Anaconda-devel April 2005

anaconda-devel@lists.stg.fedoraproject.org

37 participants
47 discussions

modifying redhat-logos.rpm
by bhavin dalal 03 May '05

03 May '05

hello everyone. i am trying to modify the red hat logos and want to make my own redhat-logos.rpm. i dont know how to modify it because .png file cannot be patched. please tell me how to do so. i do have redhat-logos.src.rpm with me. i had untarred the redhat-logos.bz2 now please tell me how to proceed further... please give me list of commands that i need to execute. thank you in advance linuxbh

2 1

Device Parameter with Anaconda
by Ronald Reed 03 May '05

03 May '05

Hmm, let me ask this another way. What part of the code reads the kickstart file and parses it? I am digging thru the source now, but I am not finding it. I basically want to see how the code parses the kickstart file and find out why it is only processing a single device line. -- =========================== Ron Reed RedHat Certified Engineer SGP Computer Department Manager Unix Systems Administrator ARM SGP CART Site (580)388-4053 ron.reed(a)arm.gov

2 5

Re: Kernel config file for kernel used in rescue mode?
by David Thompson 29 Apr '05

29 Apr '05

Matt Domsch wrote: >On i386 CDs <= RHEL3 it's the -BOOT config file. >= RHEL4 it's the >2.6.9-i686-config I think (not SMP). It's the i586 (not SMP) kernel/boot file for anaconda. -- Dave Thompson ( thomas(at)cs(dot)wisc(dot)edu ) Associate Researcher Department of Computer Science University of Wisconsin-Madison http://www.cs.wisc.edu/(twiddle)thomas 1210 West Dayton Street Madison, WI 53706-1685 --

1 0

Kernel config file for kernel used in rescue mode?
by Mike Dickson 29 Apr '05

29 Apr '05

I’d like to be able to regenate a kernel like the one used when booting the rescue CD but can’t find the config file used to do so. Can anyone help point me in the right direction for this? Thanks in advance. Mike HYPERLINK "mailto:Mike.dickson@rlx.com"Mike.dickson(a)rlx.com -- No virus found in this outgoing message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.0 - Release Date: 4/29/2005

2 1

Device Parameter in Kickstart File
by Ronald Reed 28 Apr '05

28 Apr '05

How does anaconda handle the device parameter in a kickstart file? I have a kickstart file with 5 device lines in it, but anaconda only acts like it is seeing the first one and skipping the rest. -- =========================== Ron Reed RedHat Certified Engineer SGP Computer Department Manager Unix Systems Administrator ARM SGP CART Site (580)388-4053 ron.reed(a)arm.gov

1 0

Kickstart in Rescue Mode?
by Brian Vitiello 27 Apr '05

27 Apr '05

Does anyone know if it is possible to use a kickstart file in Rescue Mode. If so, do you know how to do this or where I can look to find this out? Basically, I need to boot up in rescue mode and get through all those screens to select rescue image locations, NFS server name, DHCP or static config. I need to be able to boot into rescue mode and just run a few commands but it needs to be entirely automated/without any user interaction. Thanks!

