Red Hat Bugzilla – Bug 507411
Install fails at "finding storage devices"
Last modified: 2010-01-20 13:49:31 EST
Created attachment 348945 [details]
anaconda bug report
Description of problem:
Install fails at "finding storage devices". Last week I tried the install with the install DVD when I ran into the anaconda bug, but I had now way to get the report out of the install environment. Now I've tried the install with the Live CD and "Install to Hard Drive" with the same result. I've attached the anaconda bug report.
Version-Release number of selected component (if applicable):
Fedora 11 x86_64 DVD & Fedora 11 x86_64 Live KDE CD
Start install DVD or run "Install to Hard Drive" from live CD.
See attached anaconda bug report
This system has several software raids that were created with the Fedora 9 Installer. Here is the /proc/partitions output:
[liveuser@localhost ~]$ more /proc/partitions
major minor #blocks name
7 0 4 loop0
7 1 8 loop1
7 2 700160 loop2
7 3 3145728 loop3
7 4 524288 loop4
8 0 244198584 sda
8 1 20482843 sda1
8 2 92156872 sda2
8 3 20482875 sda3
8 4 1 sda4
8 5 20482843 sda5
8 6 90590503 sda6
8 16 244198584 sdb
8 17 20482843 sdb1
8 18 92156872 sdb2
8 19 20482875 sdb3
8 20 1 sdb4
8 21 20482843 sdb5
8 22 90590503 sdb6
253 0 3145728 dm-0
253 1 3145728 dm-1
9 0 20482752 md0
9 3 92156800 md3
9 1 20482752 md1
259 0 20466778 md1p1
9 2 20482752 md2
9 4 181180416 md4
The only other oddity in this machine is that the motherboard has integrated GeForce 8200 graphics, but I'm using a PCIE Radeon HD 3450 card.
I have the same problem. It may be related to having software raided devices as all my partitions are raid-1.
major minor #blocks name
8 0 312571224 sda
8 1 987966 sda1
8 2 3911827 sda2
8 4 1 sda4
8 5 19543041 sda5
8 6 19543041 sda6
8 7 19543041 sda7
8 8 19543041 sda8
8 9 229496526 sda9
8 16 312571224 sdb
8 17 987966 sdb1
8 18 3911827 sdb2
8 20 1 sdb4
8 21 19543041 sdb5
8 22 19543041 sdb6
8 23 19543041 sdb7
8 24 19543041 sdb8
8 25 229496526 sdb9
9 4 19542976 md4
9 1 3911744 md1
9 2 19542976 md2
9 5 19542976 md5
9 3 19542976 md3
9 0 987840 md0
9 6 229496448 md6
253 0 20971520 dm-0
253 1 20971520 dm-1
253 2 41943040 dm-2
253 3 20971520 dm-3
253 4 20971520 dm-4
253 5 20971520 dm-5
253 6 20971520 dm-6
Created attachment 351354 [details]
logfile that were in the tmp directory
Seems that anaconda wasn't quite dead, Ctrl-Alt-F2 got me a terminal and I mounted a thumbdrive and copied over the log files that were in tmp. Is there anything else that might be useful?
(In reply to comment #2)
> Created an attachment (id=351354) [details]
> logfile that were in the tmp directory
> Seems that anaconda wasn't quite dead, Ctrl-Alt-F2 got me a terminal and I
> mounted a thumbdrive and copied over the log files that were in tmp. Is there
> anything else that might be useful?
I see no evidence of a crash anywhere in the logs you provided. Can you describe the behavior that made you think there was a crash?
Please attach logs individually as plain text attachments in the future.
(In reply to comment #0)
> Created an attachment (id=348945) [details]
> anaconda bug report
> Description of problem:
> Install fails at "finding storage devices". Last week I tried the install with
> the install DVD when I ran into the anaconda bug, but I had now way to get the
> report out of the install environment. Now I've tried the install with the Live
> CD and "Install to Hard Drive" with the same result. I've attached the anaconda
> bug report.
Fedora 11 does not support the use of partitioned md devices. We are working on at least not crashing when we find them for F12.
Anaconda didn't crash, it just stopped progressing at the search, leaving up the message. What do you mean by partitioned md device? I have a raid1 partition containing a lvm volume, and various other raid1 partitions, is that the problem? And if so, why has the ability to deal with such a configuration gone away? Seems like a regression to me.
As to the attachment, bugzilla didn't let me post it as a text file. Maybe I needed to change the extension?
(In reply to comment #5)
> Hi Dave,
> Anaconda didn't crash, it just stopped progressing at the search, leaving up
How long did it sit before you decided it was stuck?
> the message. What do you mean by partitioned md device? I have a raid1
It is an md device with a partition table instead of a filesystem, but you don't have one -- the original reporter does/did.
> As to the attachment, bugzilla didn't let me post it as a text file. Maybe I
> needed to change the extension?
You created a compressed tar archive of the entire /tmp directory instead of just attaching /tmp/storage.log, /tmp/anaconda.log, and /tmp/program.log. What you did is certainly better than no logs at all, but it's troublesome to download/uncompress/extract as compared to being able to view the logs directly from the bug.
I compressed it *because* bugzilla wouldn't let me attach the log files. But let's not argue that small point. As to the time period, I tried the last install just to try getting more information. On my first attempt I waited about five minutes, time for a cup of coffee.
One of the virtues of Fedora as opposed to Ubuntu was it's transparent handling of md devices during install, the normal Ubuntu install disk can't handle that. So I would sure like to see this problem fixed.
Do you remember what bugzilla told you when not allowing you to attach log files? (I'm not interested in arguing -- I'd like to resolve the problem if I can.)
If you can, try the install again. This time, when you're pretty sure it is stuck (give it a few minutes), go to tty2 and run this command:
killall -USR2 anaconda
Then, attach the file /tmp/anacdump.txt to this bug report.