Try setting threadFactory=Y in freenet.conf or wherever it is in Windows. YThreadFactory was designed to avoid exactly this problem. Be sure to remove the % from the front of the line.
This helps only if those threads would have exited quickly if you had lots more physical memory than you have. You could also try increasing tfTolerableQueueDelay. See the "command line info" page accessible from the web interface page. tfTolerableQueueDelay specifies how long jobs must wait before new threads are created for [them. milliseconds.] -- Ed Huff On Wed, 2003-11-19 at 05:46, S wrote: > Hi all: > > I'm running 5038. The problem indicated below generally crops up after a > few hours of running my node, it's been taking place for the last few > builds. Everything runs just fine for a few hours, then all of a sudden, > the number of threads skyrockets out of control until Freenet (or > perhaps Java itself) collapses with Out Of Memory errors. > > Any idea what might cause this? Symptoms are on a P3 600 with a 10 gig > store, if there are other relevant settings, ask and I'll tell. > > Current routingTime 0ms > Curent messageSendTimeRequest 0ms > Pooled threads running jobs 6526 (1,274.6%) > Pooled threads which are idle 396 > Current upstream bandwidth usage 21 bytes/second (0.1%) > Reason for refusing connections: activeThreads(6520) >= maximumThreads (512) > > I have the thread limit set to 512. When this problem crops up, the disk > starts thrashing like all holy hell, and eventually Freenet dies with > Out of Memory as the most recent log entries. It's sporadic; I have kept > recent builds up for 24+ hours with no problem, then the thread overage > kicks in after a couple of hours on the next run. > > -s > _______________________________________________ > Devl mailing list > [EMAIL PROTECTED] > http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl _______________________________________________ Devl mailing list [EMAIL PROTECTED] http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl
