fix: Add livenessProbe to workload pod (#124)

Adds the condition from integration tests as a `livenessProbe` to the workload container in order to ensure that the workload container restarts if it's not functional. Closes canonical/bundle-kubeflow#966
canonical · Aug 5, 2024 · a18b723 · a18b723
1 parent d8916ea
commit a18b723
Showing 1 changed file with 22 additions and 5 deletions.
diff --git a/src/charm.py b/src/charm.py
@@ -164,13 +164,15 @@ def set_pod_spec(self, event):
 
             interfaces = self._get_interfaces()
 
+            http_port = int(self.model.config["http-port"])
+
             image_details = self._check_image_details()
 
             upstreams = self._check_grpc(interfaces)
 
             self._send_info(interfaces)
 
-            self._send_data_to_ingress_provider(interfaces)
+            self._send_data_to_ingress_provider(interfaces, http_port)
 
         except CheckFailed as check_failed:
             self.model.unit.status = check_failed.status
@@ -190,7 +192,7 @@ def set_pod_spec(self, event):
             listeners=[
                 get_listener(
                     cluster=upstream["service"],
-                    port=int(self.model.config["http-port"]),
+                    port=http_port,
                 )
                 for upstream in upstreams
             ],
@@ -221,7 +223,7 @@ def set_pod_spec(self, event):
                             },
                             {
                                 "name": "http",
-                                "containerPort": int(self.model.config["http-port"]),
+                                "containerPort": http_port,
                             },
                         ],
                         "volumeConfig": [
@@ -236,6 +238,21 @@ def set_pod_spec(self, event):
                                 ],
                             }
                         ],
+                        "kubernetes": {
+                            "livenessProbe": {
+                                "initialDelaySeconds": 15,
+                                "httpGet": {
+                                    "path": "/",
+                                    "port": http_port,
+                                    "httpHeaders": [
+                                        {
+                                            "name": "Content-Type",
+                                            "value": "application/grpc-web-text",
+                                        }
+                                    ],
+                                },
+                            }
+                        },
                     }
                 ],
             },
@@ -282,7 +299,7 @@ def _check_grpc(self, interfaces):
             raise CheckFailed("Waiting for upstream gRPC connection information.", WaitingStatus)
         return upstreams
 
-    def _send_data_to_ingress_provider(self, interfaces):
+    def _send_data_to_ingress_provider(self, interfaces, port: int):
         """Send data to the ingress relation data bag so the VirtualServices provider configures
         a VirtualService routing traffic from `/ml_metadata` path to envoy service.
 
@@ -294,7 +311,7 @@ def _send_data_to_ingress_provider(self, interfaces):
                     "prefix": "/ml_metadata",
                     "rewrite": "/ml_metadata",
                     "service": self.model.app.name,
-                    "port": int(self.model.config["http-port"]),
+                    "port": port,
                 }
             )
         else: