Metric kube_pod_container_status_last_terminated_reason representation in Metricbeat

[sanjaruzic]

Describe the enhancement:
See the last termination reason of a running container in Metricbeat

Describe a specific use case for the enhancement or feature:

Metricbeat is deployed by the official Elastic Helm chart in conjunction with kube-state-metrics (all done by the Metricbeat Helm chart).

To find out the restart reason, the only metric available is kubernetes.container.status.reason
(docs).
This will only be reported in a Metricbeat's fetch when the container/pod is actually down. In incidents the customer experiences, the pod is restarting so fast, there is only a little chance to capture this event.

kube-state-metrics has the metric kube_pod_container_status_last_terminated_reason (added here) to see the last termination reason of a running container.

However, this metric is not represented/fetched in Metricbeat. The user wishes this feature as a part of Metricbeat.

[elasticmachine]

Pinging @elastic/integrations (Team:Integrations)