GKE: auth service LB health check

[viv-4]

23/7/22 master branch: Auth service fails the deployed LB health check of /, requires manual editing in GCP console to /auth/authority?health=true

Adding probes as below configures the LB health check for auth correctly on creation However, the container fails the probes and is restarted by kubernetes Removing the probes from the running deployment and leaving the LB health check as /auth/authority?health=true resolves Not defining probes on deployment/LB creation (current state) results in the container running fine but LB health check failing with / request Requires further investigation

          livenessProbe:
            httpGet:
              path: /auth/authority?health=true
              port: http
              scheme: HTTP
            timeoutSeconds: 1
            periodSeconds: 10
            successThreshold: 1
            failureThreshold: 3
          readinessProbe:
            httpGet:
              path: /auth/authority?health=true
              port: http
              scheme: HTTP
            timeoutSeconds: 1
            periodSeconds: 10
            successThreshold: 1
            failureThreshold: 3

[viv-4]

Resolved in https://github.com/place-labs/k8s-helm/commit/0ee4ebcb8cb03b7363b7e6a1f25996e87c6d6c51 As described above - the initial failures and restarts of the pod were unrelated