More High Availability Options and DR
Actually the only supported mechanism for HA is to use Checkmk appliance, that feature works fine but is only supported with Checkmk appliance. I would suggest to create more options to allow HA and DR approaches using the standard virtual machine that we all know and like. It would be great to use omd commands to manage that. The HA solution would work in the same way as the appliance and the DR solution would replicate the site to another location in a scheduled basis, it would also allow the failover for test purposes.
Comments: 3
-
16 Jan, '23
Lars SörensenCheckmk should support HA in all commercial versions. I could also imagine that this could be provided for a small fee as an optional module.
For the non-appliance, there should be clear instructions what has to be set up additionally on OS level and what has to be configured where and how.
The advantage of the appliance is of course that everything is already available out-of-the-box and can be easily activated via the menu. -
16 Jan, '23
AndyI'd like to add that the HA solution does not take into account the state of Checkmk it selves. If Apache becomes unresponsive, hangs or even if I stop it manually the HA process does nothing, it will continue to run the site. Adding HA into Checkmk that does not give any HA features would not make a whole lot of sense.
-
21 Jun, '23
Paulo SantanaThe new integration with Azure Blob Storage will assist in this task to have a central backup location combined with automation tools to automate omd tasks to create sites using data from our central backup.
Thanks for the clarification Lars.
Andy to assist on that we are using a separate site for health checks on all checkmk instances, later we will integrate that instance with RobotMK to test the UI.