PostgreSQL + Timescale DB crashed down by backup

[Jencek123]

Hello Zabbix Community,

I would like to ask If somebody else face this issue before? Our Zabbix server were used for small monitoring environment. All components were optimized and everything went well. Untill scheduled backup, database crashed big style. You can see logs from zabbix.server, zabbix.agent or postgresql:


PostgreSQL is in version 13. We tested all options like pg-dump, optimalization of database tunning timescaledb-tunning. All seems ok. After 1-2 days "Zabbix escalator processes more than 75% busy" trigger shows and server is unusable. Is there anything how to revert database to unfailed state?

Best Regards,
Jan

[splitek]

Recover from backup if you have good one.
I wonder how database can crash from doing backup? Script killed DB process or what? This: "database system was not properly shut down" suggest that maybe something happened in operating system... like power off machine or someone run "kill -9"

[Jencek123]

I am not sure about backup procedure. It is not up to me. But there is used Veeam. And for particular server you can choose extra database setup for backup procedure. On this one server it was empty... So I expect that Veeam ask for proper shutup and than kill it by force