Hi,

On Wed, Jul 21, 2010 at 08:55:54AM -0300, mike wrote:
> in my ha logs I have the entries that appear several times a night. Now 
> I know in a previous post I was told these were indicative of resource 
> contention. These clusters that are seeing these messages are on a zVM 
> LPAR so they share CPU, memory and so on. Previously when we saw these 
> errors, failovers occurred and sometimes things got so bad that 
> failovers didn't even work. After adding memory and CPU things settled 
> down but I still see occasional nights with these entries in the log 
> file. So how much should I concern myself with these entries? Last night 
> the clusters were fine - no failover, no mention of resource failing. Is 
> there anything I should do to mitigate these notifications, i.e. perhaps 
> increase warntime or deadtime? What do you recommend?

You can't tune heartbeat to avoid these warnings, but you should
probably set the warntime and deadtime to higher values to avoid
unneeded failovers. Don't know if it's possible to tune zVM to
make the LPARs less "lazy".

Thanks,

Dejan

> Jul 21 05:33:40 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for cpu limit was delayed 440 
> ms (> 50 ms) before being called (GSource: 0x8011cec0)
> Jul 21 05:33:40 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: info: 
> Gmain_timeout_dispatch: started at 4307813164 should have started at 
> 4307813120
> Jul 21 06:47:56 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for send local status took too 
> long to execute: 70 ms (> 50 ms) (GSource: 0x801217d0)
> Jul 21 07:42:34 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for send local status took too 
> long to execute: 60 ms (> 50 ms) (GSource: 0x801217d0)
> Jul 21 07:45:06 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for send local status was 
> delayed 1530 ms (> 1510 ms) before being called (GSource: 0x801217d0)
> Jul 21 07:45:06 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: info: 
> Gmain_timeout_dispatch: started at 4308601743 should have started at 
> 4308601590
> Jul 21 07:45:06 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for check for signals was 
> delayed 1760 ms (> 1510 ms) before being called (GSource: 0x80121aa0)
> Jul 21 07:45:06 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: info: 
> Gmain_timeout_dispatch: started at 4308601766 should have started at 
> 4308601590
> Jul 21 07:45:33 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for send local status took too 
> long to execute: 190 ms (> 50 ms) (GSource: 0x801217d0)
> Jul 21 08:44:32 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for send local status took too 
> long to execute: 200 ms (> 50 ms) (GSource: 0x801217d0)
> Jul 21 09:20:04 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for send local status took too 
> long to execute: 130 ms (> 50 ms) (GSource: 0x801217d0)
> Jul 21 10:47:41 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for cpu limit was delayed 380 
> ms (> 50 ms) before being called (GSource: 0x8011cec0)
> Jul 21 10:47:41 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: info: 
> Gmain_timeout_dispatch: started at 4309697215 should have started at 
> 4309697177
> Jul 21 11:00:03 APAUAT1A.intranet.mydomain.com heartbeat: [2763]: WARN: 
> Gmain_timeout_dispatch: Dispatch function for send local status took too 
> long to execute: 60 ms (> 50 ms) (GSource: 0x801217d0)
> 
> 
> _______________________________________________
> Linux-HA mailing list
> [email protected]
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to