On Wed, 2012-07-25 at 08:05 -0700, ParkiratBagga wrote: > Hi, > > Today, I tried to replicate the same issue. Though, we were able to > replicate, but whatever steps we have done today were not done when the > issue came for the first time. > > Steps done: > > 1. Both the consumer and producer were running and the producer over-running > the consumer to achieve enqueue-capacity-threshold for 2-3 times. > > 2. Then we increased the queue size using resize tool provided with the > persistence store, and started running the same test as above.
I assume you stopped the broker first. The store cannot be resized on a running broker. Doing so will result in undefined broker and store behavior. You must stop the broker, then resize the store, then restart the broker and recover the messages. Note also that flow-to-disk messages are not recovered on the restart of the broker unless the messages are persistent (in which case they would be written to the store in any event). > > 3. We go the RHM_IORES_EMPTY error, and on running the store_chk, it said > there is no message in the queue but there were 2k messages present inside. > > 4. Moreover "qpid-config queues" was showing previous configuration of the > queue. > > I believe, somehow Qpid Server configuration got corrupted and we saw the > error. But, we did nothing like this the first time, when it came. We were > just running plain perf test, then how the Journal got corrupted? > Moreover this error is coming on each and every queue on that particular > broker. > > I was not able to replicate this in my local just by over-running the > consumers. Is it something related to OS or deployment/installation? > > > My Questions remain the same: > > What should I do to stop the broker error to propagate to client and why > does my connection options like retries and connectiondelay does not work > here? Size the store correctly for your test/application. Most store errors (RHM_*) you have described and seen during normal use are as a result of incorrect sizing or exhausting resources (disk space, memory or file handles). > > What configurations, should I use to prevent the same error coming? > > > > Regards, > Parkirat > > > > > > -- > View this message in context: > http://qpid.2158936.n2.nabble.com/Qpid-Journal-Error-RHM-IORES-EMPTY-tp7580022p7580149.html > Sent from the Apache Qpid users mailing list archive at Nabble.com. > > --------------------------------------------------------------------- > To unsubscribe, e-mail: users-unsubscr...@qpid.apache.org > For additional commands, e-mail: users-h...@qpid.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@qpid.apache.org For additional commands, e-mail: users-h...@qpid.apache.org
