2010/7/2 Steven Dake <sd...@redhat.com>:
> Thank you for the detailed bug report.
>
> Would you mind also posting a corosync-fplay output?

There was no output:
----8<--------8<--------8<--------8<--------8<--------8<--------8<----
[r...@pm01 20100701-memo]# corosync-fplay
failed to open /var/lib/corosync/fdata: No such file or directory
[r...@pm01 20100701-memo]# ls -l /var/lib/corosync
total 4
-rwx------ 1 root root 8 Jul  1 18:12 ringid_192.168.1.1
[r...@pm01 20100701-memo]#
----8<--------8<--------8<--------8<--------8<--------8<--------8<----


>
> There was mention that the segv occured again.  Was it during startup, or
> later during runtime when pacemaker forked a process?

It was during startup.
Here is more precise steps when I've got this hang:

 - I have 'chkconfig corosync on'
 - reboot both nodes at the same time
 - waiting for pacemaker invocation by watching crm_mon - but only one
node became online and the other node was OFFLINE for ever.
 - ps status at that time looks like this:
----8<--------8<--------8<--------8<--------8<--------8<----
[r...@pm01 20100701-memo]# ps axjf
(...)
    1  2664  2664  2664 ?           -1 Ssl      0   0:00 corosync
 2664  2670  2670  2670 ?           -1 SLs      0   0:00  \_ /usr/lib64/heartbea

 2664  2671  2664  2664 ?           -1 S      101   0:01  \_ /usr/lib64/heartbea

 2664  2672  2664  2664 ?           -1 S        0   0:00  \_ /usr/lib64/heartbea

 2664  2673  2664  2664 ?           -1 S      101   0:00  \_ /usr/lib64/heartbea

 2664  2674  2664  2664 ?           -1 S        0   0:00  \_ corosync
 2664  2675  2664  2664 ?           -1 S        0   0:00  \_ corosync
----8<--------8<--------8<--------8<--------8<--------8<----
 - took the core by 'gcore 2674', which I attached the stack trace in
the previous mail.

Side notes that may affect to the reproducibility:
 - It seems only happen on boot time (as above). When I run `service
corosync start' from the shell, I've never seen this yet.
 - I'm using rsyslog and the invocation order in init.d is corosync
first (S20corosync) and rsyslog later (S26rsyslog) with other 8
services between them.

Hope it helps.
-- 
Keisuke MORI
_______________________________________________
Openais mailing list
Openais@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/openais

Reply via email to