httpd-bugs mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From bugzi...@apache.org
Subject DO NOT REPLY [Bug 10266] - apache (prefork, ssl) hangs after some hours of running
Date Mon, 07 Apr 2003 21:59:45 GMT
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10266>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10266

apache (prefork, ssl) hangs after some hours of running

dcook@cookware.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |
            Version|2.0.44                      |2.0.45



------- Additional Comments From dcook@cookware.com  2003-04-07 21:59 -------
I was sent an email the other day telling me to upgrade to .45 and to try all the MUTEX methods.
We have since upgraded to version .45 and we have confirmed that children process hang with
ALL mutex types.  Our platform allowed us to try the following MUTEX:AcceptMutex fcntlAcceptMutex
posixsemAcceptMutex sysvsemAcceptMutex pthread Our platform defaults to pthread... but interestingly
enough when I switched to 'fcntl' and restarted apache (from a cold start) apache IMMEDIATLY
hung some children.  I killed and restarted it again and it was fine for a few hours but then
hung some children.Same for sysvsem... it hung a few minutes ago and our automated software
did a complete kill of apache and then a restart - and the restart hung immediatly.As I've
said repeatedly, this is a very random thing. Sometimes it will run for 2 or 3 days with no
problem, sometimes it will hang once an hour for 4 or 5 hours (our software catches the hang
and fixes it in 15 seconds of the hang by issuing a -USR1 signal).I would really really really
like to figure this out since it makes our servers much more problematic to maintain - not
to mention all the specialized software we have had to write to identify hung servers so we're
not woken up at 3:00 AM to system alarms.We are running SUN V100 boxes, but have seen this
on our ULTRA boxes that we replaced (with the V100's) as well.So anyway, MUTEX does not seem
to change it... the problem occurs in ALL MUTEX that we can test under.We also do not feel
that this is in any way related to SSL.  We run a separate apache just for SSL, so the apaches
in question, while having SSL compiled into them, are not receiving SSL hits.Also, this does
not appear to be related to hits except for the fact that of the 3 apaches we run, the one
that runs the most virtual sites hangs much more frequently than the others.

---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org


Mime
View raw message