Bugs item #987389, was opened at 2004-07-08 15:58 Message generated for change (Comment added) made by ejort You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=376685&aid=987389&group_id=22866
Category: JBossServer Group: v3.2 >Status: Open >Resolution: None >Priority: 5 Submitted By: wtam (wtam) Assigned to: Adrian Brock (ejort) Summary: Deadlock in AbstractInstanceCache.tryToPassivate() Initial Comment: The background thread that passivates an aged bean is deadlocking with the EntityLockInterceptor also trying to passivate the same bean. See the attached thread dump. 11:24:15,822 INFO [ServerInfo] Java version: 1.4.2_02,Sun Microsystems Inc. 11:24:15,823 INFO [ServerInfo] Java VM: Java HotSpot (TM) Server VM 1.4.2_02-b03,Sun Microsystems Inc. 11:24:15,823 INFO [ServerInfo] OS-System: Linux 2.4.20-24.7smp,i386 ---------------------------------------------------------------------- >Comment By: Adrian Brock (ejort) Date: 2004-09-20 19:37 Message: Logged In: YES user_id=9459 My previous comment on this problem is incorrect. I updated the wrong bug report. ---------------------------------------------------------------------- Comment By: Adrian Brock (ejort) Date: 2004-07-20 01:59 Message: Logged In: YES user_id=9459 Fixed for 3.2.6RC2 ---------------------------------------------------------------------- Comment By: Scott M Stark (starksm) Date: 2004-07-19 22:15 Message: Logged In: YES user_id=175228 Ok, don't know why I could not match up the source but the attached file looks correct. The problem is inconsistent acquisition order of the two locks, m_cacheLock and cache id BeanLock. AbstractInstanceCache.get -> synchronizes(m_cacheLock) -> CachePolicy.insert() ... -> LRUEnterpriseContextCachePolicy.ageOut -> AbstractInstanceCache.release -> AbstractInstanceCache.tryToPassivate -> BeanLock.sync() + deadlocked waiting for BeanLock LRUEnterpriseContextCachePolicy$OveragerTask.run -> AbstractInstanceCache.tryToPassivate -> BeanLock.sync() -> synchronizes(m_cacheLock) + deadlocked waiting for m_cacheLock ---------------------------------------------------------------------- Comment By: wtam (wtam) Date: 2004-07-19 18:51 Message: Logged In: YES user_id=1079596 We have double checked that the source code we have (which matches the stack trace) is identical to a fresh download of 3.2.3 from SourceForge. I've attached a copy of LRUEnterpriseContextCachePolicy.java which is revision 1.18.2.6. We believe this is the one shipped as JBoss 3.2.3. If this is not the right version, could you please identify which version was shipped as part of JBoss 3.2.3. FWIW, the following version was printed on the server log. 13:45:53,326 INFO [Server] Starting JBoss (MX MicroKernel)... 13:45:53,326 INFO [Server] Release ID: JBoss [WonderLand] 3.2.3 (build: CVSTag= JBoss_3_2_3 date=200311301445) ---------------------------------------------------------------------- Comment By: wtam (wtam) Date: 2004-07-19 18:49 Message: Logged In: YES user_id=1079596 We have double checked that the source code we have (which matches the stack trace) is identical to a fresh download of 3.2.3 from SourceForge. I've attached a copy of LRUEnterpriseContextCachePolicy.java which is revision 1.18.2.6. We believe this is the one shipped as JBoss 3.2.3. If this is not the right version, could you please identify which version was shipped as part of JBoss 3.2.3. FWIW, the following version was printed on the server log. 13:45:53,326 INFO [Server] Starting JBoss (MX MicroKernel)... 13:45:53,326 INFO [Server] Release ID: JBoss [WonderLand] 3.2.3 (build: CVSTag= JBoss_3_2_3 date=200311301445) ---------------------------------------------------------------------- Comment By: Scott M Stark (starksm) Date: 2004-07-19 04:37 Message: Logged In: YES user_id=175228 This is a conflict between the age and size based cache eviction policies, however, this has been fixed as of 3.2.2 and the thread dump you have for the LRUEnterpriseContextCachePolicy does not match that version. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=376685&aid=987389&group_id=22866 ------------------------------------------------------- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php _______________________________________________ JBoss-Development mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-development
