Symptom: After a SQL Server failover, Zabbix stopped receiving data. Manual zabbix_get worked, but the server showed “no data for 30m”.
Root cause: The ODBC DSN pointed to SQLCLUSTER01 (virtual name). After failover to node 2, the DSN resolved to the correct IP, but the Zabbix agent on node 2 had a different service account SID that lacked permissions. Failover did not replicate SQL logins. zabbix mssql failed to fetch info data -or no data for 30m-
Solution: Create the zabbix_svc login on all nodes of the cluster and grant permissions via a server-level trigger that executes on failover. Symptom: After a SQL Server failover, Zabbix stopped
Before diving into logs and permissions, verify the basics. If the service is stopped, Zabbix obviously cannot
If the service is stopped, Zabbix obviously cannot fetch data. Start it and check the logs for why it stopped (usually memory pressure or disk errors).
If the SQL service is running, the issue might be between the Zabbix Server and the Agent on the SQL host.
On the Zabbix Server, run:
zabbix_get -s <TARGET_IP> -k agent.ping