Werk #22495: Redfish: an unavailable chassis no longer turns Check_MK CRIT
| Component | Checks & agents | ||
| Title | Redfish: an unavailable chassis no longer turns Check_MK CRIT | ||
| Date | Oct 29, 2026 | ||
| Level | Trivial Change | ||
| Class | Bug Fix | ||
| Compatibility | Compatible - no manual interaction needed | ||
| Checkmk versions & editions |
|
Since Werk #22290, the Check_MK
service of Redfish hosts could go CRIT with an error like
/redfish/v1/Chassis/2 is temporarily unavailable. This happened
whenever one of several chassis of a device was not ready to answer,
which HPE iLO 5, for example, reports intermittently. On some devices,
such as HPE iLO 4, Cisco CIMC or Supermicro, the same could happen for a
management controller.
The agent now ignores such a chassis or controller in that check cycle and monitors the rest as usual. If none of them can answer, the Check_MK service still goes CRIT.
Werk 22290 also caused services to lose their data for a check cycle if a part of the device, such as a chassis, a system or a storage controller, could not be queried. These services now keep their last known data for up to five minutes, provided the Check_MK service runs at least that often.
The Firmware health service no longer reports "All firmware in optimal state" if the firmware data could not be queried.