I am trying to understand why a server restart would cause mod_jk to not find a worker. I checked all tomcat workers were running and I was able to telnet to the ajp port on each app server.
Just to give background I have 1 web server that is running apache with mod_jk for load balancing. The apache web server is configured to run as a Service and is setup to run Automatically. I have 5 tomcat workers (app engines) that serve http requests from the web server.
---------------------
The mod_jk log showed around the time of patch application
[Mon Aug 27 00:11:07 2007] [error] ajp_connection_tcp_get_message::jk_ajp_common.c (933): ERROR: can't receive the response message from tomcat, network problems or tomcat is down (10.3.200.187:8060), err=-54
[Mon Aug 27 00:11:07 2007] [error] ajp_get_reply::jk_ajp_common.c (1432): Tomcat is down or network problems. No response has been sent to the client (yet)
[Mon Aug 27 00:11:07 2007] [info] ajp_service::jk_ajp_common.c (1645): Receiving from tomcat failed, recoverable operation attempt=2
[Mon Aug 27 00:11:07 2007] [info] ajp_service::jk_ajp_common.c (1665): Sending request to tomcat failed, recoverable operation attempt=2
[Mon Aug 27 00:11:07 2007] [error] ajp_service::jk_ajp_common.c (1673): Error connecting to tomcat. Tomcat is probably not started or is listening on the wrong port. worker=app1 failed errno = 54
----------------------------
The errog log showed this message
------------------
[Sun Aug 26 23:52:18 2007] [info] (OS 10054)An existing connection was forcibly closed by the remote host. : core_output_filter: writing data to the network
----------------
The mod_jk log then showed this message after the reboot and when the web server was started.
[Mon Aug 27 00:48:07 2007] [info] jk_handler::mod_jk.c (1883): Could not find a worker for worker name=loadbalancer
[Mon Aug 27 00:48:14 2007] [info] jk_handler::mod_jk.c (1883): Could not find a worker for worker name=loadbalancer
----------------
The message right above kept registering every time I would try to bring up a webpage.
Do you think this is somehow caused by the abrupt shutdown and restart? I don't see anything in the error.log indicating a problem with the restart except for the warning message saying "pid file apache.pid - unclean shutdown of previous Apache run?"
I am able to restart the web and resume normal operation but would like to understand the cause so I can document it.
Any help is greatly appreciated.
Thanks.
Kuljit
Just to give background I have 1 web server that is running apache with mod_jk for load balancing. The apache web server is configured to run as a Service and is setup to run Automatically. I have 5 tomcat workers (app engines) that serve http requests from the web server.
---------------------
The mod_jk log showed around the time of patch application
[Mon Aug 27 00:11:07 2007] [error] ajp_connection_tcp_get_message::jk_ajp_common.c (933): ERROR: can't receive the response message from tomcat, network problems or tomcat is down (10.3.200.187:8060), err=-54
[Mon Aug 27 00:11:07 2007] [error] ajp_get_reply::jk_ajp_common.c (1432): Tomcat is down or network problems. No response has been sent to the client (yet)
[Mon Aug 27 00:11:07 2007] [info] ajp_service::jk_ajp_common.c (1645): Receiving from tomcat failed, recoverable operation attempt=2
[Mon Aug 27 00:11:07 2007] [info] ajp_service::jk_ajp_common.c (1665): Sending request to tomcat failed, recoverable operation attempt=2
[Mon Aug 27 00:11:07 2007] [error] ajp_service::jk_ajp_common.c (1673): Error connecting to tomcat. Tomcat is probably not started or is listening on the wrong port. worker=app1 failed errno = 54
----------------------------
The errog log showed this message
------------------
[Sun Aug 26 23:52:18 2007] [info] (OS 10054)An existing connection was forcibly closed by the remote host. : core_output_filter: writing data to the network
----------------
The mod_jk log then showed this message after the reboot and when the web server was started.
[Mon Aug 27 00:48:07 2007] [info] jk_handler::mod_jk.c (1883): Could not find a worker for worker name=loadbalancer
[Mon Aug 27 00:48:14 2007] [info] jk_handler::mod_jk.c (1883): Could not find a worker for worker name=loadbalancer
----------------
The message right above kept registering every time I would try to bring up a webpage.
Do you think this is somehow caused by the abrupt shutdown and restart? I don't see anything in the error.log indicating a problem with the restart except for the warning message saying "pid file apache.pid - unclean shutdown of previous Apache run?"
I am able to restart the web and resume normal operation but would like to understand the cause so I can document it.
Any help is greatly appreciated.
Thanks.
Kuljit