| 摘要 |
A server system in typical operation has a process manager, multiple listeners (each to receive requests for its protocols) and multiple worker processes that are each able to handle requests in multiple protocols. At server start-up, each listener connects with the process manager via a pipe published by the process manager. The listener then receives information via the process manager that includes information defining the applications(s) for which that listener is to "listen" and associating selected application(s) to application pool(s). When the listener detects a request for such an application, the listener starts a queue for the associated application pool. In response, the process manager launches an appropriate worker process to handle requests in the listener's protocol. The worker process then makes a connection with the listener.
|
