And this is a 32-bit Ubuntu:
$> uname -a Linux 2.6.32-33-generic-pae #72-Ubuntu SMP Fri Jul 29 22:06:29 UTC 2011 i686 GNU/Linux Probably, all goes better on 64-but systems. Irina. From: Irina Arkhipets [mailto:ia...@unipro.ru] Sent: Wednesday, February 13, 2013 10:25 AM To: dev@openmeetings.apache.org Subject: RE: OM stress testing Hi Sebastian, Just two additional comments from my side: 1. "-Xms512M -Xmx1024M" java keys were used for OM starting 2. All the 19 users in the room had their video windows opened Previously, when we run OM without "-Xmx" option specified, it died on the 14th user in the room approximately under the similar conditions. Regards, Irina. From: seba.wag...@gmail.com [mailto:seba.wag...@gmail.com] Sent: Wednesday, February 13, 2013 5:38 AM To: dev Subject: Re: OM stress testing What script did you use to startup the server? You should use the red5-highperf.sh and have a look at the values for Xmx (max memory) and Xms (min memory). Please include those runtime informations / configurations from your start script in the stress test report. I tried to summarize some links and documentation about performance tuning once: http://openmeetings.apache.org/JVMPerformanceTuning.html Se for example the link to the JVM settings: http://www.oracle.com/technetwork/java/javase/tech/vmoptions-jsp-140102.html #PerformanceTuning But Xmx and Xms is really the most important thing. There is also some param regard Heap and Thread, that are quite important for performance. I think by default the JVM will aquire 60 MB. Sebastian 2013/2/13 Stephen Cottham <stephen.cott...@robertbird.com.au> At a guess I'd say you need more then 2 gig of RAM based on the exception from the java heap space. Sent from my iPad On 12 Feb 2013, at 22:25, "Joseph Karwat" <jkar...@jellnet.com> wrote: Vasiliy: Terrific information and test. I look forward to seeing more of your results. It would be helpful to understand what caused the error. Would a more powerful computer handle the load or is there a physical limit of 20 regardless of environment. Let me know how I can help. Joseph Joseph Karwat | CEO | 415-462-0263 | <http://www.jellvideo.com/> JellVideo | <http://www.jellnet.com/> www.jellnet.com | Error! Filename not specified. Jell Networks, Inc. Enterprise Video Collaboration Solutions Click Here to Register for <http://www.jellvideo.com/> your JellVideo Collaboration Account Stephen Cottham Group IT Manager (Associate) Robert Bird Group Level 5, 333 Ann St Brisbane, Queensland, 4000, Australia Phone: +6173 319 <tel:%2B6173%20319%202777> 2777 (AUS) Phone: +44207 <tel:%2B44207%C2%A0633%202880> 633 2880 (UK) Fax: +6173 <tel:%2B6173%20319%202799> 319 2799 Mobile: +61400 <tel:%2B61400%20756%20963> 756 963 (AUS) Mobile: +447900 <tel:%2B447900%20918%20616> 918 616 (UK) Web: <http://www.robertbird.com/> www.robertbird.com <http://www.robertbird.com.au/> Error! Filename not specified. This email and any attachments are confidential and may contain legally privileged information or copyright material. Unless expressly stated, confidentiality and/or legal privilege is not intended to be waived by the sending of this email. The contents of this email, including any attachments, are intended solely for the use of the individual or entity to whom they are addressed. If you are not an intended recipient, please contact us immediately by return email and then delete both messages. You may not otherwise read, forward, copy, use or disclose this email or any attachments. Any views expressed in this email are those of the individual sender except where the sender expressly, and with authority, states otherwise. It is your responsibility to check any attachments for viruses or defects before opening or sending them on. None of the sender or its related entities accepts any liability for any consequential damage resulting from this email containing computer viruses. Disclaimer added by CodeTwo Exchange Rules www.codetwo.com From: Vasya [mailto:va...@unipro.ru] Sent: Tuesday, February 12, 2013 12:49 AM To: dev@openmeetings.apache.org Subject: OM stress testing Hello all! Today we have started stress testing of Openmeetings to detect reliability and collect statistics. We plan to run stress tests weekly (on Tuesdays) and send reports to dev@openmeetings.apache.org . Tested server configuration is: OS: Ubuntu 10.04.3 LTS . RAM 2 GB . CPU Intel(R) Celeron (R) E3300 @ 2.50GHz Tested scenario is: . Start OM conference . Start recording . Start screen sharing . Connect to conference while server is not crashed. Today results meet our expectations, please see a report below. Maximum number of the users in the system - 20 Maximum number of the users in one room of the type "Conference" with the screen sharing and recordings turned on - 19. All the users were logged from two Windows workstations. Result: After 19 connections server is down. Error message in log file is: "2013-02-12 08:02:07,844 [NioProcessor-7] WARN o.r.s.net.rtmp.RTMPMinaIoHandler - Exception caught Java heap space" Log file is attached. Thanks, Vasiliy -- Sebastian Wagner <https://twitter.com/#%21/dead_lock> https://twitter.com/#!/dead_lock <http://www.webbase-design.de> http://www.webbase-design.de <http://www.wagner-sebastian.com> http://www.wagner-sebastian.com <mailto:seba.wag...@gmail.com> seba.wag...@gmail.com
