banshee/hardware-collection/hardware/redfish/https/health

Overall health of a server, by Redfish

Returns, for each computer system a server's management card serves by Redfish, its health and its power state, and, for each chassis, its health: the verdict the card rolls up from everything under it (power supplies, fans, temperatures, drives, memory, processors). Four watches are offered, each one set on the probe. It is the first probe to put on a server; the other Redfish collectors say which part is at fault.

Versions

Generic: written for the DMTF Redfish specification (DSP0266) and its ComputerSystem and Chassis schemas, whose Status and PowerState have kept their form since the first version (2016). Any management card that serves Redfish: Dell iDRAC, HPE iLO, Supermicro, Lenovo XCC, OpenBMC, in the firmware versions that serve Redfish (which firmware first does, per brand, is not checked here). Hence the common level.

How it works

The collector reads, the account in Basic:

GET https://<host:address>/redfish/v1/Systems, then each system it lists;
GET https://<host:address>/redfish/v1/Chassis, then each chassis it lists.

{"Id": "1", "PowerState": "On", "Status": {"State": "Enabled", "Health": "OK", "HealthRollup": "Warning"}}

An instance is named system <Id> or chassis <Id>: system 1, chassis System.Embedded.1.
health is HealthRollup when the card gives it, the health of the resource and of everything under it; Health otherwise; unknown when it gives neither. In lower case: ok, warning, critical.
power-state is Redfish's, in lower case: on, off, poweringon, poweringoff, paused. Only a system has one.
A chassis the card says Absent (an empty slot of an enclosure) is left out.
A code 401 or 403 stops the collector misconfigured (the account is refused); any other code but 200 stops it unreachable, with the path it was reading. A service that lists neither a system nor a chassis stops it misconfigured.

The password comes from protocol:http:password, a secret variable: it only goes in the Basic header. The answer is not secret. Redfish only reads with GET: nothing is changed on the card.

Prerequisites

The management card's address as the host's address, reachable from the poller in HTTPS (443). Its certificate follows the TLS profile of the probe: the card's self-signed certificate is trusted by pinning its fingerprint (protocol:http:fingerprint) or by adding its authority under protocol:tls:authorities.
A read-only account on the card, in protocol:http:user and protocol:http:password: the ReadOnly role of Redfish (iDRAC: Read Only; iLO: a user with the Login privilege only; XCC: ReadOnly; Supermicro: User). Redfish services accept Basic authentication, as the specification asks; a card where it was turned off refuses the account.

Outputs

Name	Label	Unit	Per instance
`health`	Health	—	yes, per component
`power-state`	Power state	—	yes, per component

Watches

Name	Title	Kind	Defaults	Checked by default
`unhealthy`	Health degraded	state to avoid, on `health`	warning on `warning`, critical on `critical`	yes
`unknown`	Health unknown	state to avoid, on `health`	warning on `unknown`	no
`powered-off`	System not powered on	expected value, on `power-state`	critical when the state is anything but `on`	yes
`missing`	Component missing	expected instances, on `health`	no component named: it raises nothing until names are set on the probe	no

A server kept off on purpose has powered-off unchecked on its probe.

Limits

The address is written into the URL as is: an IPv6 address must be given as a name.
A collection cut into pages (Members@odata.nextLink) is read on its first page only: a card lists its systems and chassis on one.
A JSON null reads as '' in Keen, which the collector takes for "not given": a card that writes "Health": null reads unknown, which raises nothing by default; a "PowerState": null is not returned.
Not checked against a card: the paths and fields are those of the DMTF Redfish schemas (ComputerSystem, Chassis, Resource.Status), read on redfish.dmtf.org.