Comments on: Spring Integration on dm Server http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/ The voice of SpringSource Sat, 20 Mar 2010 12:14:29 -0500 http://wordpress.org/?v=2.8.5 hourly 1 By: Sriram C http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/comment-page-1/#comment-167826 Sriram C Wed, 05 Aug 2009 21:48:16 +0000 http://blog.springsource.com/?p=1086#comment-167826 Could you please let us the probable date for the Spring Integration 2.0 release. We are very interested in the runtime, message-content-based, dynamic determination of "destinations" for different channele adapter like JMS , HTTP , Remoting , WS etc. Also the fact that we could do these using Spring EL is appealing. This will avoid us to hardcode all the destination URL statically in the application context XML file Thanks Sriram C Could you please let us the probable date for the Spring Integration 2.0 release. We are very interested in the runtime, message-content-based, dynamic determination of "destinations" for different channele adapter like JMS , HTTP , Remoting , WS etc. Also the fact that we could do these using Spring EL is appealing. This will avoid us to hardcode all the destination URL statically in the application context XML file

Thanks

Sriram C

]]> By: Rob Harrop http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/comment-page-1/#comment-167813 Rob Harrop Tue, 04 Aug 2009 06:35:48 +0000 http://blog.springsource.com/?p=1086#comment-167813 @Harshad, Thanks for reporting the strange performance problems you are seeing. I've raised a JIRA issue at https://issuetracker.springsource.com/browse/DMS-1260 for you you track the progress of the fix. Regards, Rob @Harshad,

Thanks for reporting the strange performance problems you are seeing. I've raised a JIRA issue at https://issuetracker.springsource.com/browse/DMS-1260 for you you track the progress of the fix.

Regards,

Rob

]]> By: Harshad Penkar http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/comment-page-1/#comment-167810 Harshad Penkar Mon, 03 Aug 2009 22:08:51 +0000 http://blog.springsource.com/?p=1086#comment-167810 The example demonstrates a great way to decouple your systems in SI unaware (or aware) manner. I was able to get it to run without any problems. I further extended the example to have multiple TownCrier's sending many announcements to its Citizens i.e. I had 100 TownCriers each sending 1000 announcements i.e. spawned 100 threads each invoking town.cry() 1000 times (just trying to measure the round-trip of the call invocation -- ie. wrapped town.cry() to print the time it takes to execute. I had consumer.onCry() implemented as an empty method. The following are my observations -- 1) When I run from command line (using java directly) it doesn't take more than 10% of my CPU (its a 4 Processor linux box) and the round-trip is not more than 0 or 1ms. 2) The same test when running under Spring DM 2.0M3 on the same machine takes more than 50-60% of CPU and the round-trip time is approx 8-10ms. Are there any performance results that are comparable to my findings (why SI under OSGi is so CPU intensive)? Thanks, -Harshad The example demonstrates a great way to decouple your systems in SI unaware (or aware) manner. I was able to get it to run without any problems. I further extended the example to have multiple TownCrier's sending many announcements to its Citizens i.e. I had 100 TownCriers each sending 1000 announcements i.e. spawned 100 threads each invoking town.cry() 1000 times (just trying to measure the round-trip of the call invocation — ie. wrapped town.cry() to print the time it takes to execute. I had consumer.onCry() implemented as an empty method. The following are my observations –
1) When I run from command line (using java directly) it doesn't take more than 10% of my CPU (its a 4 Processor linux box) and the round-trip is not more than 0 or 1ms.
2) The same test when running under Spring DM 2.0M3 on the same machine takes more than 50-60% of CPU and the round-trip time is approx 8-10ms.

Are there any performance results that are comparable to my findings (why SI under OSGi is so CPU intensive)?

Thanks,

-Harshad

]]> By: Spring Integration 1.0.3 Samples: just add OSGi | SpringSource Team Blog http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/comment-page-1/#comment-167723 Spring Integration 1.0.3 Samples: just add OSGi | SpringSource Team Blog Tue, 28 Jul 2009 15:12:19 +0000 http://blog.springsource.com/?p=1086#comment-167723 [...] more ideas and a more sophisticated example you might also enjoy reading this from Spring Integration team member Iwein [...] [...] more ideas and a more sophisticated example you might also enjoy reading this from Spring Integration team member Iwein [...]

]]> By: links for 2009-06-09 « Xume linklog http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/comment-page-1/#comment-167246 links for 2009-06-09 « Xume linklog Tue, 09 Jun 2009 13:01:36 +0000 http://blog.springsource.com/?p=1086#comment-167246 [...] SpringSource Team Blog » Spring Integration on dm Server (tags: spring integration osgi) [...] [...] SpringSource Team Blog » Spring Integration on dm Server (tags: spring integration osgi) [...]

]]> By: Iwein Fuld http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/comment-page-1/#comment-147821 Iwein Fuld Sun, 01 Mar 2009 08:48:10 +0000 http://blog.springsource.com/?p=1086#comment-147821 Thank you for your comment William. I could remove the ambiguity and then disprove the opposite, but as it is not the point I'm trying to make I'll just rephrase it and avoid this argument that isn't very constructive either way. Do you have another alternative you would like to share perhaps? Thank you for your comment William.

I could remove the ambiguity and then disprove the opposite, but as it is not the point I'm trying to make I'll just rephrase it and avoid this argument that isn't very constructive either way.

Do you have another alternative you would like to share perhaps?

]]> By: William Louth http://blog.springsource.com/2009/02/27/spring-integration-on-dm-server/comment-page-1/#comment-147722 William Louth Sat, 28 Feb 2009 16:31:34 +0000 http://blog.springsource.com/?p=1086#comment-147722 "To evolve large applications that should run on a single jvm for optimal performance OSGi is the only viable alternative." Could you please explain the above statement which seems to have been pulled out of thin air and with not actual supporting arguments. - large? - single! jvm? - optimal performance? - only viable? William "To evolve large applications that should run on a single jvm for optimal performance OSGi is the only viable alternative."

Could you please explain the above statement which seems to have been pulled out of thin air and with not actual supporting arguments.

- large?
- single! jvm?
- optimal performance?
- only viable?

William

]]>