tag:blogger.com,1999:blog-4869268408030448256.post276509446585380828..comments2023-08-12T16:03:51.559+02:00Comments on Erwin Vervaet: MQ's MQRFH2 Header and JMSklr8http://www.blogger.com/profile/13417996252591392996noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-4869268408030448256.post-9968447983514190162010-04-14T16:43:26.899+02:002010-04-14T16:43:26.899+02:00Just for perspective, by "high-volume" I...Just for perspective, by "high-volume" I'm referring to a queue where the volume is so high that it routinely backs up to a depth of 400,000 messages. Its at depths over 250,000 that the issue really raises its head - under that depth the performance is definitely better and less of an issue.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-4869268408030448256.post-74814600700150605782010-04-13T20:13:27.583+02:002010-04-13T20:13:27.583+02:00We also have quite high volume queues that use mes...We also have quite high volume queues that use messages with MQRFH2 headers, but haven't noticed any considerable performance overhead. Having said that, we also haven't done any side by side testing to see whether or not we could gain something by using plain MQSTR headers. So at least to me the MQRFH2 overhead is not a familiar gotcha.klr8https://www.blogger.com/profile/13417996252591392996noreply@blogger.comtag:blogger.com,1999:blog-4869268408030448256.post-15170836195680043082010-04-12T21:59:21.487+02:002010-04-12T21:59:21.487+02:00I'm currently diagnosing an issue with a high-...I'm currently diagnosing an issue with a high-volume application that seems to reflect that MQRFH2 format messages incur a significant performance hit on the server side at high queue depths, in contrast to MQSTR format messages. Does this sound like a familiar "gotcha" ?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-4869268408030448256.post-57703286143756697462010-03-06T07:49:14.947+01:002010-03-06T07:49:14.947+01:00Thanks for letting me know!
We're indeed still...Thanks for letting me know!<br />We're indeed still on WebSphere MQ 6.0.klr8https://www.blogger.com/profile/13417996252591392996noreply@blogger.comtag:blogger.com,1999:blog-4869268408030448256.post-58205406067524524592010-03-05T21:33:47.386+01:002010-03-05T21:33:47.386+01:00Check out the latest version of WebSphere MQ. All...Check out the latest version of WebSphere MQ. All of the message properties are available as JMS properties now. Also, you can have a JMS and a non-JMs program both subscribing to the same topic and both successfully receive messages that were put with RFH2 headers. This has all become largely transparent in v7.0.1.Unknownhttps://www.blogger.com/profile/01812662034053959894noreply@blogger.com