Task · SKT-0087.07

Add a Redfish BMC hardware construct (inventory, health, power, thermal, events)

Description

Hardware faults, per-node power, firmware inventory and serial numbers come from the server BMC (for example HPE iLO, Dell iDRAC, or the BMCs in NVL72 compute and switch trays) over Redfish. Alloy cannot scrape Redfish JSON, so a Redfish exporter polls the BMC and a central collector scrapes the exporter, while the BMC also sends hardware events as remote syslog. Critical events need serial number, BMC hostname and rack attached so a fault arrives carrying what failed, where and how it is supported. There is no single canonical Redfish exporter, so the signal contract has to follow one real, maintained exporter.

Acceptance Criteria

Definition of Done

Implementation Notes

loop47: admitted R07 vendor-contract research; implementation waits for fixture/syslog landing

loop47: SAPCC exporter metric contract sourced, but required AC4 actual Dell iDRAC BMC event message format/examples unavailable after three official403 responses and bounded correction. Exporter metrics are not BMC syslog. Resume with authoritative family/version-pinned event manual or authorized capture and fixture/syslog dependencies; pin actual prometheus-client0.23.1 wire translation for demonstrated temperature name. Implementation0, infra0, review-repairs0, not rejected.

References

View the source file on GitHub