We have a DRBD volume that is synchronized between 2 servers. The volume contains all the observium install and RRD’s. Mysql is set up to replicate the database. With this set up it is possible for the data on disk and the database to get out of synch with one another. That is the risk with this approach. You also need good monitoring of DRBD and Mysql to verify that your replication is working. This is definitely an active/passive type setup. The failover is a manual process to promote the DRBD and Mysql to the primary role on the backup server. If you have a good run book to perform the actions, this can be done in an automated manner.
There are lots of things to look out for, such as hostname references in the config, which can be fixed by using the /etc/hosts files. Each setup is fairly unique as there isn’t a standard way to do this, and everyone will make different choices about how to solve these problems.
On Sep 27, 2021, at 1:28 AM, Toky ANDRIANASOLO via observium observium@observium.org wrote:
Hello, we just installed a new second observium on a separated server located in another site to ensure monitoring service high availability. these are set with two different IP and the primary server is supposed to do system and Mysql replication to the second server every day so that we always have a same data level on both.
Could you please guide us on that?
Thank you and Regards,
Toky
observium mailing list observium@observium.org http://postman.memetic.org/cgi-bin/mailman/listinfo/observium