Page MenuHomePhabricator

Investigate high API latency on LW k8s
Open, Needs TriagePublic

Description

The alert KubeletOperationalLatency has been on warning level of and on, but more frequently in the last two weeks.

We should investigate:

a) whether the alert's limits are too tight
b) if the alert is indicating an actual problem, and what the root cause is.

The CPU throttling dashboard by Luca may be a start for looking into this: https://grafana-rw.wikimedia.org/d/Q1HD5X3Vk/elukey-k8s-throttling?orgId=1