The source project of this merge request has been removed.
Update kube-state-metrics to prevent crashing
What does this MR do?
The Prometheus chart version was recently updated, bringing with it a newer version of kube-state-metrics
. This version, v1.9.1, has a bug causing it to crash almost immediately and never recover. Simply updating to a newer version fixes the issue. This MR causes the default to be v1.9.7 instead, which is the default in the most recent version of kube-state-metrics
that is used in the most recent version of the Prometheus chart in its old location.
FWIW I tested this in a cluster I own, and it worked beautifully.
Manual QA
-
Test this change using a Cluster management project