On 07/01/2010 09:26 PM, Keisuke MORI wrote: > 2010/7/2 Andrew Beekhof<[email protected]>: >> On Thu, Jul 1, 2010 at 3:09 PM, Keisuke MORI<[email protected]> >> wrote: >>> Here is the stack trace and the corosync.conf when I reproduce it with >>> corosync-1.2.6. >>> According to the core, fileno=10 looks broken, while filno=0,1,2,3 seems >>> sane. >> >> Any chance you could do some digging to figure out where fileno=10 is >> coming from? > > I'm trying to, but no idea yet. > I also checked the parent corosync process but 10 was not in use. > It could be just a garbage value. > >
In the glibc output there appears to be logging data. I have asked Andrew to make you a patch for pacemaker which removes log output from the "child" process inside the plugin to see if problem persists. I'm not sure when he will get to this, but if you want to give it a try, this data point would be helpful. If rsyslog is started first, does this problem persist? Why starting logging daemon after corosync? corosync depends on syslog if using syslog output. Does this problem persist if to_syslog is set to no? Regards -steve _______________________________________________ Openais mailing list [email protected] https://lists.linux-foundation.org/mailman/listinfo/openais
