Anaconda-devel May 2006

anaconda-devel@lists.fedoraproject.org

40 participants
57 discussions

installing anaconda
by amin saberian 10 May '06

10 May '06

Hello. I want build a new distro. I want customize anaconda installer for my distro, but i cant compail and install it. How can i compail an install anaconda tarbal on fedora core 4 or redhat enterprice 4. Regards. --------------------------------- Yahoo! Mail goes everywhere you do. Get it on your phone.

2 1

anaconda bug: loader2/cdinstall.c's use of probeDevices
by Steven Augart 09 May '06

09 May '06

Thanks to some debugging statements, I've tracked down the source of the /sbin/loader crash I was experiencing that was hurting Mazu's installation process. The bug is in loader2/cdinstall.c's use of Kudzu's probeDevices: devices = probeDevices(CLASS_CDROM, BUS_UNSPEC, 0); if (!devices) { logMessage(ERROR, "No CDROM devices found!"); return 1; } Unfortunately, Kudzu does not return a null pointer in this case. Here is the official documentation (out of /usr/include/kudzu/kudzu.h) for what is actually returned: /* Probe for devices of the specified class, on the specified bus, * with the specified class. Returns a NULL-terminated array of * device (or subclass) pointers */ struct device ** probeDevices (enum deviceClass probeClass, enum deviceBus probeBus, int probeFlags ); A NULL-terminated array of zero elements is not a NULL pointer. So, first of all, if someone can tell me how I can go about submitting a patch for the bug in cdinstall.c, I would appreciate knowing what to do. Second of all, it appears that Kudzu 1.2.24.3, (as distributed in the FC 5 RPM kudzu-devel-1.2.34.3-1), is not detecting the CD drive on the IBM BladeCenter I've been testing on. This is an IBM Machine Type 8677, model # 2XX. (8677-2XX), manufactured in April of 2003. I'll write directly to Bill Notting, the current Kudzu maintainer, and see if I can help make that probing work in Kudzu. Thanks to Dan Carpenter for his helpful reply when I wrote to this list initially about debugging /sbin/loader. Steven -- Steven Augart Mazu Networks Cambridge, Massachusetts, USA

3 6

Re: anaconda bug: loader2/cdinstall.c's use of probeDevices
by Bill Nottingham 09 May '06

09 May '06

Steven Augart (saugart(a)mazunetworks.com) said: > When you have a chance to look at the USB devices list dump I sent you, > please let me know if I can provide more helpful information. For > Mazu's internal purposes (of getting an installer to work on the blade > center), I can probably add some kludges that will detect the CD ROM and > will use mknod() to create the appropriate block device in /tmp so that > we can read the ks.cfg file off of the CD. However, I'd certainly > prefer to make the changes be more generally useful to other users of > libkudzu. I'm sorry if I wasn't clear... if you're getting to the second stage at *all*, with the installer CD in the USB cd-rom, then the detection is working - there's nothing to fix. Bill

1 0

Locating %post scripts, files...
by Philip Prindeville 09 May '06

09 May '06

On other quick question that I didn't find documented... Obvious the more stuff you put into isolinux.img the less RAM you have left over for other stuff, so this is best left with the fewest additional files... That said, is it "normal" and "proper" (etc) for a %post script to make references to stuff in /mnt/source/? For instance, let's say I have a mini-tar overlay of files I want to explode into the new root that's being built (since everything is prefixed in the tarchive by ./inst-helper/ anyway)... is: %post --nochroot (cd /mnt/sysimage ; gunzip -c | tar xf -) < /mnt/source/overlay-root.tgz %post chvt 3 # for debugging... set -x /inst-helper/mkprt --all ... is that a good practice? Or is there a better way to do this? And does the output of all of the %pre (well, that might be tricky) and %post scripts get captured/logged somewhere? Both on a vty and later into a saved log file? Thanks, -Philip

