CMK version:
we test both Checkmk Raw Edition 2.2.0p7 and Checkmk Enterprise Edition 2.2.0p11
OS version:
docker image
Error message:
none
Output of “cmk --debug -vvn hostname”: (If it is a problem with checks or plugins)
We tested dynamic host management with Enterprise edition for 4 Kubernetes clusters and were not satisfied with its performance.
We have 700 hosts and dynamic host management always adds and removes hosts even if configured not to do so.
Which is we believe causes 100% cpu usage.
If we compare it to our other checkmk instance with same specs (16gb ram, 4 cores) there are even less hosts ( not kubernetes hosts but normal ones like server, vms ) and the performance is very good.
The performance in checkmk for k8s is very bad which causes the instance to be very slowly.
We suspect the cause to be dynamic host management.
Our second test was with checkmk raw edition and
bash /opt/omd/sites/cmk/share/doc/check_mk/treasures/find_piggy_orphans
instead of dynamic host management.
The above script only outputs hosts but doesnt add them in checkmk.
Is it a bug or intended bevaviour?
Do we need to process the output to somehow add the k8s hosts to checkmk? (at least 200 hosts per k8s cluster)
Did the developers of checkmk test the performance of dynamic host management with multiple kubernetes clusters?