[jboss-user] [Clustering/JBoss] - mod_jk workers un-balanced under load

[jimyliu]

Hi, 

I am doing a performance test with our application. The test was run with 1,2,4,8,16,32,64,128,200,256 users and I collect the page throughput through Load Runner. 

My configuration:
One apache web server (2.2.9) with mod_jk 1.2.26 (on Red Hat Linux)
Two application servers. EAP 4.2 Each run with two JBoss instances. 
So the apache's workers.properties file point to these 4 JBoss instances.

Everything went well until I ran with 256 virtual users. I found most(98%) of the load went to single JBoss instance. So the throughput went down a lot. 

I also view the mod_jk status from browser, same thing. All load went to only one worker. 

Any idea or suggestion ? 

Jimy

Following are the workers.properties and ajp1.3  setting. 
  | workers.properties file:
  | worker.list=loadbalancer,status
  | worker.node0.port=8009
  | worker.node0.host=192.168.11.31
  | worker.node0.type=ajp13
  | worker.node0.lbfactor=1
  | worker.node0.connect_timeout=10000
  | worker.node0.prepost_timeout=10000
  | worker.node0.socket_timeout=10
  | worker.node0.connection_pool_timeout=1200
  | worker.node1.port=8109
  | worker.node1.host=192.168.11.31
  | worker.node1.type=ajp13
  | worker.node1.lbfactor=1
  | worker.node1.connect_timeout=10000
  | worker.node1.prepost_timeout=10000
  | worker.node1.socket_timeout=10
  | worker.node1.connection_pool_timeout=1200
  | worker.node2.port=8009
  | worker.node2.host=192.168.11.32
  | worker.node2.type=ajp13
  | worker.node2.lbfactor=1
  | worker.node2.connect_timeout=10000
  | worker.node2.prepost_timeout=10000
  | worker.node2.socket_timeout=10
  | worker.node2.connection_pool_timeout=1200
  | worker.node3.port=8109
  | worker.node3.host=192.168.11.32
  | worker.node3.type=ajp13
  | worker.node3.lbfactor=1
  | worker.node3.connect_timeout=10000
  | worker.node3.prepost_timeout=10000
  | worker.node3.socket_timeout=10
  | worker.node3.connection_pool_timeout=1200
  | 
  | worker.loadbalancer.type=lb
  | worker.loadbalancer.balance_workers=node0,node1,node2,node3
  | worker.loadbalancer.sticky_session=1
  | worker.status.type=status
  | 
  | 
  | This is one of the AJP setting in server.xml in JBoss:
  | <Connector port="8009" address="${jboss.bind.address}" protocol="AJP/1.3"  maxThreads="400" connectionTimeout="10000"
  |      emptySessionPath="true" enableLookups="false" redirectPort="8443" />
  |     <Engine name="jboss.web" defaultHost="localhost" jvmRoute="node0">





View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4167782#4167782

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4167782