- Plesk Onyx for Linux
Note: This article has the reference to the issues with the fixes available:
- #PPP-36948 "The Wdcollect service is now stopped correctly. It is no longer killed instead of being stopped and the system reboot
time is not increased by 90 seconds."
- Plesk Onyx 17.9 Preview 6 25 September 2018 (Linux)
- #PPPM-8683 "The wdcollect.service is not stopped and the process is killed after this due to timeout (90 seconds) what increase the system reboot time for 90 seconds (as PPP-36948)"
The following error appears in an attempt to reboot the server:
CONFIG_TEXT: A stop job is running for watchdog wdcollect (for 1min 32s)
After 90 seconds the server is restarted successfully.
wdcollectservice is stopped via systemctl after 90 seconds:
# time systemctl stop wdcollect.service
wdcollectservice status (after it was stopped) shows that the service corresponding process was killed after 90 seconds of the timeout:
# systemctl status wdcollect
● wdcollect.service - Watchdog wdcollect
Loaded: loaded (/lib/systemd/system/wdcollect.service; disabled; vendor preset: enabled)
Active: failed (Result: signal) since Mon 2018-06-04 13:29:24 +07; 9s ago
Jun 04 00:00:00 example.com systemd: Stopping Watchdog wdcollect...
Jun 04 00:01:31 example.com systemd: wdcollect.service: State 'stop-sigterm' timed out. Killing.
Jun 04 00:01:31 example.com systemd: wdcollect.service: Main process exited, code=killed, status=9/KILL
It is the bug #PPPM-8683 which will be fixed in future product updates.
As a workaround, perform the following actions:
Connect to the server via SSH.
Make a backup of the
# cp /lib/systemd/system/wdcollect.service /lib/systemd/system/wdcollect.service.back
/lib/systemd/system/wdcollect.servicefile via any text editor and make the following changes:
ExecStop=/bin/kill -9 $MAINPID
Reload the daemon to apply changes:
# systemctl daemon-reload