Having just taken a look at the code, it does seem that this is a bug for Tomcat 4.1 with the CoyoteConnector. Could you please submit it at http://nagoya.apache.org/bugzilla?
"Dan Higgins" <[EMAIL PROTECTED]> wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > Hi, > > If I set my maxProcessors sufficiently low (say 5), then run a script from > another machine that floods Tomcat with requests for a given JSP, eventually > Tomcat stops accepting socket connections at all. > > Ok, that's to be expected when every processor thread is busy. BUT, when the > flood stops, all the threads let their socket connections remain in > CLOSE_WAIT and they never close them. This causes Tomcat to NEVER accept new > connections, until it's restarted. > > I'm seeing this problem happen on a site with 600 max processors. > Eventually, it stops accepting connections and has to be restarted, every > couple days. > > This is Tomcat 4.0.6 on Windows 2000, JDK 1.4.1_01. > > Any idears?? > > Regards, > -Dan --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]