[Pacemaker] cib and attrd processes segfault

[Andrew Beekhof]

On Tue, Feb 16, 2010 at 4:36 PM, Alessandro Federico
<alessandro.federico at caspur.it> wrote:
> Hi all,
> we have just installed the latest versions of pacemaker/corosync software:
> cluster-glue-1.0.3-1.el5.x86_64
> cluster-glue-libs-1.0.3-1.el5.x86_64
> corosync-1.2.0-1.el5.x86_64
> corosynclib-1.2.0-1.el5.x86_64
> heartbeat-3.0.2-2.el5.x86_64
> heartbeat-libs-3.0.2-2.el5.x86_64
> pacemaker-1.0.7-4.el5.x86_64
> pacemaker-libs-1.0.7-4.el5.x86_64
> resource-agents-1.0.1-1.el5.x86_64
> on a Scientific Linux SL 5.4 box (kernel 2.6.18-128.7.1.el5).
> The configuration file corosync.conf is attached.
> The problem is that we get segfault of cib and attrd processes as soon
> as we start the corosync service (see the attached messages.log file).
> Can anybody help us, please?

Please
- enable coredumps (set "ulimit -c unlimited" at the top of the
corosync init file)
- use hb_report to create a support tarball covering the problem
- attach the tarball to a new bug:
http://developerbugs.linux-foundation.org/enter_bug.cgi?product=Pacemaker
Thats the minimum we'd need to be able to assist.