Hi everyone,
I've been struggling trying to set up Zenoss for two sites with these two constraints :
1) Geographic redundancy : We need to have the Zenoss service up even if one of the two sites has an outage
2) Site Separation (security) : The Zenoss instance(s) present in site 1 should not collect metrics from site 2 and vice versa.
An anwser to the first constraint would be to use DRBD for database sync and Corosync/Pacemaker for failover.
As for the second constraint (without taking into account the first one), the distributed collector configuration is enough.
It gets tricky when you try to combine both of them.
I am thinking about testing a configuration which consists of 2 Zenoss instances (in active/passive mode) and 2 collectors. 1 collector and Zenoss instance in each site. The master (as opposed to collector) Zenoss instances use a floating IP address and DRBD/Corosync/Pacemaker like the first described case. I am about to test this configuration. At first glance, this seems to be OK...
I tried another option, which would have been neat if it had worked. I tried using a configuration somewhat like the distributed collector one. Then, via Corosync/Pacemaker, starting zenhub/zopectl... on the collector once the master Zenoss instance was down. I got stuck tried to get zenhub start on the collector manually with no success...
So, what would be the best course of action for this situation. Any advice would be most welcome.
Thanks for your time.