In order to properly account subscription info, clusters must have a recording rule that captures the following data and reports it via telemetry, from 4.1 onwards. Clusters must report a metric that includes number of cores summed across nodes, subdivided by node-role label (e.g. master, infra, etc), os arch (kube label), os type (rhcos / rhel / other), and os version (7, 8, etc). This then must be back ported to 4.1.
The os_version label that was initially intended to be exposed was dropped with the following PR: https://github.com/openshift/machine-config-operator/pull/657 The following issue is a replacement to provide the OS major version in the future: https://github.com/openshift/machine-config-operator/issues/582 For the time being we are deferring the version until the above issue is implemented as it is not necessary for the subscription related metric we need. When the above OS major version label is added this metric may be updated, but it was decided not to pull in the os_image string to reduce parsing issues. More context of this discussion can be found here: https://github.com/openshift/cluster-monitoring-operator/pull/466#issuecomment-530094389
Note that for 4.1 backporting, we are waiting for a bugzilla to reference here: https://github.com/openshift/kube-state-metrics/pull/17. If there is no bugzilla in verified we cannot go ahead with 4.1. If that's ok and expected, then feel free to ignore this comment, I just want to ensure there is no cyclic dependency of people waiting on each other :) .
See https://bugzilla.redhat.com/show_bug.cgi?id=1749451#c10
Based on Comment 8 and Comment 10, move it to VERIFIED
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:2922