2 3

RE: Default for ks=cdrom
by Shawn Starr 09 May '06

09 May '06

Sure, you can put it into the initrd, if you do this you'll need to change the isolinux.cfg as ks=file:/ks.cfg (being / of initrd). You can also put it into CD-ROM image before you make the .iso. In this case you'd use ks=cdrom:/ks.cfg or some other location you like on the CD. Shawn. -----Original Message----- From: Philip Prindeville [mailto:philipp_subx@redfish-solutions.com] Sent: Tuesday, May 09, 2006 5:20 PM To: Shawn Starr Cc: anaconda-devel-list(a)redhat.com Subject: Re: Default for ks=cdrom So it needs to go in "/" in initrd.img? Is that what I'm understanding? -Philip Shawn Starr wrote: > The kickstart file is only found once the initrd is first loaded via isolinux, so this is after the isolinux stuff has over to the initrd, by passing the ks=cdrom option the loader will check the CD/DVD once its mounted inside Linux for ks.cfg in /. > > Files found in isolinux are not visible to Linux as they are a different CD layer. > > -----Original Message----- > From: Philip Prindeville [mailto:philipp_subx@redfish-solutions.com] > Sent: Tuesday, May 09, 2006 4:39 PM > To: Shawn Starr > Cc: anaconda-devel-list(a)redhat.com > Subject: Re: Default for ks=cdrom > > > Ok, I've only been working on writing Kickstart images for a couple of > days now, > so... what is / mapped to? The isolinux.img filesystem? Is that where > you're > saying the ks.cfg needs to be? > > Repacking the isolinux.img is (for me, anyway) a fair amount of work > that's best > avoided... I'd just as soon have the default location for the file be > /mnt/source/ks.cfg > or /mnt/source/isolinux/ks.cfg ... > > Am I alone in this? > > -Philip > > Shawn Starr wrote: > >> I thought it tries the toplevel '/' by default. >> >> -----Original Message----- >> From: anaconda-devel-list-bounces(a)redhat.com >> [mailto:anaconda-devel-list-bounces@redhat.com]On Behalf Of Philip >> Prindeville >> Sent: Tuesday, May 09, 2006 4:03 PM >> To: anaconda-devel-list(a)redhat.com >> Subject: Default for ks=cdrom >> >> >> Forgot to ask... When one specifies "ks=cdrom" with no path, >> where does Anaconda end up looking? In the CD's root, or in >> isolinux? Or what? >> >> Should this be more explicitly documented? >> >> -Philip >> >> _______________________________________________ >> Anaconda-devel-list mailing list >> Anaconda-devel-list(a)redhat.com >> https://www.redhat.com/mailman/listinfo/anaconda-devel-list >> >> > >

1 0

RE: Default for ks=cdrom
by Shawn Starr 09 May '06

09 May '06

The kickstart file is only found once the initrd is first loaded via isolinux, so this is after the isolinux stuff has over to the initrd, by passing the ks=cdrom option the loader will check the CD/DVD once its mounted inside Linux for ks.cfg in /. Files found in isolinux are not visible to Linux as they are a different CD layer. -----Original Message----- From: Philip Prindeville [mailto:philipp_subx@redfish-solutions.com] Sent: Tuesday, May 09, 2006 4:39 PM To: Shawn Starr Cc: anaconda-devel-list(a)redhat.com Subject: Re: Default for ks=cdrom Ok, I've only been working on writing Kickstart images for a couple of days now, so... what is / mapped to? The isolinux.img filesystem? Is that where you're saying the ks.cfg needs to be? Repacking the isolinux.img is (for me, anyway) a fair amount of work that's best avoided... I'd just as soon have the default location for the file be /mnt/source/ks.cfg or /mnt/source/isolinux/ks.cfg ... Am I alone in this? -Philip Shawn Starr wrote: > I thought it tries the toplevel '/' by default. > > -----Original Message----- > From: anaconda-devel-list-bounces(a)redhat.com > [mailto:anaconda-devel-list-bounces@redhat.com]On Behalf Of Philip > Prindeville > Sent: Tuesday, May 09, 2006 4:03 PM > To: anaconda-devel-list(a)redhat.com > Subject: Default for ks=cdrom > > > Forgot to ask... When one specifies "ks=cdrom" with no path, > where does Anaconda end up looking? In the CD's root, or in > isolinux? Or what? > > Should this be more explicitly documented? > > -Philip > > _______________________________________________ > Anaconda-devel-list mailing list > Anaconda-devel-list(a)redhat.com > https://www.redhat.com/mailman/listinfo/anaconda-devel-list >

2 1

RE: Default for ks=cdrom
by Shawn Starr 09 May '06

09 May '06

I thought it tries the toplevel '/' by default. -----Original Message----- From: anaconda-devel-list-bounces(a)redhat.com [mailto:anaconda-devel-list-bounces@redhat.com]On Behalf Of Philip Prindeville Sent: Tuesday, May 09, 2006 4:03 PM To: anaconda-devel-list(a)redhat.com Subject: Default for ks=cdrom Forgot to ask... When one specifies "ks=cdrom" with no path, where does Anaconda end up looking? In the CD's root, or in isolinux? Or what? Should this be more explicitly documented? -Philip _______________________________________________ Anaconda-devel-list mailing list Anaconda-devel-list(a)redhat.com https://www.redhat.com/mailman/listinfo/anaconda-devel-list

2 1

Default for ks=cdrom
by Philip Prindeville 09 May '06

09 May '06

Forgot to ask... When one specifies "ks=cdrom" with no path, where does Anaconda end up looking? In the CD's root, or in isolinux? Or what? Should this be more explicitly documented? -Philip

