On Saturday 20 October 2007 00:14, Jeff Rogers wrote: > Do you still get the lockup in your scenario with keepalives off? If > not, it might be a workaround until it gets fixed. If memory serves, it > would not be the first time that keepalives have been broken.
What I suspect is that there isn't a race condition, just a test condition which causes the system to appear stuck. I doubt I'll ever see this in production. But a low number of threads may lead to this type of thing where you have a few clients requesting a bunch of files using keepalive. When the thread exits, it could leave unprocessed requests in the queue. Or maybe a few clients could essentially hog all the threads within the keep-alive loop. I'm not 100% sure this is a short circuit of the regular queue, but it could be. If these greedy clients hog all the connection threads, and work fast enough, the connection threads will never exit (due to the new patch). This might actually work against sharing threads, but I'll have to test if exiting a thread and creating another one gives queued requests a chance to execute. tom jackson -- AOLserver - http://www.aolserver.com/ To Remove yourself from this list, simply send an email to <[EMAIL PROTECTED]> with the body of "SIGNOFF AOLSERVER" in the email message. You can leave the Subject: field of your email blank.
