[OpenAFS] OpenAFS 1.8.4 Linux kernel BUG

Benjamin Kaduk kaduk@mit.edu
Thu, 2 Apr 2020 19:50:25 -0700


The "disk cache read error" makes me wonder how the hardware is doing; are
you in a position to run (e.g.) SMART tests?

-Ben

On Wed, Apr 01, 2020 at 01:53:04PM +0100, Chris Cooke wrote:
> Hi,
> 
> A machine of ours recently became unresponsive - these are the messages reported by journalctl for the time it happened:
> 
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 320036 off 25602900/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 319975 off 25598020/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 231566 off 18525300/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 320007 off 25600580/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 239740 off 19179220/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 229899 off 18391940/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 319838 off 25587060/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 653166 off 52253300/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 653166 off 52253300/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: failed to store file (5/0)
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: afs: disk cache read error in CacheItems slot 653166 off 52253300/113096500 code -5/80
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: openafs: afs_InvalidateAllSegments tdc count
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: ------------[ cut here ]------------
> Mar 31 22:51:07 lute.inf.ed.ac.uk kernel: kernel BUG at /builddir/build/BUILD/openafs-1.8.4/src/libafs/MODLOAD-3.10.0-1062.7.1.el7.x86_64-SP/afs_segments.c:556!
> 
> Nothing further was logged until a reboot nearly an hour later.
> The machine runs Scientific Linux 7.6, and here's the output of "rpm -q kernel openafs" :
> 
> kernel-3.10.0-1062.7.1.el7.x86_64
> openafs-1.8.4-1.el7.x86_64
> 
> Chris Cooke.

> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.