Anaconda-devel September 2013

anaconda-devel@lists.stg.fedoraproject.org

12 participants
12 discussions

anaconda: reuse of btrfs subvolumes
by Gene Czarcinski 27 Sep '13

27 Sep '13

OK, this bugzilla report has not gotten much action -- https://bugzilla.redhat.com/show_bug.cgi?id=892747 so lets try this list. I would think that a common installation would involve two types of btrfs subvolumes: a new subvolume for root ("/") and one or more existing subvolumes for /home, etc. Anaconda can handle creation of new subvolumes nicely and adds them to /etc/fstab. However, for existing subvolumes, you need to specify "--useexisting" on the btrfs command which specifies the mount point or anaconda crashes saying that the subvolume already exists .. OK, that makes sense. But, if you then specify --useexisting, anaconda proceeds BUT the mount point is NOT added to /etc/fstab. This problem has existed since Fedora 18 and is still true in Fedora 20 alpha. How about if someone fixes this, explains why it is not necessary, or explains how to specify reuse of a btrfs mount point? Is this too much to ask? I do not like cross posting but I am going to post a separate copy of this message on the Fedora devel list. Gene

1 0

Re: F20 anaconda progress bar
by Kamil Paral 16 Sep '13

16 Sep '13

----- Original Message ----- > On 12/09/13 05:54, Richard Ryniker wrote: > > On my last installation (TC4) anaconda's blue progress bar stalled at > > > approximately 40% while thousands of packages were installed. I suspect > > > this is "Working as Designed". There is a continuing display of package > > > names as they are installed, as well as counts of installed packages and > > > total number of packages to be installed, therefore a user is not likely > > > to think the installation process is stuck just because there is no > > > movement of the progress bar. The little slide show below the progress > > > bar also continues. > > > This does, however, give an impression the installer is a bit crude: > > > during the longest part of the installation process, there is no movement > > > of the progress bar. > > > If someone does see better behavior of the anaconda progress indicator, > > > please post to say it is my experience that is wacky. > > I have observed the same as you. I suspect it only get s updated after each > stage of installation is completed. > I think there should be a progress bar for each stage, updating according to > the number of micro-stages/files/byt es/... (as appropriate) processed. From > memory, as the last install I did was for F19, there should be 4 bars > labelled appropriate ly . This way, not only would be people be aware of > what stage the installation was at, but they would have an indication of the > progress in the current stage, and what other s tages were yet to be > started. Obviously, i f 2 or more stage could be done in parallel, then > those progress ba rs would be updated concurrently. > More importantly, the progress bar ( if there is only to be one) should be > finer grained than up dating after each major stage. > Just my 2 pennies worth. > Cheers, > Gavi n Right, anaconda doesn't have a progress bar, but a "stage indicator". Unfortunately it is displayed as... a progress bar. I tried to explain the differences and propose some improvements to Martin Sivak, a former Anaconda developer, but he wouldn't listen. He doesn't work on Anaconda anymore, so maybe someone else might be more open to this dialogue. I suggest that you write into anaconda-devel mailing list [1], or/and create a bugzilla report with your suggestions. Kamil [1] https://www.redhat.com/mailman/listinfo/anaconda-devel-list

1 0

Re: rawhide boot.iso i386 - Could not boot
by poma 14 Sep '13

14 Sep '13

On 10.09.2013 10:38, poma wrote: > > $ checkisomd5 --verbose boot.iso > boot.iso: 58c742f8c6aa17da32e82c4ca541ec24 > Fragment sums: 2f42915afa918218c24a59a23e346a3f931376eafddb5f75443daecaf88f > Fragment count: 20 > Press [Esc] to abort check. > Checking: 100.0% > > The media check is complete, the result is: PASS. > > It is OK to use this media. > > boot.iso 09-Sep-2013 13:07 > https://bugzilla.redhat.com/show_bug.cgi?id=1006210 poma

