As an MSP we have multiple data centers with many remote probes. We know that remote probes will not switch over automatically to a backup core server so in order to achieve redundancy we are contemplating setting up a secondary standby core server in a backup data center and several times a day backing up the main core server to the secondary location.
When we need to switch to the secondary site we are thinking of restoring the database from backup and then changing the IP address associated with the remote probes so that they now point to the secondary data center server.
So my question is... of course will this work? Are there any gotchas we may have missed? Do you have anyone already doing this? Is there a better option?
Thank You!
Article Comments
OK that's great! I was not clear on the IP address change though. What I meant to say is the probes will be set to point to a DNS address "core.server.com" of the core server. When it comes time to switch to the secondary we will make a single change to the DNS A record to point to a new IP address. The probes will still point to the DNS address "core.server.com" so the A record change should just be picked up with a probe restart...correct?
Thank You!
Oct, 2014 - Permalink
Yes, that should be correct. The Probes could also reconnect directly once the connection is possible to again (to the 2nd Core), with the dns-record re-registered.
Oct, 2014 - Permalink
Hello,
Thank you very much for using PRTG. This procedure should work, has one caveat though, which is when changing the IP, on the Remote Probes, to point them to the new/backup Core Server. The IP Change will mean a restart of the Probe, and thus they will loose their cached data (since the main Core went down).
best regards.
Oct, 2014 - Permalink