On 31/01/12 09:16, [email protected] wrote:
Hi,
Sometime my server crash, the SOGo process take 100% of cpu, if I do
"/etc/init.d/sogo stop", sogo didn't stop, I'm forced to kill process
with "kill -KILL".
In log, I have a lot of "the socket was shutdown".
An extract from my logs:
server.domain.loc - - [25/Jan/2012:14:36:35 GMT] "REPORT
/SOGo/dav/user3/Calendar/user4_C495AEAB-88F0-0001-D1B4-653E46904F40/
HTTP/1.1" 207 125/245 0.009 - - 0
server.domain.loc - - [25/Jan/2012:14:36:35 GMT] "OPTIONS
/SOGo/dav/user3/Calendar HTTP/1.1" 200 0/0 0.005 - - 0
server.domain.loc - - [25/Jan/2012:14:36:35 GMT] "REPORT
/SOGo/dav/user3/Calendar/personal/ HTTP/1.1" 207 125/245 0.009 - - 0
server.domain.loc - - [25/Jan/2012:14:36:35 GMT] "OPTIONS
/SOGo/dav/user3/Contacts HTTP/1.1" 200 0/0 0.011 - - 0
Jan 25 14:37:38 sogod [8438]: [WARN] <0x0x19a10fc8[WOWatchDogChild]>
pid 30829 has been hanging in the same request for 1 minutes
I had similar problems.
Jan 25 14:38:38 sogod [8438]: [WARN] <0x0x19a10fc8[WOWatchDogChild]>
pid 30829 has been hanging in the same
);
SxVMemLimit = 512;
WOMessageUseUTF8 = YES;
WOParsersUseUTF8 = YES;
WOPort = 20000;
WOUseRelativeURLs = YES;
};
}
How many WOWorkers do you have ?
One thing to check is if you use a debian based system, the setting for
WOWorkersis actually overwritten by the value in /etc/default/sogo ie
the value in the main .GNUStep config is not taken
ie my /etc/default/sogo looks like :-
# The amount of processes that should be spawned (Default: 1)
PREFORK=5
# The name of the account under which SOGo will be running (Default: sogo)
# USER=sogo
Hope this helps
Julian
--
[email protected]
https://inverse.ca/sogo/lists
