> We're having a killer problem on our production server, where tomcat 4.0.3 > goes crazy allocating threads in its thread pool, until it either runs out > of threads or heap (for us it's heap). > > Searching the list I found a reference to a (closed, not reproducible) bug > that looks the same: > http://nagoya.apache.org/bugzilla/show_bug.cgi?id=5735 > > We are using tomcat standalone, with the https connector. > > I have re-opened the bug, as it looks very much like what is happening to > us. I'd really like to know from people is what further specific > information we can contribute to help diagnose this problem. > > Failing that we are looking at an emergency fallback to Tomcat 3, where we > didn't see this problem.
There are a few possibilities: - it happens because of GC (see the analysis posted by Glenn a while ago) - it happens because of some bug in the thread pool; maybe, but I doubt it; the most recent Tomcat 4 connector (Coyote b7 and up) uses the Tomcat 3 thread pool, so it would be relatively easy to solve your problem in that case - it happens because of some memory leak (I mentioned the possibility of having too many active sessions, but it could be something else) Remy -- To unsubscribe: <mailto:[EMAIL PROTECTED]> For additional commands: <mailto:[EMAIL PROTECTED]> Troubles with the list: <mailto:[EMAIL PROTECTED]>
