Hello,

If you have enabled all logs and this does not give any additional information, 
you have to break with gdb on the pid hanging and provide the backtrace : 
https://www.sogo.nu/support/faq/how-do-i-debug-sogo.html
 
You can also attach to a running sogod process using gdb. For example, if a 
sogod process is consuming 100% and its PID is 1234, you can attach to it using 
“gdb -p 1234”. At this point, you can also produce back traces (“bt” or “bt 
full”), put the process back in the running state (“c”) or interrupt it 
(Control-C).Sebastien

Le Mardi, Février 07, 2023 16:44 CET, "Jose David Bravo A" 
([email protected]) <[email protected]> a écrit:
 Hello, I have a server running SOGo 5.8.0 (Ubuntu 22.04.1) that is realy slow, 
I'm seeing that some processes are getting stuck in the same request: # tail -f 
/var/log/sogo/sogo.log|grep hangingFeb 07 10:28:17 sogod [853676]: [WARN] 
<0x0x555e931e4df0[WOWatchDogChild]> pid 853771 has been hanging in the same 
request for 1 minutesFeb 07 10:28:25 sogod [853676]: [WARN] 
<0x0x555e931e5820[WOWatchDogChild]> pid 853775 has been hanging in the same 
request for 1 minutesFeb 07 10:28:27 sogod [853676]: [WARN] 
<0x0x555e931f51b0[WOWatchDogChild]> pid 853777 has been hanging in the same 
request for 1 minutesFeb 07 10:28:40 sogod [853676]: [WARN] 
<0x0x555e931e4790[WOWatchDogChild]> pid 853774 has been hanging in the same 
request for 2 minutes If I check the SOGo log, the latest log that I get from 
specific PID is:
Feb 07 10:30:50 sogod [853679]: <0x0x555e932cade0[NGImap4Client]> TLS started 
successfully. So, I suppose that the problem is between SOGo and our 
dovecot/imap service. But I can't see any related dovecot log.

If I set the WOWatchDogRequestTimeout to 10, those stuck processes gets killied 
after 10 minutes, but the problem persists. And this is causing the CPU of the 
server to overload.

My WOWorkersCount is 100, and my SxVMemLimit is 512. I have 2818 users in this 
server, but not everybuddy is using SOGo. I tried enabling ImapDebugEnabled, 
but the last log line I receive from the stuck pid is "TLS started 
successfully.", and I don't get any information about what is hanging the 
process.

Any idea what can I do to determine what is hanging those processes. Thank 
you!,   Jose David Bravo

 

Reply via email to