1 0

Where is the error ? FC5 Customization
by Fabricio Lopes de Souza 09 May '06

09 May '06

I followed this steps : 1-copy all files from all cds to /fedora_dev/fc5/i386/ 2- # createrepo -g Fedora/base/comps.xml . 3- #pkgorder /fedora_dev/fc5/i386/ i386 /fedora_dev/fc5/i386/Fedora/RPMS/ | tee pkgfile 4- #/usr/lib/anaconda-runtime/buildinstall --comp 5 --pkgorder pkgfile --version 4 --product "Fedora Core" --release "Fedora Core 5" --prodpath Fedora --discs "Fedora Core 5" /fedora_dev/fc5/i386/ 5- # mkisofs -R -J -T -no-emul-boot -boot-load-size 4 -boot-info-table \ -V "Fedora Core 5" \ -A "Fedora Core 5" \ -b isolinux/isolinux.bin \ -c isolinux/boot.cat \ -o FC5-i386-disc1.iso \ i386 and booted in QEMU with : # qemu -boot d -cdrom fc/FC5-i386-disc1.iso hdc.img and the system boots and tell me to choose a intall method , then i choose LOCAL CD-ROM , but "No Fedora core cd Found" !!!!! what is wrong ???? thanks ... []'s

1 0

FC5 buildinstall changes
by Alan Milligan 06 May '06

06 May '06

Hi, I'm now porting our FC3/4 build environment to anaconda-11.0.5-1 - please do advise if I should be using fedora-development trunk instead. I have a very sophisticated Zope-based application to manage all of this, and thus am rather meticulous about reliably and repeatably being able to create custom installers. A fundamental tenet of the build process is working backwards from a bunch of objects that dynamically create a comps.xml and thus the requisites such would indicate. I have a range of anomalies that I'd like explained, or at least contemplated - please excuse me if I've missed something in the release notes ;) 1. Regardless of shipping these, it appears that all anaconda-runtime requires dependencies need to be slipped into the RPMS for the second buildinstall parse. I expect this is a general-enough rule to consider fundamental and I could derive these from whatever version's requires dependencies suggest? Any idea's for what one could/should do to figure/separate a curses install instead of the full-X that the requires dependencies suggests? 2. I get some buildinstall errors about /usr/share/X11/fonts/* needs to be present - I can fix this by including xorg-x11-fonts-* - what's the recommended minimal set of such? 3. The pango-querymodules step requires libXdmcp - this and I suspect a whole range of other dependencies is not in the requires tags. I also get GLib-WARNING **: getpwuid_r(): failed due to unknown user id (0) - which is rather strange, indicating that root should not invoke it ... 4. I seem to have lost bogl and ash from FC5 - should I be worried?? 5. /usr/share/hwdata/videoaliases/* for trimpciids is missing - this seems to require xorg-x11-drv-* - what would be a/the recommended set of video drivers to include?? 6. buildinstall now invokes pkgorder which seems to completely ignore the possibility of defining groups within the comps.xml. There are a whole bunch of hard-coded addGroups() calls which obviously barfs when a large number of these are not present in my comps.xml - if I don't want X Windows, or mysql - surely I should be accomodated. What am I missing here?? 7. When unpacking my kernel (kernel-2.6.15-1.2054_FC5.root.i686.rpm), I have warnings that a worryingly large number of modules "not found in kernel rpm": 3c501 3c503 3c505 3c507 3c515 82596 AM53C974 abyss aic94xx airport arlan at1700 bcm5700 bmac cs89x0 ctc dpt_i2o eepro eepro100 eexpress es3210 eth16i ewrk3 fcal fmv18x hid hp hp-plus i82365 ibmtr ibmveth ibmvscsic ide-cd ipr iseries_veth lance lcs lne390 mace mambonet md mousedev myri_sbus ncr53c8xx ne ne3210 netiucv ni5010 ni65 pcmcia pluto qeth qla2x00 qla6322 qlogicisp qlogicpti siimage sk98lin sktr spidernet sunbmac sunlance sunqe sym53c8xx_2 tcic tcmscsim tmspci tulip_old veth viodasd wavelan xenblk xennet yenta_socket Some of these (ide-cd, mousedev, yenta_socket) seem quite critical to proceedings... 8. Unless I include fedora-logos (or it's equivalent), the anaconda-runtime/splashtolss.sh script cannot be found. Is this in the correct package?? 9. Lastly, I do a bit of an integrity check on files I expect to see present if there is to be any hope at all of booting the CD. I'm now not finding: /usr/src/redhat/BUILD/i386/isolinux/snake.msg /usr/src/redhat/BUILD/i386/BastionLinuxLite/base/hdlist /usr/src/redhat/BUILD/i386/BastionLinuxLite/base/hdlist2 /usr/src/redhat/BUILD/i386/BastionLinuxLite/base/hdstg2.img /usr/src/redhat/BUILD/i386/BastionLinuxLite/base/netstg2.img I do instead see base/minstg2.img - does this file deprecate these others? I'd like to hear opinions about all of this in the hopes of avoiding large production runs of anaconda-installer coffee mug coasters ;) Alan

2 1

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Anaconda-devel May 2006