Comments on: How to deal with concurrent mode failures in the Hotspot JVM http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/ Been there, Done that, Wanna Remember! Sat, 05 May 2012 20:33:01 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: mrkips http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/comment-page-1/#comment-403 mrkips Wed, 16 Jun 2010 19:18:09 +0000 http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/#comment-403 @Sandor: Yes, you must try to avoid even a single occurrence of concurrent mode failure. Using iCMS with JRE 5 does help, but you should ensure you have sufficient processor time available. G1 replaced CMS from JRE 7 onwards and that's supposed to be preferable. @Sandor: Yes, you must try to avoid even a single occurrence of concurrent mode failure. Using iCMS with JRE 5 does help, but you should ensure you have sufficient processor time available. G1 replaced CMS from JRE 7 onwards and that’s supposed to be preferable.

]]> By: Sandor Vass http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/comment-page-1/#comment-402 Sandor Vass Tue, 15 Jun 2010 10:55:59 +0000 http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/#comment-402 Hi, Even the 1-2 concurrent mode failure is unacceptable (stop-the-world collection for a minute). I know this post is pretty old, but in these cases you might try out the incremental CMS, so the GC will be executing (almost) all the time in the background. This should work fine if you have 1-2 CPUs in your system (GC is not superseeded by app threads) ... and use more recent JRE version (1.6 G1 collector?) Hi,
Even the 1-2 concurrent mode failure is unacceptable (stop-the-world collection for a minute).

I know this post is pretty old, but in these cases you might try out
the incremental CMS, so the GC will be executing (almost) all the time in the background. This should work fine if you have 1-2 CPUs in your system (GC is not superseeded by app threads)

… and use more recent JRE version (1.6 G1 collector?)

]]> By: Deepak Devadathan http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/comment-page-1/#comment-393 Deepak Devadathan Wed, 10 Mar 2010 19:40:18 +0000 http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/#comment-393 excellent article! one question, what kind of gc occurs in the young region, does it stop the world? Did you perform the tuning a test env or production , because this results would probably have come after trial and error method. excellent article!
one question, what kind of gc occurs in the young region, does it stop the world?

Did you perform the tuning a test env or production , because this results would probably have come after trial and error method.

]]> By: WebLogic – IP Multicast : A primer :Mr.KIPS http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/comment-page-1/#comment-38 WebLogic – IP Multicast : A primer :Mr.KIPS Fri, 25 Sep 2009 09:02:10 +0000 http://cybergav.in/2009/09/22/how-to-deal-with-concurrent-mode-failures-in-the-hotspot-jvm/#comment-38 [...] SOLUTION: The JVMs hosting the WebLogic servers were tuned to minimize stop-the-world GC pauses to allow the servers to multicast properly. For the specific GC problem I encountered, you may refer to the tuning details here. [...] [...] SOLUTION: The JVMs hosting the WebLogic servers were tuned to minimize stop-the-world GC pauses to allow the servers to multicast properly. For the specific GC problem I encountered, you may refer to the tuning details here. [...]

]]>