All, I am running Tomcat 3.2.3.
I have modified my server.xml file to specify some of the sizing parameters on my PoolTcpConnectors, (both the Tomcat listen port and the AJP12 port, used for integration with Web servers) like so: <Connector className="org.apache.tomcat.service.PoolTcpConnector"> <Parameter name="handler" value="org.apache.tomcat.service.http.HttpConnectionHandler"/> <Parameter name="port" value="8080"/> <Parameter name="backlog" value="100"/> <Parameter name="thread_pool" value="on"/> <Parameter name="max_threads" value="500"/> <Parameter name="max_spare_threads" value="50"/> <Parameter name="min_spare_threads" value="25" /> </Connector> I believe this setup would in theory, allow me to have up to 500 concurrent requests on the port specified by the "port" attribute. I have two questions: 1) Is there any significant performance degradation I would see by having this set so high in an installation where this level of concurrency may never be reached? That is, can I just set these attributes as a "just in case it's a big installation" and not be concerned that they are too high when the level of concurrency is low? 2) Am I correct in understanding that this only affects listener threads and does not affect actual worker threads dispatched to handle these requests? Is there pooling for the worker threads and how is it managed? Thanks for any help, Wes Gamble -- To unsubscribe: <mailto:[EMAIL PROTECTED]> For additional commands: <mailto:[EMAIL PROTECTED]> Troubles with the list: <mailto:[EMAIL PROTECTED]>
