Redundancy in Zabbix servers

[crb]

Hi guys,

The last thread I could find on this issue was one year and two versions old, so I thought I'd see if the situation has changed.

We are currently testing Zabbix on one server, with a HA MySQL database.

Is there any way to have (preferably multi-master redundancy) in the server layer, so that alerts are reported even when the primary Zabbix server is offline? Do any of the distributed monitoring improvements help here?

It seems from the thread I linked that if we were to have the primary server listed first, it would be the only one that the agents sent active check results to, and we might be OK. Any thoughts?

Thanks!

[alixen]

Hi,

Since there is no redundancy built-in Zabbix, we have chosen to put the whole Zabbix server in a VM running on an HA cluster.

It would be better to have HA features within Zabbix since our solution experiences some downtime during VM restart on secondary server but it is still better than nothing.

Regards,
Alixen