Commits on Feb 19, 2021

1. jsonnet/rules: etcdInsufficientMembers is wrong when etcd is in a pod …

   The upstream etcd alert is incorrect because it only excludes instance,
   but OpenShift runs etcd in a pod and therefore the pod label must be
   excluded.
   
   Exclude the upstream alert, improve the resiliency of the alert
   expression, target the alert to the expected job for the cluster etcd
   (job="etcd"), update the description and health text to include a
   clearer description of what insufficient members means and consequences
   and some impact actions, and separate the alert into its own rule
   group to prepare (in the future) of moving the alert into the
   cluster-etcd-operator repo. The alert now includes
   etcd_server_has_leader == 1 to ensure that if an instance from a
   previous quorum appears we will not consider it part of the majority
   calculation. This also flags when we can't establish quorum due to
   failures in communication between nodes (but not between monitoring
   and etcd).

   

   smarterclayton authored and openshift-cherrypick-robot committed Feb 19, 2021
   Configuration menu

   • View commit details
   • Copy full SHA for 238a598
   • Browse repository at this point

   Copy the full SHA

   238a598 View commit details

   Browse the repository at this point in the history