OOMKilled at 03:00

firefight

Pods keep getting OOMKilled. Diagnose on the node, raise the limit, verify the node survives.

mode

kubectl · bad deploy · difficulty 4

kubectl -n prod get pods -l app=api -o wide

oncall@ops-01:~$kubectl -n prod describe pod worker-uwy8j | grep -A5 'Last State'

oncall@ops-01:~$kubectl -n prod top pods --sort-by=memory

Enter submits a line · Tab then Enter restarts · Esc releases focus

Playing as a guest. Runs are saved against a device-local id — sign in with GitHub to keep your history and appear on the leaderboards.