[SOGo] No child available to handle incoming request / pid xxx has been hanging in the same request for x minutes

[Alejandro Vargas]

I am trying to add SOGo to an already working server with Postfix +
Dovecot.

The server is Centos 7, 2 core with 3Gb RAM with less than 10 users.
Sogo installed from official repo: sogo-2.3.8-1.el7.centos.x86_64

Sogo is configured with 10 workers with this configuration:

WOWorkersCount = 10;
WOListenQueueSize=10;
WOWatchDogRequestTimeout=60;

SOGoMaximumPingInterval = 354;
SOGoMaximumSyncInterval = 354;

SOGoInternalSyncInterval = 15;
SOGoMaximumSyncWindowSize = 50;
SOGoMaximumSyncResponseSize = 2048;

The problem is SOGo processes starts eating all the RAM and sometimes
hangs. The log file reports:

Feb 19 13:30:26 sogod [13164]:
<0x0x7f1357056030[SOGoActiveSyncDispatcher]> Sleeping 15 seconds while
detecting changes in Ping...
Feb 19 13:30:26 sogod [13163]:
<0x0x7f13575da6d0[SOGoActiveSyncDispatcher]> Sleeping 15 seconds while
detecting changes in Ping...
Feb 19 13:30:26 sogod [13150]: [ERROR] <0x0x7f1357379d20[WOWatchDog]> No
child available to handle incoming request!
Feb 19 13:30:26 sogod [13155]:
<0x0x7f135797a390[SOGoActiveSyncDispatcher]> Sleeping 15 seconds while
detecting changes in Ping...
Feb 19 13:30:27 sogod [13152]:
<0x0x7f13576d30d0[SOGoActiveSyncDispatcher]> Sleeping 15 seconds while
detecting changes in Ping...
Feb 19 13:30:27 sogod [13150]: [WARN]
<0x0x7f13572efac0[WOWatchDogChild]> pid 13168 has been hanging in the
same request for 3 minutes
Feb 19 13:30:28 sogod [13150]: [ERROR] <0x0x7f1357379d20[WOWatchDog]> No
child available to handle incoming request!
Feb 19 13:30:28 sogod [13150]: [WARN]
<0x0x7f13572ef490[WOWatchDogChild]> pid 13164 has been hanging in the
same request for 3 minutes
Feb 19 13:30:29 sogod [13150]: [ERROR] <0x0x7f1357379d20[WOWatchDog]> No
child available to handle incoming request!
Feb 19 13:30:29 sogod [13150]: [WARN]
<0x0x7f13572ef2d0[WOWatchDogChild]> pid 13163 has been hanging in the
same request for 2 minutes
Feb 19 13:30:30 sogod [13168]:
<0x0x7f135770fe90[SOGoActiveSyncDispatcher]> Sleeping 15 seconds while
detecting changes in Ping...
Feb 19 13:30:30 sogod [13150]: [WARN]
<0x0x7f13572e4fa0[WOWatchDogChild]> pid 13151 has been hanging in the
same request for 1 minutes

Feb 19 13:35:03 sogod [13150]: [WARN]
<0x0x7f13572dd9a0[WOWatchDogChild]> pid 13153 has been hanging in the
same request for 5 minutes
Feb 19 13:35:04 sogod [13150]: [ERROR] <0x0x7f1357379d20[WOWatchDog]> No
child available to handle incoming request!
Feb 19 13:35:06 sogod [13150]: [ERROR] <0x0x7f1357379d20[WOWatchDog]> No
child available to handle incoming request!
Feb 19 13:35:07 sogod [13153]:
<0x0x7f13576691c0[SOGoActiveSyncDispatcher]> Sleeping 15 seconds while
detecting changes in Ping...
Feb 19 13:35:07 sogod [13150]: [ERROR] <0x0x7f1357379d20[WOWatchDog]> No
child available to handle incoming request!
Feb 19 13:35:08 sogod [13164]:
<0x0x7f13573c0570[SOGoActiveSyncDispatcher]> Sleeping 15 seconds while
detecting changes in Ping...
Feb 19 13:35:09 sogod [13150]: [ERROR] <0x0x7f1357379d20[WOWatchDog]> t!


I used gdb to get a trace of one of the hanging processes. The response
is this:

