You are on call. The Monster Market in wrecks-u585f1 is down, and five tickets are open at once. Triage before you touch anything:

kubectl get deploy,pods,svc,pvc -n wrecks-u585f1
kubectl events -n wrecks-u585f1 --types=Warning
kubectl describe rs -n wrecks-u585f1 | grep -B2 FailedCreate

The tickets, in the order they arrived:

1. Website: market-web Pods never start.
2. API: market-api Pods never start either, for a different reason.
3. Database: market-db is Pending.
4. Checkout: even with Pods up, market-api:8080 refuses connections.
5. Worker: market-worker has no Pods at all and no errors.

Every fix is one change in the layer the symptom points at. When all five are done, prove it from the client Pod:

kubectl exec client -n wrecks-u585f1 -- wget -qO- -T 2 market-api:8080/healthz