Question 46347 in Debian Question posted in
The oficial documentation can be found here.

Understand the reasons why Mariadb has been stopped and which is the best way to retablish the service – Debian

I have 2 servers installed in 2 differents localisation, with a synchronisation files managed by heartbeat + drbd. They are running since few months and the date of the last main server reboot is : 2022-12-07 15:44

This night on the main server, the Mariadb service has been stopped. See below the logs :

Dec 18 06:50:14 nodegra systemd[1]: Starting Daily apt upgrade and clean activities...
Dec 18 06:50:20 nodegra pacemaker-controld[1131]:  notice: High CPU load detected: 1.370000
Dec 18 06:50:50 nodegra pacemaker-controld[1131]:  notice: High CPU load detected: 1.620000
Dec 18 06:51:20 nodegra pacemaker-controld[1131]:  notice: High CPU load detected: 1.980000
Dec 18 06:51:50 nodegra pacemaker-controld[1131]:  notice: High CPU load detected: 2.120000
Dec 18 06:52:20 nodegra pacemaker-controld[1131]:  notice: High CPU load detected: 2.210000
Dec 18 06:52:36 nodegra systemd[1]: Stopping MariaDB 10.5.15 database server...
Dec 18 06:52:36 nodegra mariadbd[1949]: 2022-12-18  6:52:36 0 [Note] /usr/sbin/mariadbd (initiated by: unknown): Normal shutdown
Dec 18 06:52:36 nodegra mariadbd[1949]: 2022-12-18  6:52:36 0 [Note] Event Scheduler: Purging the queue. 0 events
Dec 18 06:52:36 nodegra kayelles.service[1967382]: Warning: got packets out of order. Expected 2 but received 1
Dec 18 06:52:37 nodegra kayelles.service[1967382]: Warning: got packets out of order. Expected 2 but received 1
Dec 18 06:52:37 nodegra kayelles.service[1967382]: Warning: got packets out of order. Expected 2 but received 1
Dec 18 06:52:37 nodegra kayelles.service[1967382]: Warning: got packets out of order. Expected 2 but received 1
Dec 18 06:52:37 nodegra kayelles.service[1967382]: Warning: got packets out of order. Expected 15 but received 1
Dec 18 06:52:37 nodegra mariadbd[1949]: 2022-12-18  6:52:37 0 [Note] InnoDB: FTS optimize thread exiting.
Dec 18 06:52:37 nodegra kayelles.service[1967382]: db error Error: Connection was killed

I don’t find/understand the reason of the MariaDb service stop.
Is there a way to find the requester ?

Tags:

2

Answers


  1. Chosen as BEST ANSWER
    0

    Ths for your reply. Heartbeat log are empty fot this day and there is nothing special about in the pacemaker log :

    Dec 18 06:50:50 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 1.620000 Dec 18 06:51:20 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 1.980000 Dec 18 06:51:50 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 2.120000 Dec 18 06:52:20 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 2.210000 Dec 18 06:52:50 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 3.080000 Dec 18 06:53:20 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 5.630000 Dec 18 06:53:50 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 5.670000 Dec 18 06:54:20 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 6.120000 Dec 18 06:54:50 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 7.070000 Dec 18 06:55:20 nodeGRA pacemaker-controld [1131] (throttle_check_thresholds) notice: High CPU load detected: 7.360000

    I don't know what I have to check on more or that I have to set to avoid this issue again.


  2. 0

    "initiated by: unknown" is an indicator that it wasn’t a user doing SHUTDOWN and was as a result of a signal.

    Base on the close occurrence of the packmaker-control notice message I suspect it may have been the cause. Look closer at your pacemaker/heartbeat logs.

    Login or Signup to reply.
Please signup or login to give your own answer.

Back To Top
Search