tag:blogger.com,1999:blog-169571142208866968.post2674056014618001994..comments2023-03-19T01:23:20.425-07:00Comments on The HornetQ Team Blog: 8.2 million messages / second with SpecJMSAnonymoushttp://www.blogger.com/profile/09981498972343885451noreply@blogger.comBlogger8125tag:blogger.com,1999:blog-169571142208866968.post-46364620634017778962011-07-29T19:56:28.825-07:002011-07-29T19:56:28.825-07:0047K messages / second that's nothing for Horne...47K messages / second that's nothing for HornetQ. (especially on this hardware).<br /><br />I will have someone who knows the benchmark in more detail to provide more background. The run rules will also provide you some information.Clebert Suconichttps://www.blogger.com/profile/05972640198013851304noreply@blogger.comtag:blogger.com,1999:blog-169571142208866968.post-685918801389324422011-07-29T13:07:54.457-07:002011-07-29T13:07:54.457-07:00Thanks Clebert. In my previous comment I'd lis...Thanks Clebert. In my previous comment I'd listed a sum of the total messages produced in the test (which was around about 9.5 million I think), divided against the number of seconds used in the test (2000) and that came out as about 47k msg/sec. I wasn't questioning the "spec calculation" as such.<br /><br />Either way, my own understanding of the SPECjms tests is that the important number (because it gives a true side-by-side comparison of messaging platform performance) is the "score" of 491 for horizonal and 1317 for vertical.<br /><br />Thanks for taking the time to discuss.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-169571142208866968.post-27729092499693379572011-07-29T11:10:24.851-07:002011-07-29T11:10:24.851-07:00@andypiper: I made a mess with publish / unpublish...@andypiper: I made a mess with publish / unpublish your comment, and I ended up deleting it for a complete accident.<br /><br />You may repost it if you can... <br /><br />But as far as I remember you said that you looked at the total sum of messages consumed versus the time and the amount didn't match with the calculations made by spec.<br /><br />As far as I know the calculations are correct, and you would need to look at the full disclosure. I don't think that information is out there on the HTML.<br /><br />The calculation was done by the spec.org software.<br /><br />All the other results are showing equivalent figures. just navigate on the other publications. (the other HornetQ and ActiveMQ publications).Clebert Suconichttps://www.blogger.com/profile/05972640198013851304noreply@blogger.comtag:blogger.com,1999:blog-169571142208866968.post-63004401867691985702011-07-29T05:24:02.962-07:002011-07-29T05:24:02.962-07:00(and to clarify my last comment, I was looking at ...(and to clarify my last comment, I was looking at the horizontal results there, totalling the number of messages processed across the interactions)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-169571142208866968.post-76215456259983719962011-07-28T13:02:47.648-07:002011-07-28T13:02:47.648-07:00@adypiper: No, the statement is correct. This is 8...@adypiper: No, the statement is correct. This is 8.2 Million messages Per second.<br /><br />This is the total throughput (messages / second) on the server involving all the queues (topic subscriptions + jms queues).<br /><br />The graph here is just an img referencing the SpecJMS website directly. And if you look at the graph produced by the SUT, you will see messages per second on the graph:<br /><br />http://www.spec.org/jms2007/results/res2011q2/jms2007-20110614-00029.png<br /><br />The information was generated by the SUT which is a spec.org software. It was peer reviewed by the SPEC group and we would already have had lots of complains from the Spec.org if we made a bad mistake like that.<br /><br />If you look at the publication itself, you will have the entire submission open (that is, anyone can look at it).<br /><br />This is the total ammount of messages per second handled by the server, including all the queues (topic subscriptions and JMS queues).Clebert Suconichttps://www.blogger.com/profile/05972640198013851304noreply@blogger.comtag:blogger.com,1999:blog-169571142208866968.post-57906284659012799572011-07-28T11:34:30.702-07:002011-07-28T11:34:30.702-07:00"With this latest result HornetQ sustained a ..."With this latest result HornetQ sustained a load of about 8 million messages per second" - are you 100% certain that you are quoting a per second figure there, or is there a chance that you were talking about the total number of messages sent in the context of the test runs? I think the way this is phrased is somewhat confusing.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-169571142208866968.post-3446141242412734052011-07-21T11:43:29.779-07:002011-07-21T11:43:29.779-07:00Just awesome!Just awesome!Luishttps://www.blogger.com/profile/08104776054909570165noreply@blogger.comtag:blogger.com,1999:blog-169571142208866968.post-73679755407240359542011-07-19T08:46:24.124-07:002011-07-19T08:46:24.124-07:00Congratulations to the teams involved in this effo...Congratulations to the teams involved in this effort!Mark Littlehttps://www.blogger.com/profile/15072917010265365428noreply@blogger.com