Return-Path: Delivered-To: apmail-activemq-users-archive@www.apache.org Received: (qmail 45804 invoked from network); 7 Apr 2010 07:57:09 -0000 Received: from unknown (HELO mail.apache.org) (140.211.11.3) by 140.211.11.9 with SMTP; 7 Apr 2010 07:57:09 -0000 Received: (qmail 22983 invoked by uid 500); 7 Apr 2010 07:57:09 -0000 Delivered-To: apmail-activemq-users-archive@activemq.apache.org Received: (qmail 22830 invoked by uid 500); 7 Apr 2010 07:57:09 -0000 Mailing-List: contact users-help@activemq.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: users@activemq.apache.org Delivered-To: mailing list users@activemq.apache.org Received: (qmail 22809 invoked by uid 99); 7 Apr 2010 07:57:08 -0000 Received: from nike.apache.org (HELO nike.apache.org) (192.87.106.230) by apache.org (qpsmtpd/0.29) with ESMTP; Wed, 07 Apr 2010 07:57:08 +0000 X-ASF-Spam-Status: No, hits=0.0 required=10.0 tests=FREEMAIL_FROM,SPF_HELO_PASS,SPF_PASS,T_TO_NO_BRKTS_FREEMAIL X-Spam-Check-By: apache.org Received-SPF: pass (nike.apache.org: domain of lists@nabble.com designates 216.139.236.158 as permitted sender) Received: from [216.139.236.158] (HELO kuber.nabble.com) (216.139.236.158) by apache.org (qpsmtpd/0.29) with ESMTP; Wed, 07 Apr 2010 07:57:01 +0000 Received: from isper.nabble.com ([192.168.236.156]) by kuber.nabble.com with esmtp (Exim 4.63) (envelope-from ) id 1NzQ7s-0001pv-TM for users@activemq.apache.org; Wed, 07 Apr 2010 00:56:40 -0700 Message-ID: <28161669.post@talk.nabble.com> Date: Wed, 7 Apr 2010 00:56:40 -0700 (PDT) From: Richard Holt To: users@activemq.apache.org Subject: No Message Loss With Intermitent Services MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Nabble-From: richard_holt@btopenworld.com X-Virus-Checked: Checked by ClamAV on apache.org Hi, Our system is distributed where individual services may be offline for an unknown period of time, during these periods the activemq server is required to buffer and store the messages. It is not acceptable for us to slow down producers as we purposefully keep these clients thin and they cannot handle backlogs gracefully. Disk space is not really an issue so we are happy to use this. So my ideal solution is to use memory until close to the ceiling (allowing enough for processing to still occur) and then page messages out to disk. Also some of our queues get a lot more messages than others (and are less important) so i am quite happy to page those out sooner than the business messages (critical) >From my research it appears a fileCursor within a destinationPolicy is the way to go with a systemUsage section. However during my testing on 5.3 i can generate OOM messages still. I am running the system below in Pure Master/Slave configuration. I have attached the 2 configs i am using. I have used a low memory activemq (32mb) on purpose to try and hit hard limits quicker. Any advice is appreciated. And please don't direct me to the documentation... Thanks In Advance http://old.nabble.com/file/p28161669/activemq.xml activemq.xml http://old.nabble.com/file/p28161669/activemq-slave.xml activemq-slave.xml -- View this message in context: http://old.nabble.com/No-Message-Loss-With-Intermitent-Services-tp28161669p28161669.html Sent from the ActiveMQ - User mailing list archive at Nabble.com.