-
Incident report
-
Resolution: Won't fix
-
Major
-
None
-
4.2.7
-
Centos7 - Zabbix Server 4.2.2
We have a Zabbix environment that we migrated from version 3.0.5 to 4.2.2.
Just after the migration we noticed a very different behavior in housekeeper, it started running for almost 4 hours and we detected that not all history and trends data was being deleted.
Our approach to the problem was:
1) We stopped housekeeper and deleted data by manually;
2) We started housekeeper only for history and it started running for about 20/25 minutes and deleting all data needed.
3) We started housekeeper for history + trends and we had the same issue, housekeeping running almost 4 hours.
We found that after housekeeper starts it stays for almost 3 hours running this query: "select itemid,min(clock) from trends group by itemid"
After that query completes it starts deleting. We decided to disable housekeeper from trends and doing it manually. Is this is a known issue?
Note:
we have some partitioned tables on our MariaDB Zabbix Database:
'trends','trends_uint','history','history_uint','history_str'