Add startup probes and relax probe timings for CI stability

Add startupProbe to porch-server, porch-controllers, and function-runner
to prevent premature pod kills during slow CI startup. Increase readiness
failureThreshold and periodSeconds to tolerate transient delays.

Also adds a livenessProbe to function-runner which previously had none.

Signed-off-by: Fiachra Corcoran <fiachra.corcoran@est.tech>

Author: efiacor

deployments/porch/2-function-runner.yaml

@@ -67,12 +67,34 @@ spec:
             - containerPort: 9464
               name: metrics
           # Add grpc readiness probe to ensure the cache is ready
+          startupProbe:
+            exec:
+              command:
+                - /home/nonroot/grpc-health-probe
+                - -addr
+                - localhost:9445
+            initialDelaySeconds: 5
+            periodSeconds: 5
+            failureThreshold: 30
           readinessProbe:
             exec:
               command:
                 - /home/nonroot/grpc-health-probe
                 - -addr
                 - localhost:9445
+            periodSeconds: 5
+            failureThreshold: 3
+            timeoutSeconds: 3
+          livenessProbe:
+            exec:
+              command:
+                - /home/nonroot/grpc-health-probe
+                - -addr
+                - localhost:9445
+            initialDelaySeconds: 30
+            periodSeconds: 30
+            failureThreshold: 3
+            timeoutSeconds: 5
           resources:
             requests:
               memory: "64Mi"

deployments/porch/3-porch-server.yaml

@@ -94,13 +94,19 @@ spec:
             - --repo-operation-retry-attempts=3
             - --max-request-body-size=6291456 # Keep this in sync with function-runner's corresponding argument
             - --cache-type=db
-          #adding livenessProbes and readinessProbes for porch server
+          startupProbe:
+            httpGet:
+              path: /healthz
+              port: 4443
+              scheme: HTTPS
+            initialDelaySeconds: 5
+            periodSeconds: 5
+            failureThreshold: 30
           livenessProbe:
             httpGet:
               path: /healthz
               port: 4443
               scheme: HTTPS
-            initialDelaySeconds: 30
             periodSeconds: 30
             failureThreshold: 3
             successThreshold: 1
@@ -110,11 +116,10 @@ spec:
               path: /readyz
               port: 4443
               scheme: HTTPS
-            initialDelaySeconds: 5
-            periodSeconds: 5
-            failureThreshold: 3
+            periodSeconds: 10
+            failureThreshold: 6
             successThreshold: 1
-            timeoutSeconds: 3                            
+            timeoutSeconds: 5                            
 ---
 apiVersion: v1
 kind: Service

deployments/porch/9-controllers.yaml

@@ -86,23 +86,29 @@ spec:
           limits:
             memory: 1Gi
             cpu: 1000m
+        startupProbe:
+          httpGet:
+            path: /healthz
+            port: 8081
+            scheme: HTTP
+          initialDelaySeconds: 5
+          periodSeconds: 5
+          failureThreshold: 30
         livenessProbe:
           failureThreshold: 3
           httpGet:
             path: /healthz
             port: 8081
             scheme: HTTP
-          initialDelaySeconds: 30
           periodSeconds: 30
           successThreshold: 1
           timeoutSeconds: 5
         readinessProbe:
-          failureThreshold: 3
+          failureThreshold: 6
           httpGet:
             path: /readyz
             port: 8081
             scheme: HTTP
-          initialDelaySeconds: 5
-          periodSeconds: 5
+          periodSeconds: 10
           successThreshold: 1
-          timeoutSeconds: 3            
+          timeoutSeconds: 5