2 1

PATCH - Optimization to isomd5sum media check
by Dustin kirkland 26 Apr '05

26 Apr '05

Hi- I'm submitting a patch that provides an optimization to the isomd5sum media check functionality. Presently, the algorithm calculates a single md5sum after reading the entire disc, and reports a "PASS" or "FAIL" at that time. When there is corrupted data on the disc, a user will not know until the entire disc has been read--which can take quite a while. It would be optimal for the algorithm to exit the checking and report a failure as soon after reading the corrupted section as possible. I've accomplished this by adding two extra fields to the application data header of the ISO image. The ISO 9660 spec allows for a ~512-character string, of which were 224 characters are used. I've added two fields, "FRAGMENT SUMS" and "FRAGMENT COUNT", after which 326 characters are used. The "FRAGMENT COUNT" is an integer representing the number of fragments in which to break the entire ISO. The "FRAGMENT SUMS" is an ascii string of concatenated leading characters of md5sums of subsequent fragments. I've #define'd the length of FRAGMENT SUMS to 60 characters, as this number is evenly divisible by (2, 3, 4, 5, 6, 10, 12, 15, 20, 30). I've also #define'd FRAGMENT COUNT to be 20, which means that ISO will be broken into 20 fragments, each fragment represented by a 3-character string (the first 3 characters of the fragment's md5sum) in FRAGMENT SUMS. Looking at an example, the application data header used to look like: ISO MD5SUM = 1fca2222a947399d5b4a4226629be254;SKIPSECTORS = 15;RHLISOSTATUS=1;THIS IS NOT THE SAME AS RUNNING MD5SUM ON THIS ISO!! It now looks like: ISO MD5SUM = 1fca2222a947399d5b4a4226629be254;SKIPSECTORS = 15;RHLISOSTATUS=0;FRAGMENT SUMS = 67bd3263cccfda2d851a493b9a238e8877c45bca78e276e94437dd9f26e4;FRAGMENT COUNT= 20;THIS IS NOT THE SAME AS RUNNING MD5SUM ON THIS ISO!! The string [67bd3263cccfda2d851a493b9a238e8877c45bca78e276e94437dd9f26e4] is actually a concatenation of: md5(frag(0))[0-2] = 67b md5(frag(1))[0-2] = d32 md5(frag(2))[0-2] = 63c md5(frag(3))[0-2] = ccf md5(frag(4))[0-2] = da2 md5(frag(5))[0-2] = d85 md5(frag(6))[0-2] = 1a4 md5(frag(7))[0-2] = 93b md5(frag(8))[0-2] = 9a2 md5(frag(9))[0-2] = 38e md5(frag(10))[0-2] = 887 md5(frag(11))[0-2] = 7c4 md5(frag(12))[0-2] = 5bc md5(frag(13))[0-2] = a78 md5(frag(14))[0-2] = e27 md5(frag(15))[0-2] = 6e9 md5(frag(16))[0-2] = 443 md5(frag(17))[0-2] = 7dd md5(frag(18))[0-2] = 9f2 md5(frag(19))[0-2] = 6e4 Thus, if any one of these prescribed strings does not match what is calculated, the checking exits immediately, as a corruption has been detected. In the cases where the error is near the beginning of the disc, the time to identify the problem is greatly reduced. Note that there is a chance of collision. When the disc is broken into 20 fragments, each fragment is represented by a 3-character string, yielding a collision 1 in 16^3=4096 cases. Such a collision only means that the user will not benefit from the optimization, as the entire disc will be checked by the last operation just like it has always been done and the error will be detected at that time. One can reduce the chances of collision by decreasing the number of fragments, or expanding the length of the FRAGMENT SUMS string. This optimization occurs at virtually no performance cost. Media checking is overwhelmingly I/O bound, reading from disc. The additional calculations of smaller sums adds little additional processing time, as most of that time was being spent waiting on new data from disc. Please consider for inclusion in Anaconda, and I'll be glad to answer any questions or address concerns. I think the patch is pretty simple, and isolated to two files. Thanks! Dustin Kirkland

5 11

Re: Re: Ask for help about customizing Fedora Core 1, Thanks in advance!:-)
by 王秀文 25 Apr '05

25 Apr '05

David, Thanks for your reply.^-^ And now i got another problem.When i run the command below. ./splittree.py --arch=i386 --total-discs=6 --bin-discs=3 --src-discs=3 --release-string="Fedora Core 1" --pkgorderfile=/root/pkgfile --distdir=/home/i386 --srcdir=/home/SRPMS --productpath=Fedora what i got is as follows, First package on disc1:a.rpm Last package on disc1:b.rpm i386-disc1 size:617M First package on disc2:c.rpm Last package on disc2:d.rpm i386-disc2 size:638M First package on disc3:e.rpm Last package on disc3:f.rpm i386-disc3 size:629M No more discs to put packages,overflowing on disc2 First package on disc2:e.rpm Last package on disc2:g.rpm i386-disc2 size:640M First package on disc3:h.rpm Last package on disc3:i.rpm i386-disc3 size:639M ... ... ... ... What is the rule of overflowing,will this action still keep the dependency of packages. Thanks! mosquite

1 0

Re: Re: Ask for help about customizing Fedora Core 1, Thanks in advance!:-)
by 王秀文 25 Apr '05

25 Apr '05

Greg Morgan, Thanks very much for your reply! In my distribution,I don't want to include any sources.so directory "/home/SRPMS" does not exist.when i run the command splittree.py below,I got Error messages. If there is a way to designate the situation of including no source.For example using the option --src-discs=0 and something like that. ./splittree.py --arch=i386 --total-discs=6 --bin-discs=3 --src-discs=3 --release-string="Fedora Core 1" --pkgorderfile=/root/pkgfile --distdir=/home/i386 --srcdir=/home/SRPMS --productpath=Fedora Traceback (most recent call last): File "./splittree.py", line 455, in ? logfile = timber.main() File "./splittree.py", line 381, in main self.splitSRPMS() File "./splittree.py", line 350, in splitSRPMS for srpm in os.listdir("%s" % self.src_dir): OSError: [Errno 2] No such file or directory: '/root/SRPMS' Thanks a lot.^-^ mosquite

1 0

Ask for help about customizing Fedora Core 1,Thanks in advance!:-)
by 王秀文 25 Apr '05

25 Apr '05

Anaconda-devel-list, hello,everyone. What i want to do is to cut out the original three cds into one cd containing only our own necessary rpm packages.Because in every cd,there are different TRANS.TBL fiels such as $DISK[123]/TRANS.TBL $DISK[123]/Fedora/TRANS.TBL $DISK[123]/Fedora/RPMS/TRANS.TBL $DISK[1]/Fedora/base/TRASN.TBL I wonder if these files are created automatically by running a particular command or i must manually edit these files to ensure the consistency. Anyway,how to create one installing cd from the original three cds?or if i just want to cut out into two cds. What steps should i take?Is there any document on this subject. Please help,Thanks a lot lot lot lot! mosquite

3 2

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Anaconda-devel April 2005