Here is a timeline distribution of the GC activity. The period is quite short (a little bit more that 3 minutes) so it is hard to deduce anything, but here we can see that there is a period of 30 seconds where the GC activity is more intense, between 45 and 75 seconds. But all full GC takes around 100ms, so I tend to think that the GC is not the culprit here, it is only a good indicator of the broker slowdown.
You can try to analyze your GC logs from your production with the GCHisto tool (can be downloaded here: http://pietrowski.info/wp-content/uploads/2009/06/GCHisto.tgz), maybe it can help in diagnosing this issue. Regards, Reynald On Tuesday, December 14, 2010 at 14:40 , robert.sl...@misys.com wrote: > > Here is the gc.log and activemq logs, once again taken from a run where I > allowed it to process, block and then continue processing: > > http://activemq.2283324.n4.nabble.com/file/n3087092/logs_verbose_gc.rar > logs_verbose_gc.rar > > -- > View this message in context: > http://activemq.2283324.n4.nabble.com/Slow-throughput-after-several-hundred-messages-tp3082431p3087092.html > Sent from the ActiveMQ - User mailing list archive at Nabble.com. > > > >
