@andreas-doehler Im cmc.log habe ich folgenden Eintrag gefunden, der nach dem Pushen von Changes generiert wurde:
2020-04-15 11:52:42 [4] [Check_MK helper 4644] >> Fatal Python error: GC object already tracked
2020-04-15 11:52:42 [2] [Check_MK helper 4644] cannot read response: Resource temporarily unavailable
2020-04-15 11:52:44 [4] [Check_MK helper 4644] restarting dead helper
2020-04-15 11:52:44 [3] [Check_MK helper 4644] died with signal 6
2020-04-15 11:52:44 [5] [Check_MK helper 11060] started, commandline: /omd/sites/MySite/bin/cmk --debug --keepalive
Allerdings habe ich diesen Eintrag auch nur ein Mal gefunden und nicht nach jedem Pushen von Changes.
@simonniz The helpers log into the cmc.log. An entry looks like this:
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] open file descriptors changed, currently open:
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 0 → socket:[39216996]
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 1 → pipe:[39216997]
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 2 → pipe:[39216997]
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 3 → socket:[39216996]
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 4 → socket:[42330895]
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 5 → socket:[40604550]
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 6 → socket:[43809839]
2020-04-15 15:02:48 [4] [Check_MK helper 4645] >> [cycle 3498, host “xxx.xxx.xxx.xxx”] 13 → /dev/urandom
But Andreas is right, the amount of output is indeed very large.