, DC (202) 502-2394 kickstart-list-bounces at redhat.com wrote on 07/22/2005 11:31:17 AM: > > > > > print "Running Array Config...\n" > > os.system("/usr/local/ao/ao-stuff.sh") > > > > > Inside that shell script (ao-stuff.sh), I run HP's array configuration > > utility. It configures the hardware RAID array. It creates the logical > > drives. Anaconda will fail if you attempt a kickstart with no logical > > drives. I'm sure you can apply this to what you are trying to do. > > This is cool. How does ao-stuff.sh get installed in /usr/local/ao? > Also, is the hack to Anaconda really needed? Could you just modify the > proper Python .py file and put it in an updates.img file? > > /Brian/ > > -- > Brian Long | | | > IT Data Center Systems | .|||. .|||. > Cisco Linux Developer | ..:|||||||:...:|||||||:.. > Phone: (919) 392-7363 | C i s c o S y s t e m s > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list From ltbraswell at yahoo.com Tue Jul 26 17:30:42 2005 From: ltbraswell at yahoo.com (Lance Braswell) Date: Tue, 26 Jul 2005 10:30:42 -0700 (PDT) Subject: customizing anaconda through kickstart Message-ID: <20050726173042.73133.qmail@web52108.mail.yahoo.com> There have been a number of threads on this list dealing with getting input from the user to generate a custom kickstart file on the fly. Looks like the Rocks Cluster project (http://www.rocksclusters.org) has done some pretty clever things by hacking Anaconda directly to insert custom screens and to generate kickstart stuff from xml. As far as I can tell it's not very well documented but maybe it would be useful for someone on this list. These slides are a good overview of the hack: http://www.rocksclusters.org/rocks-a-palooza-slides/rolls-overview.pdf Here is link to a a sample ks.cfg which has a lot of their customizations inline. It takes advantage of the %installclass kickstart directive. http://www.rocksclusters.org/ftp-site/rocks-3.0.0/rocks-dist/enterprise/2.1AW/en/os/ia64/ks.cfg I'm not associated with this project in any way. Just thought it was pretty interesting. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com From drkludge at cox.net Fri Jul 29 09:04:50 2005 From: drkludge at cox.net (Greg Morgan) Date: Fri, 29 Jul 2005 02:04:50 -0700 Subject: Anaconda Wiki In-Reply-To: <20050729002543.GP897@thomson.net> References: <20050729002543.GP897@thomson.net> Message-ID: <42E9F132.4000709@cox.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Klaus, I decided to copy the anaconda ks list after I finished writing the notes below. Klaus Steden wrote: > Hello, > > I would like to contribute something to the Anaconda Wiki. That's great news. > > I have developed an approach for inserting code (not just '%pre' stuff) into > the kickstart procedure, and a number of people on the kickstart list have > asked me for more info. I thought it would be easier and more helpful if I > were to make this info available through the official Anaconda channels, > rather than emailing a bunch of people and having the info get lost in the > ether. That sounds very interesting. > > I understand that you maintain this section of the Fedora Wiki; what is the > best way to get this info online (if you think it's a bad idea, feel free to > tell me to go lump it :>). Go here http://fedoraproject.org/wiki/FrontPage Do this.... Welcome to the [WWW] Fedora Project Wiki. The purpose of this wiki is to provide an external resource for Fedora Developers and Users inside and outside of Red Hat to share information and work together on plans. In order to help in the editing you must first setup an account by going to UserPreferences. After that, inform someone within the EditGroup, so that they can add you. ...please let me know when you are done so that I can add you. MoinMoin likes "CamelCase" wiki names and user names i.e. KlausSteden, if you are not familiar with wiki links made this way. Klaus..It would be great to have you contribute what you'd like. The Anaconda page is located here http://fedoraproject.org/wiki/Anaconda. I started converting the rau wiki with notes here http://fedoraproject.org/wiki/GregMorgan. Then I decided half way through that Mr Rau was in process of revising the wiki and I was trying to convert two pages per page in some cases. I also realized that the Rau wiki covered just part of what anaconda needed in the way of documentation. I have been trying to get this wiki, http://phpwebsite-comm.sourceforge.net/wiki/index.php , to a place that I can switch back to the Anaconda wiki. I am Dr kludge on that wiki. (SourceForge.net is having mysql database resource issues right now so the wiki paints slow. You may have to click the links twice.) Anyhow, It looks like Jeremy Katz has recently reorganized the anaconda focus along similar lines to what I have been thinking https://www.redhat.com/archives/kickstart-list/2005-June/msg00064.html: 1.) A Basic CD install was needed. Stuart Ellis and Paul W. Frields have created an updated CD installation guide here http://fedoraproject.org/wiki/AnacondaBuildDocProject. Docbook drives me crazy at this point but that's where the doc project would like to take all the documentation in the wiki too, if you didn't know. 2.) A linux prompt cmdline boot guide. New users don't always know how to combine the basic encyclopedic list of parms into something useful. The section needs to be "cookbook oriented". One of these topics would include creating an install tree and show new users how to install from that or how to use the iso directory method. 3.) The install tree topic could then be a segue into remastering the Fedora distribution. There are a whole bunch of small skills a user needs to learn to know how to remaster the four CD Fedora set into one CD, for example. Remastering may be the place to talk about creating install classes. 