1 1

"""cannot retrieve repository metadata (repomd.xml)"""
by Kun Huang 11 Sep '13

11 Sep '13

[image: Inline image 1] Now I have faced this problem in my customised iso. The iso is simple one from CentOS-6.4-x86_64-minimal.iso<http://mirrors.sohu.com/centos/6.4/isos/x86_64/CentOS-6.4-x86_64-minimal.iso>. In this iso, I created a new directory 'os' and move directory 'Packages' and 'repodata' into it. And passed a boot option repo=cdrom:///dev/sr0:/os to tell anaconda where the repo is, but result always is this warning window. I have tried a lot of hacking ways. Such as "repo.baseurl=['file:/mnt/source/os']", but it doesn't work. And at the same time, I went into tty2, I could manually find my resources in /mnt/source/os. BTW, my anaconda version is 13.21.195 Best Regards

1 0

Help with translations
by Gianluca Sforna 09 Sep '13

09 Sep '13

Hi, I am traslating Anaconda to Italian and having a bit of "fun" trying to find a good term for "rescan", used in 5 strings. I did not try the installer (yet) so a bit of context about where this is used would help. I think it is safe to assume it indicate some sort of analysis a user can/need to perform on disks but why the re- part? Additionally, using it as verb/noun does not help as well. Is it some technical term so I better leave it alone, or maybe I can translate it as analysis/analize to the same effect? thanks in advance Gianluca -- Gianluca Sforna http://morefedora.blogspot.com http://identi.ca/giallu - http://twitter.com/giallu

3 4

20TC5: not nice selected partition setup
by Felix Miata 09 Sep '13

09 Sep '13

1-left pane reproduces *5* items "Unknown Linux" with twisties to open up. I open up all 5 only to find each contains the exact same partitions but in different randomized (non) order. e.g. sda11, sda24, sda1, sda16, sda10 sda9, sda17, sda14, sda2 and so forth. 2-After I finally *find* the (pre-formatted, pre-labeled, empty EXT4) partition (in "Unknown", not "Unknown Linux") that is to be the / partition, I put / in the mount point box, whereupon EXT4 appears as the file system selected under mount point and label and desired capacity and device type. Looks good, so I click on Update Settings. At the bottom I see it complain I didn't do it right, as according to Fedora policy the / device must *always* be formatted by Anaconda. So I click the reformat box, then update settings, and I get another message I did something wrong. What happened? Anaconda cleared the previously entered / mount point. Why? As long as Fedora insists the / device must be reformatted, then Anaconda should be smart enough to check that box itself. At the least it should not clear prior entered entries unless a different device is selected before update settings gets accepted without error. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/

1 0

How to skip bootloader updating in kickstart?
by Alexander Todorov 06 Sep '13

06 Sep '13

Hi guys, what is the command to skip the bootloader update when performing upgrades via kickstart? I see no relevant options here (or I'm blind): https://fedoraproject.org/wiki/Anaconda/Kickstart#bootloader Thanks, Alex

3 5

Re: Bug 1000889 AttributeError: 'DeviceFormat' object has no attribute 'mount'
by Felix Miata 05 Sep '13

05 Sep '13

On 2013-09-05 02:17 (GMT) bugzilla(a)redhat.com composed: > https://bugzilla.redhat.com/show_bug.cgi?id=1000889 > --- Comment #30 from stan > If you haven't downloaded TC3 yet, don't waste your time. It still has I only downloaded squashfs.img, vmlinua and initrd.img, but that only duplicated comment 25 with comment 29. > anaconda 20.9.1. The fix for this problem is in 20.10.1. I guess > communication just got crossed. The new version is on koji, in fact there is a > 20.10.2 out there. Or I might just wait for it to get fixed. Is there a squashfs.img somewhere on http://alt.fedoraproject.org/pub/alt/ or elsewhere that has 20.10.2 in it? If I could have found one I'd already be at least attempting another HTTP TC3 installation. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/

2 1

boot options of anaconda
by MaShimiao 04 Sep '13

04 Sep '13

Hi all, I want to know all boot options supported by anaconda. How can I get them with a specified version? Thanks.

2 1

HD install from exploded tree
by Nuno Fernandes 04 Sep '13

04 Sep '13

Hi, I'm very interested in this feature. It was implemented in http://www.redhat.com/archives/anaconda-devel-list/2006-November/msg00050.h… but never make it to in-tree code. I'm in the process of patching anaconda-13.21.195-1.el6 to do that. Can anyone help a newbie in anaconda devel where in the code i should start looking and/or changing? Thanks, Nuno Fernandes

1 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Anaconda-devel September 2013