#0 0x00007f176ddcc49d in nanosleep () from /lib64/libc.so.6
#1 0x00007f176ddcc334 in sleep () from /lib64/libc.so.6
#2 0x00007f17608e8a99 in -[SOGoActiveSyncDispatcher
processPing:inResponse:] () from
/usr/lib64/GNUstep/SOGo/ActiveSync.SOGo/./ActiveSync
#3 0x00007f17608eee4b in -[SOGoActiveSyncDispatcher
dispatchRequest:inResponse:context:] () from
/usr/lib64/GNUstep/SOGo/ActiveSync.SOGo/./ActiveSync
#4 0x00007f1760d50d84 in -[SOGoMicrosoftActiveSyncActions
microsoftServerActiveSyncAction] () from
/usr/lib64/GNUstep/SOGo/MainUI.SOGo/./MainUI
#5 0x00007f1773e61113 in -[WODirectAction performActionNamed:] () from
/lib64/libNGObjWeb.so.4.9
#6 0x00007f1773ee3834 in -[SoActionInvocation
callOnObject:withPositionalParametersWhenNotNil:inContext:] () from
/lib64/libNGObjWeb.so.4.9
#7 0x00007f1773edee98 in -[SoObjectMethodDispatcher dispatchInContext:]
() from /lib64/libNGObjWeb.so.4.9
#8 0x00007f1773ee0f09 in -[SoObjectRequestHandler
handleRequest:inContext:session:application:] () from
/lib64/libNGObjWeb.so.4.9
#9 0x00007f1773e72753 in -[WORequestHandler handleRequest:] () from
/lib64/libNGObjWeb.so.4.9
#10 0x00007f1773e3433c in -[WOCoreApplication
dispatchRequest:usingHandler:] () from /lib64/libNGObjWeb.so.4.9
#11 0x00007f1773e3463f in -[WOCoreApplication dispatchRequest:] () from
/lib64/libNGObjWeb.so.4.9
#12 0x00007f17751fbb4d in -[SOGo dispatchRequest:] ()
#13 0x00007f1773ed1a85 in -[WOHttpTransaction _run] () from
/lib64/libNGObjWeb.so.4.9
#14 0x00007f1773ed1de5 in -[WOHttpTransaction run] () from
/lib64/libNGObjWeb.so.4.9
#15 0x00007f1773ecd9e4 in -[WOHttpAdaptor runConnection:] () from
/lib64/libNGObjWeb.so.4.9
#16 0x00007f1773ecdc02 in -[WOHttpAdaptor _handleAcceptedConnection:] ()
from /lib64/libNGObjWeb.so.4.9
#17 0x00007f1773ecdff7 in -[WOHttpAdaptor _handleConnection:] () from
/lib64/libNGObjWeb.so.4.9
#18 0x00007f1773ece2c3 in -[WOHttpAdaptor acceptControlMessage:] () from
/lib64/libNGObjWeb.so.4.9
#19 0x00007f177261613f in -[NSNotificationCenter _postAndRelease:] ()
from /lib64/libgnustep-base.so.1.24
#20 0x00007f17732a0e3d in -[NSObject(FileObjectWatcher)
receivedEvent:type:extra:forMode:] () from /lib64/libNGExtensions.so.4.9
#21 0x00007f177271ceea in -[GSRunLoopCtxt pollUntil:within:] () from
/lib64/libgnustep-base.so.1.24
#22 0x00007f177265d870 in -[NSRunLoop acceptInputForMode:beforeDate:] ()
from /lib64/libgnustep-base.so.1.24
#23 0x00007f177265dd22 in -[NSRunLoop runMode:beforeDate:] () from
/lib64/libgnustep-base.so.1.24
#24 0x00007f1773e33b94 in -[WOCoreApplication run] () from
/lib64/libNGObjWeb.so.4.9
#25 0x00007f17751fb1fe in -[SOGo run] ()
#26 0x00007f1773e7bc5e in -[WOWatchDog _runChildWithControlSocket:] ()
from /lib64/libNGObjWeb.so.4.9
#27 0x00007f1773e7c0f1 in -[WOWatchDog _spawnChild:] () from
/lib64/libNGObjWeb.so.4.9
#28 0x00007f1773e7c7d9 in -[WOWatchDog _ensureChildren] () from
/lib64/libNGObjWeb.so.4.9
#29 0x00007f1773e7d7f6 in -[WOWatchDog run:argc:argv:] () from
/lib64/libNGObjWeb.so.4.9
#30 0x00007f1773e7df21 in WOWatchDogApplicationMain () from
/lib64/libNGObjWeb.so.4.9
#31 0x00007f17751fa491 in main ()

Any help please?
--
ALEJANDRO VARGAS
_Departamento de Sistemas_
_607760045 || 5045_
***@zenerplus.com_
_www.zener.es [1]__ || __Linkedin: GrupoZener [2]_



Links:
------
[1] http://www.zener.es
[2] https://www.linkedin.com/company/grupo-zener

[Christian Mack]

Hello

I am trying to add SOGo to an already working server with Postfix + Dovecot.
The server is Centos 7, 2 core with 3Gb RAM with less than 10 users. Sogo
installed from official repo: sogo-2.3.8-1.el7.centos.x86_64
WOWorkersCount = 10;

< cut >

This setting is ignored.
Please instead increase PREFORK in /etc/sysconfig/sogo, as that is set
in the service start script as number of workers.


Kind regards,
Christian Mack

[Alejandro Vargas]

Post by Christian Mack
Hello

I am trying to add SOGo to an already working server with Postfix + Dovecot. The server is Centos 7, 2 core with 3Gb RAM with less than 10 users. Sogo installed from official repo: sogo-2.3.8-1.el7.centos.x86_64 Sogo is configured with 10 workers with this configuration: WOWorkersCount = 10;

< cut >
This setting is ignored.
Please instead increase PREFORK in /etc/sysconfig/sogo, as that is set
in the service start script as number of workers.

I already added "PREFORK=10" to /etc/sysconfig/sogo. But it does not
solve the problem.

[anv]

Post by Alejandro Vargas
I am trying to add SOGo to an already working server with Postfix +
Dovecot.
The server is Centos 7, 2 core with 3Gb RAM with less than 10 users.
Sogo installed from official repo: sogo-2.3.8-1.el7.centos.x86_64
WOWorkersCount = 10; WOListenQueueSize=10; WOWatchDogRequestTimeout=60;
SOGoMaximumPingInterval = 354; SOGoMaximumSyncInterval = 354;
SOGoInternalSyncInterval = 15; SOGoMaximumSyncWindowSize = 50;
SOGoMaximumSyncResponseSize = 2048;

Just configured default values and the problem is semi-solved. Now
sometimes there is 100% CPU usage. I will start a new thread for this.