Thanks Mladen for your suggestions. But I have to make clear that we just started with a minimum configuration for the workers.
So we had 3 workers defined using the following config: worker.worker_1.host=localhost worker.worker_1.port=8009 worker.worker_1.type=ajp13 worker.worker_2.host=localhost worker.worker_2.port=8109 worker.worker_2.type=ajp13 worker.worker_3.host=localhost worker.worker_3.port=8209 worker.worker_3.type=ajp13 There was nothing about connection recycling - just using the standard. So I am not sure about that what you have said about making sure that not only one endpoint of the connection is dropped, which should not happen in this case?! Afterwards, we've checked the manuals again and set up a quite similar configuration just as Jonathon did (except the keep_alive parameter, since we were running Apache and Tomcat on the same host). That's where I could follow your objection. But furthermore I cannot imagine how the whole setup could be running for about 2 weeks and then go down after a couple of minutes (after a server reboot). Even under heavy load conditions this seems very mysterious, since even with the "recycle configuration", the connections should not be dropped that fast, so that there are dead sockets hanging around. Even the problem with server reboot at all is quite strange, since it doesn't seem to be a problem with the OS. After Tomcat was down, I could connect to the server remotely using the MS Terminal Client (RDP - Connection). So no general socket problem at all, in my opinion... Is it possible that after Tomcat shutdown any Java related thread or process is still keeping the dead sockets and no connections can be made until this thread/process is killed during server shutdown? Additional information: Tomcat is running with jdk1.5.0_04, using the server JVM. @Jonathon: Do you use any special switches for the JVM? We're using the same switches which worked fine for us when running the system on Win2K Server (where the system ran fine before using mod_jk 1.2.6 - ;) ), such as settings PermSize etc. Regards, Robert. Jonathon Reeve wrote: > Well I'm glad someone else is getting this, even if you have no idea > either. We wondered if we were missing some configuration from > workers.properties, since the "quick start" section > (http://tomcat.apache.org/connectors-doc/howto/quick.html) on > Tomcat's site gives an example "minimum workers.properties": > Huh, you really tried everything :) Couple of notes: 1. AJP connection is supposed to last forever. 2. cache_timeout, recycle_timeout, MaxRequestsPerClient tend to close the socket, so you have half-closed sockets, and that's why you are leaking resources. 3. Whenever there is a need to recycle one end of the channel the other end must be recycled too. This is done by setting connectionTimeout="milliseconds" in server.xml for AJP/1.3 connector. 4. If you need full failover add 'connection_timeout' and 'prepost_timeout' to each worker. This implicates that CPING/CPONG packets will be send before each request is forwarded to Tomcat. Regards, Mladen. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]