-1

I have one Nagios XI server monitoring many hosts in multiple data centers.

Now I want to setup DR in another datacenter for my primary Nagios XI.

Whats the best way to do this? Please share your experience and suggestions.

My research and constraints:

My network does not support corosync or pacemaker. So I want to do master-slave replication and Manual failover. How to manage Nagios XI with minimal dataloss.

RMK
  • 107
  • See https://assets.nagios.com/downloads/nagiosxi/docs/High-Availability-Options-For-Nagios-XI.pdf and https://assets.nagios.com/downloads/nagiosxi/docs/Monitoring-A-Nagios-XI-Server.pdf – Michael Hampton Feb 19 '18 at 02:44

1 Answers1

0

You can use DRBD for replication without Pacemaker. LINBIT has a guide that tells you exactly which directories need replicating in XI. If I recall correctly, you'll want to replicate the database(s), /usr/local/nagios, and /usr/local/nagiosxi. The rest is up to you.

I would add a DRBD device to the systems, make a filesystem on it, then move everything you want replicated to the DRBD device leaving symlinks where the files were originally.

rsync might cut it too, but I'm partial to DRBD.

Matt Kereczman
  • 1,899
  • 9
  • 12
  • Don't you need something like DRBD Proxy to run DRBD over a WAN, as the OP wants to do? – Michael Hampton Feb 19 '18 at 17:48
  • Thanks Matt. What about Nagios Xi database, can we replicate db also using drdb? Is it recommended? – RMK Feb 19 '18 at 18:31
  • DRBD Proxy helps over slow WAN links by buffering bursts of writes from whatever sits in top of DRBD. Depending on the environment and IO it may or may not be required. – Matt Kereczman Feb 20 '18 at 20:49
  • Database replication with DRBD gives you a crash consistent copy on the Secondary when the Primary fails without cleanly stopping the database. This is essentially the same thing you'd have in a single server environment after rebooting from a crashed server. – Matt Kereczman Feb 20 '18 at 20:52