Hello guys,
I'm running Zabbix 5.0.24 with 30 proxies, a PostgreSQL + TimescaleDB backend with a 2 To database (14000 hosts, 1,1 millions of items).
When I perform maintenance tasks on Zabbix which require to restart the zabbix-server process (minor upgrade for example), the database is overloaded by statements like "UPDATE trends_uint SET...". As the database is fully busy by these requests, no new data are inserted to the history tables during at least an hour and it creates a lot of chaos in my monitoring system.
The documentation says:
So if I understand correctly, should I stop Zabbix at 6:58 for example, and start it at 7:02 in order to avoid overlap? The last time I did this, I fell again in the "UPDATE" situation (screenshot below).
Am I missing something? What's the good process to restart Zabbix without this kind of issue please?
Thank you!
I'm running Zabbix 5.0.24 with 30 proxies, a PostgreSQL + TimescaleDB backend with a 2 To database (14000 hosts, 1,1 millions of items).
When I perform maintenance tasks on Zabbix which require to restart the zabbix-server process (minor upgrade for example), the database is overloaded by statements like "UPDATE trends_uint SET...". As the database is fully busy by these requests, no new data are inserted to the history tables during at least an hour and it creates a lot of chaos in my monitoring system.
The documentation says:
When server flushes trend cache and there are already trends in the database for this hour (for example, server has been restarted mid-hour), server needs to use update statements instead of simple inserts. Therefore on a bigger installation if restart is needed it is desirable to stop server in the end of one hour and start in the beginning of the next hour to avoid trend data overlap.
Am I missing something? What's the good process to restart Zabbix without this kind of issue please?
Thank you!
Comment