[Debian-ha-maintainers] Bug#605025: Segmentation fault of corosync running without pacemaker

[Mehdi Dogguy]

tags 605025 + unreproducible
severity 605025 important
thanks

On  0, Mehdi Dogguy <mehdi at dogguy.org> wrote:
> 
> I tested on my Squeeze box and I wasn't able to reproduce the segfault.

I also tested in Lenny to make the test relevant:

root at lenny-chroot:/# apt-cache policy corosync
corosync:
  Installed: 1.2.1-1~bpo50+1
  Candidate: 1.2.1-1~bpo50+1
  Version table:
 *** 1.2.1-1~bpo50+1 0
          1 http://www.backports.org lenny-backports/main Packages
        100 /var/lib/dpkg/status
root at lenny-chroot:/# cat /etc/debian_version 
5.0.5
root at lenny-chroot:/# corosync
Dec 05 19:50:15 corosync [MAIN  ] Corosync Cluster Engine ('1.2.1'): started and ready to provide service.
Dec 05 19:50:15 corosync [MAIN  ] Corosync built-in features: nss
Dec 05 19:50:15 corosync [MAIN  ] Successfully read main configuration file '/etc/corosync/corosync.conf'.
Dec 05 19:50:15 corosync [MAIN  ] parse error in config: The consensus timeout parameter (1500 ms) must be atleast 1.2 * token (3600 ms).
Dec 05 19:50:15 corosync [MAIN  ] Corosync Cluster Engine exiting with status -9 at main.c:1423.
root at lenny-chroot:/# echo $?
247
root at lenny-chroot:/# sed -i "s/\(.*consensus:\).*$/\1 4000/g" /etc/corosync/corosync.conf
root at lenny-chroot:/# corosync
root at lenny-chroot:/# echo $?
0

No segfault here.

Same even after fixing the configuration file. (Also tested without
pacemaker in service section).

Thus, I'm lowering the severity of this bugreport and tagging it as
unreproducible.

If you still encounter the problem, please send us a copy of the
configuration files you're using and precise steps leading to the
segfault.

Regards,

-- 
Mehdi Dogguy