Add gRPC probe for agent (flyteorg#4990)

Signed-off-by: Kevin Su <[email protected]>
yubofredwang · Mar 26, 2024 · a400805 · a400805
1 parent bcd45cb
commit a400805
Show file tree

Hide file tree

Showing 7 changed files with 32 additions and 10 deletions.
diff --git a/charts/flyteagent/README.md b/charts/flyteagent/README.md
@@ -20,7 +20,7 @@ A Helm chart for Flyte agent
 | fullnameOverride | string | `""` |  |
 | image.pullPolicy | string | `"IfNotPresent"` | Docker image pull policy |
 | image.repository | string | `"ghcr.io/flyteorg/flyteagent"` | Docker image for flyteagent deployment |
-| image.tag | string | `"1.10.7"` | Docker image tag |
+| image.tag | string | `"1.10.8b4"` | Docker image tag |
 | nameOverride | string | `""` |  |
 | nodeSelector | object | `{}` | nodeSelector for flyteagent deployment |
 | podAnnotations | object | `{}` | Annotations for flyteagent pods |
@@ -30,6 +30,7 @@ A Helm chart for Flyte agent
 | ports.containerPort | int | `8000` |  |
 | ports.name | string | `"agent-grpc"` |  |
 | priorityClassName | string | `""` | Sets priorityClassName for datacatalog pod(s). |
+| readinessProbe | object | `{"grpc":{"port":8000},"initialDelaySeconds":1,"periodSeconds":3}` | https://kubernetes.io/blog/2022/05/13/grpc-probes-now-in-beta/#trying-the-feature-out |
 | replicaCount | int | `1` | Replicas count for flyteagent deployment |
 | resources | object | `{"limits":{"cpu":"500m","ephemeral-storage":"200Mi","memory":"200Mi"},"requests":{"cpu":"500m","ephemeral-storage":"200Mi","memory":"200Mi"}}` | Default resources requests and limits for flyteagent deployment |
 | securityContext | object | `{"allowPrivilegeEscalation":false}` | Security context for container |

diff --git a/charts/flyteagent/templates/agent/deployment.yaml b/charts/flyteagent/templates/agent/deployment.yaml
@@ -39,6 +39,10 @@ spec:
         ports:
         - containerPort: {{ .Values.ports.containerPort }}
           name: {{ .Values.ports.name }}
+        readinessProbe:
+        {{- with .Values.readinessProbe -}}
+          {{ tpl (toYaml .) $ | nindent 10 }}
+        {{- end }}
         securityContext:
           {{- toYaml .Values.securityContext | nindent 12 }}
         resources: {{- toYaml .Values.resources | nindent 10 }}

diff --git a/charts/flyteagent/values.yaml b/charts/flyteagent/values.yaml
@@ -23,7 +23,7 @@ image:
   # -- Docker image for flyteagent deployment
   repository: ghcr.io/flyteorg/flyteagent
   # -- