CMK version: 2.2.0p33
OS version: RHEL 8.10
Error message:
Status information: 8 CRIT messages (Last worst: “2025-09-10 23:00:42.047000000:116:4:Transparent Huge Pages (THP) are activated or not readable on bd0hdb. M_HOST_INFORMATION.KEY: kernel.transparent_hugepage.enabled, M_HOST_INFORMATION.VALUE: “[always] madvise never”.”)
1/ I just started using the mk_sap_hana plugin about a week ago. It correctly detected that THP was still enabled on some of my HANA systems. I then fixed it using the SAP documented procedure “[always madvise [never]”, and even using SAP Studio, the alerts have gone away. But for some reason, Checkmk is still showing the alerts.
I found a suggestion via Google search to acknowledge the alert and then do a service discovery on the host, but that didn’t help. Also not sure why the “Duration” is still the date I enabled the check (9/11) while last check was about 45s ago today (9/18).
In the attached picture, it seems to run every min, but why is the age 7 days ago? For example, I disabled THP on these systems on 9/15.
2/ A related issue - my databases don’t use the same schedule for backups e.g., training DBs are backed up say twice a week, but it appears Checkmk default is to report any systems whose last successful backup is over a day ago. I did find a procedure on Google to deactivate the default check and then build a custom script around it that calls the default check, but I am yet to test it out. Thank you.
3/ Another related issue - the “Last Worst” alert for various issues (backup age, THP, certificates validity, memory usage, etc.) does not always go away even when the issue has been resolved?
