You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@activemq.apache.org by Geurt Schimmel <GS...@schubergphilis.com> on 2013/03/15 17:25:55 UTC

bad performance, and loosing messages

We have a number of broker-networks with static bridges in between. We are experiencing bad performance, and loosing messages. Jvm -Xms1024M -Xmx1024M, queue memorylimit 64Mb, producerflowcontrol true. 500 Messages of 200 bytes are posted by a Stomp-client to only one node at a rate of 200 msg/sec.  Interesting enough, exactly 75% is lost. Decreasing the rate or message size hardly influences the results.

Please advise...

The typical config of all nodes (with a different uri for each network-of-brokers) is:

  <transportConnector name="openwire" uri="ssl://0.0.0.0:61616?jms.prefetchPolicy.queuePrefetch=4000" updateClusterClients="true" rebalanceClusterClients="true" updateClusterClientsOnRemove="true"/>

     <networkConnector name="BrokerLinkPmq101ToOmq" userName="system" password="${activemq.password}" uri="static:(failover:(ssl://omq101-front:61616,ssl://omq401-front:61616)?randomize=false&amp;priorityBackup=true)"  conduitSubscriptions="false" decreaseNetworkConsumerPriority="true" duplex="true" networkTTL="10" prefetchSize="4000" staticBridge="true" suppressDuplicateQueueSubscriptions="true" >
        <staticallyIncludedDestinations>
        ....
        </staticallyIncludedDestinations>
      </networkConnector>

Thanks !

Re: bad performance, and loosing messages

Posted by Geurt Schimmel <GS...@schubergphilis.com>.
Fixed by setting jms.optimizeAcknowledge=true:

http://activemq.apache.org/optimized-acknowledgement.html

Thanks

On 15 mrt. 2013, at 17:26, "Geurt Schimmel" <GS...@schubergphilis.com>> wrote:

We have a number of broker-networks with static bridges in between. We are experiencing bad performance, and loosing messages. Jvm -Xms1024M -Xmx1024M, queue memorylimit 64Mb, producerflowcontrol true. 500 Messages of 200 bytes are posted by a Stomp-client to only one node at a rate of 200 msg/sec.  Interesting enough, exactly 75% is lost. Decreasing the rate or message size hardly influences the results.

Please advise...

The typical config of all nodes (with a different uri for each network-of-brokers) is:

 <transportConnector name="openwire" uri="ssl://0.0.0.0:61616?jms.prefetchPolicy.queuePrefetch=4000" updateClusterClients="true" rebalanceClusterClients="true" updateClusterClientsOnRemove="true"/>

    <networkConnector name="BrokerLinkPmq101ToOmq" userName="system" password="${activemq.password}" uri="static:(failover:(ssl://omq101-front:61616,ssl://omq401-front:61616)?randomize=false&amp;priorityBackup=true)"  conduitSubscriptions="false" decreaseNetworkConsumerPriority="true" duplex="true" networkTTL="10" prefetchSize="4000" staticBridge="true" suppressDuplicateQueueSubscriptions="true" >
       <staticallyIncludedDestinations>
       ....
       </staticallyIncludedDestinations>
     </networkConnector>

Thanks !