4.) There needs to be a demo section on how to create a kickstart file. There's pieces all over the net on how to create the kickstrt file. One problem is that people don't understand that they have a live machine that they can script at two points during the installation. So I thought of taking http://www.gforge.org (they just released 4.5) or http://meta.wikimedia.org/wiki/Help:Installation http://meta.wikimedia.org/wiki/Apache_config as an example project. You see if you create an install tree, or remaster the CDs you can put all the files you need on the last CD and script the rest of the install of the application on the new Fedora box. Although I have not tried it yet, I believe yum can be used in the kickstart %post section. That may be another way to handle this in the having additional files available in the post section. The demo would show a person how to collect all the pieces to create a kickstart file. This guide would also then refer back to the command line section of part two. It looks like you have some interesting ideas for this area too....go for it. 5.) Part three is also a segue into the last part that is required for exposing all of Anaconda. This is the section on how to obtain the sources and modify both the phase one C boot code and phase two python installation code. I think presenting the ideas in the order above would help make mastering the programming learning curve easier to handle. I started using MoinMoin wiki at the beginning of the year. We ran into space problems on sourceforge.net. I created my very own denial of service with the MoinMoin wiki on SourceForge. It saves a whole copy of each document that is changed for diffs. There's a 100 meg limit on SF's web server. :P Anyhow we had to switch to MediaWiki. MediaWiki is the wiki engine behind http://www.wikipedia.org and all of its sister projects. I have come across some interesting tools that may help document the data structures used in Anaconda. In the link earlier, msg00064, Jeremy mentions the comps file documentation http://rhlinux.redhat.com/anaconda/comps.html. I am thinking of some key pictures of the comps file using graphviz code would help the documentation. I have been playing with this idea here http://phpwebsite-comm.sourceforge.net/wiki/index.php?title=Graphviz_Code and here http://phpwebsite-comm.sourceforge.net/wiki/index.php?title=PhpWebSite_Root Many new users get mixed up on the document root of a web server and this picture may help me revise the document once again. As noted in the Graphviz_Code page MediaWiki has extensions that takes graphviz code and generates both the web image map and graphic on the fly. Then anyone can edit the graphic without having to have gimp or dia tools. The packages for graphviz and ploticus are in extras, by the way. Anyhow, key pictures can be generated with the CLI tools and put in the web pages. I am not yet sure where to upload graphics to the fedora wiki so that they can be included in the wiki pages, yet. I think the graphics would be a nice touch for both the install tree and comps documentation. I used the pl code from the LJ article below to generated the initial images and then revised the file. 6.) Finally, The last section is to create a wiki FAQ. So if you look at this page http://phpwebsite-comm.sourceforge.net/wiki/index.php?title=User_FAQ you can see how I am using actual questions from forum posts to create the question. Then if you click on the link, it will take you to the answer. The answer is then in the context of the other surrounding documentation. You only have to maintain the answer in one spot and not two as with other FAQs. I believe MoinMoin can do something similar. I did not understand wiki's until around November 2004. Now I wonder why I haven't used them before now. Here's a link to a new page that I am writing on a new wiki module for phpWebSite that Blindman1344 has authored. It sums up my thoughts on wiki use for the last six months that I have used 'em. http://phpwebsite-comm.sourceforge.net/wiki/index.php?title=Wiki * Use the release early and release often creed. Unfortunately, the page may look like a mess at first but eventually it all pulls together. * Take the good parts of emails like this to begin a document. * Wiki's make it easy to revise and reorganize. * Don't be afraid to help revise other people's work. That's the whole point of a wiki is that it is a team effort. * Don't be upset if other people edit your work. It's a team effort. * Wiki's have diff engines and history pages so you can see how you or other writers have made. * Try to find a way to build a community around the wiki. This email is an attempt to draw more people into the Anaconda wiki. Here's some links that may be helpful. http://www.linux-magazine.com/issue/55/Wiki_And_Blog_Intro.pdf http://www.linux-magazine.com/issue/55/TWiki.pdf http://www.linux-magazine.com/issue/41/Moin_Moin_Wiki.pdf http://www.linuxjournal.com/article/7275 http://www.linuxjournal.com/article/7583 http://www.oddmuse.org/cgi-bin/wiki http://www.usemod.com/cgi-bin/mb.pl?WikiLifeCycle http://meta.wikimedia.org/wiki/Category:Mediawiki_Extensions http://meta.wikimedia.org/wiki/EasyTimeline http://www.islandseeds.org/wiki/Test:Graphviz Greg -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org iD8DBQFC6fEyxyxe5L6mr7IRAjeJAKCMs3pAULz9KOM6Zhm59Q5LbDTLCQCgn/Wu b3O2ytc3uJQd9luQlcU6fA4= =ScmS -----END PGP SIGNATURE----- From kickstart at lordcow.org Fri Jul 29 17:44:15 2005 From: kickstart at lordcow.org (gareth) Date: Fri, 29 Jul 2005 19:44:15 +0200 Subject: death In-Reply-To: References: <20050602155119.GA25662@lordcow.org> <5520b27905060209355d6f839c@mail.gmail.com> <20050602164225.GA27011@lordcow.org> Message-ID: <20050729174415.GA32294@lordcow.org> sorry i only got back into this now. changing it to '--size=1' works like a charm, thanx On Thu 2005-06-02 (17:08), Dan Carpenter wrote: > This is completely wild guess. > > What happens if you change > old: part pv.2 --size=0 --grow --ondisk=sda > new: part pv.2 --size=1 --grow --ondisk=sda > > Anyway it has something to do with partitioning... Perhaps the > existing partitions are screwing it up. I've seen that happen quite > often with software RAID for example. > > regards, > dan carpenter > > _______________________________________________ > Kickstart-list mailing list > Kickstart-list at redhat.com > https://www.redhat.com/mailman/listinfo/kickstart-list