Hi All, I have a 2-node HANA (pacemaker cluster) running on Linux. After pushing the checkmk agent to both nodes, I could see all the services on each node including the pacemaker ones. But I just realized after a recent failover I need to take further steps for Checkmk to actually treat both nodes as part of a single cluster. The only reference I could find was for configuring an MS MSSQL cluster.
Anyone have a detailed guide for setting up CheckMK to monitor a HANA cluster on pacemaker? I have using the raw/free CheckMK version. Thanks!
Hi,
Basically you need a rule ‘Clustered Services’ and add all services except the SAP HANA instance Status to this rule:
Then create the cluster host with ‘Add cluster’ and add your both nodes to the Nodes of this object. In Discovered Service the SAP HANA services should appear on the cluster host object and they disappear on the nodes.
Hope that helps
Mike
Hi Mike, please can you share the full screenshot of the rule you shared above? You can cross out identifiable hostnames etc. I don’t see how you got “Host tag: Database is SAP HANA” if I attempt to add a ”Clustered services” rule. Thank you.
Basically you need this for all the clustered services to define the behavior
:
And the clustered service rule to add the service to the cluster object:
In case you run active DB´s on both cluster nodes use ‘Clustered services for overlapping clusters’ instead.
I hope that helps
regards
Michael
1 Like