Rick Mann wrote:
> Some additional information: I tried installing 4.0.5 on my better server 
> (Resin-4.0.0, Nexenta/Open Solaris, recent, lots of RAM, faster CPUs), and 
> got similar behavior. I got it up and running, noticed some oddness at first 
> (java link errors running my app, slow page loads the first few times), then 
> it seemed to get normal.
> Then about  5 minutes later, it stopped responding. I didn't investigate, but 
> rather just switched back to using resin-4.0.0. No idea if it was the same 
> issue. In the logs, I saw it restarting, but it never came back.
Thanks for the update. Just to review the symptoms:

1. Both JVMs are around and responds to kill -QUIT (there have been 
cases where the JVM has been so frozen it doesn't respond to anything.)

2. Resin is not responding to connect() for either of two ports.

3. The thread dump shows active threads in the nativeAccept() of both ports.

4. Netstat looks normal (i.e. it's not showing lots of connections in 
strange states.)

My stress testing isn't showing anything like this kind of behavior yet, 
but any further details would help narrow the tests.

-- Scott

resin-interest mailing list

Reply via email to