Qemu-Libgfapi: periodical shutdown of virtual machines.

Bug #1464278 reported by Igor
14
This bug affects 2 people
Affects Status Importance Assigned to Milestone
qemu (Ubuntu)
Invalid
High
Unassigned

Bug Description

Hello,

Would you so kind to help me with my problem concerning libgfapi.

My host operating system is Ubuntu 14.04 LTS, version of glusterfs is
3.6.2, and version of qemu is 2.0.0. We use libfgapi library to connect to glusterfs.

In our environment virtual machines sometimes go to power-off
state 'Powered Off'(Shutdown) with ERROR in /var/log/syslog like:

kernel: [5346607.988173] qemu-system-x86[29564]: segfault at 128 ip
00007f930a90b48c sp 00007f931296fd70 error 4 in
qemu-system-x86_64[7f930a57f000+4b1000]

Please help me to understand possible reasons of the qemu-system segfault.
Maybe glusterfs tuning is required or tuning of libvirt.

Do you have any ideas?...

Best regards,
Igor.

Revision history for this message
Serge Hallyn (serge-hallyn) wrote :

Could you please show exactly which versions of which packages yo uare using (and anything you hand-built)?

If all is using stock packages, then could you please, the next time this happens , do

apport-collect 1464278

Changed in qemu (Ubuntu):
status: New → Incomplete
Revision history for this message
Igor (qemu-glusterfs) wrote :

list all packages.

Revision history for this message
Igor (qemu-glusterfs) wrote :
Revision history for this message
Serge Hallyn (serge-hallyn) wrote :

Is there a crashfile under /var/crash? If so that should container a corefile which should help to track down what actually happened.

Revision history for this message
Igor (qemu-glusterfs) wrote :

>Is there a crashfile under /var/crash?
Yes.
Size if crashfile _usr_bin_qemu-system-x86_64.0.crash is 1,9G.

Revision history for this message
Serge Hallyn (serge-hallyn) wrote : Re: [Bug 1464278] Re: Qemu-Libgfapi: periodical shutdown of virtual machines.

Ok,

you can get the CoreFile from that using apport-unpack:

mkdir crashfiles; apport-unpack /var/crash/_usr_bin_qemu-system-x86_64.0.crash crashfiles

Then you should be able to use the 'where' command in gdb

gdb /usr/bin/qemu-system-x86_64 crashfiles/CoreDump
(gdb) where

to get more information. Please paste that information here.

Revision history for this message
Igor (qemu-glusterfs) wrote :

>Please paste that information here
>gdb /usr/bin/qemu-system-x86_64 crashfiles/CoreDump

-> gdb-out.txt

Revision history for this message
Serge Hallyn (serge-hallyn) wrote :

Sorry, did you type "where" into gdb and it simply died, or did you not type 'where'?

Changed in qemu (Ubuntu):
importance: Undecided → High
Revision history for this message
Igor (qemu-glusterfs) wrote :

Thank you for your reply.
Sorry,
gdb /usr/bin/qemu-system-x86_64 crashfiles/CoreDump
 with to use the 'where'.

Revision history for this message
Serge Hallyn (serge-hallyn) wrote :

Ok, so this:

#1 0x00007f369bd2b3b1 in glfs_io_async_cbk (ret=<optimized out>, frame=<optimized out>, data=0x7f369ee536c0) at glfs-fops.c:598
#2 0x00007f369badb66a in syncopctx_setfspid (pid=0x7f369ee536c0) at syncop.c:191

shows you where you could focus (syncop.c line 191) to debug this issue. Unfortunately, since gluster is in universe and this is a hand-built version of qemu, a launchpad bug against the archive package is not an appropriate place to debug. If a look at the source doesn't help, you might contact the package owner (monotek).

Changed in qemu (Ubuntu):
status: Incomplete → Invalid
Revision history for this message
Serge Hallyn (serge-hallyn) wrote :

See also

http://www.gluster.org/pipermail/gluster-devel/2015-July/046077.html

It might be worth following up to see whether Prasanna has made progress.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.