Connection Reset by Peer error with Apache and JBoss 7.1.1

Posted by vikingz on Server Fault See other posts from Server Fault or by vikingz
Published on 2014-08-19T20:26:45Z Indexed on 2014/08/19 22:22 UTC
Read the original article Hit count: 293

Filed under:
|
|
|
|

We are seeing errors on some of our QA testing scripts that intermittently throw Connection Reset By Peer errors.

The Test scripts submit requests via F5 which forwards requests to Apache (2.2.21) with a mod_jk load_balancer with the following setting for each worker in the worker.property

    worker1 props
    worker.worker1.type=ajp13
    worker.worker1.port=8109
    worker.worker1.lbfactor=1
    worker.worker1.host=skunkhost1.com
    worker.worker1.connection_pool_timeout=30

and here is what is in the JBoss domain.xml for the AJP port from JBoss 7.1.1

    <unbounded-queue-thread-pool name="SKUNKY.APP.AJP">
                        <max-threads count="300"/>
                        <keepalive-time time="3" unit="minutes"/>
    </unbounded-queue-thread-pool>

Here is httpd.conf

    Timeout 300
    KeepAlive On
    KeepAliveTimeout 15
    MaxKeepAliveRequests 100
    TraceEnable Off

My question is that is it posisbe that apache times out and closes the connection while jboss is still ready and working on the request? What might be causing the Connection Reset By Peer error?what am i missing here? Any help is majorly appreciated!! Sincerely KK

© Server Fault or respective owner

Related posts about apache-2.2

Related posts about connection