On Wed, 28 May 2025, Robert Elz wrote:
Date: Wed, 28 May 2025 08:51:01 +0200 (CEST)
From: 6b...@6bone.informatik.uni-leipzig.de
Message-ID:
<264b8afa-54ed-2b70-c096-604ad5944...@6bone.informatik.uni-leipzig.de>
| We run a server with a large amount of disk space. The disk space for the
| server is provided via iSCSI. Unfortunately, the server is not very stable
| and crashes regularly. Attached is the last dump. Can anyone identify the
| cause of the problem?
| fmt=fmt@entry=0xffffffff81489de8 "%s: dup alloc ino=%ld on %s: mode %o/%o
gen %x/%x size %lx blocks %lx", ap=ap@entry=0xffff850ded754a98)
That suggests (strongly) a corrupted filesystem, but your message
doesn't include enough information to say which.
If you have the output from the printf that used
that format string, you'd have a lot more useful information
than a backtrace in this case (that's not helpful at all).
But somewhere you have a filesystem which (badly) needs the
attention of "fsck -f".
Nothing else (apart from not using that filesystem) will
avoid this happening again and again.
kre
Hello,
After every crash, all filesystems were marked as dirty. I performed an
fsck with -y on every reboot. The check took several hours each time.
However, the problem still occurs irregularly. Therefore, I still suspect
the iSCSI.
Thank you for your efforts
Regards
Uwe
