|
|||||||||||
|
DO NOT REPLY [Bug 10266] - apache hangs after some hours of running
From: <bugzilla(at)apache.org>
Date: Sun Nov 24 2002 - 14:50:54 EST
apache hangs after some hours of running
At the time of the hang: how many child processes and what are they doing? (run pstack against each one to see what is going on; presumably most of the children will have the same backtrace, so no sense sending in duplicate backtraces) what is the parent doing? run truss against it for a while (15 seconds) if it doesn't appear to be doing anything, run pstack against it to get a backtrace is it possible that all available children are handling connections? look at netstat output for ports served by Apache to see if there are enough hung connections to represent each available child wild and crazy idea: try a different accept mutex mechanism (e.g., "AcceptMutex fcntl" in httpd.conf) To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org For additional commands, e-mail: bugs-help@httpd.apache.org Received on Sun Nov 24 19:49:51 2002 This archive was generated by hypermail 2.1.8 : Wed Aug 23 2006 - 16:44:08 EDT |
||||||||||
|
|||||||||||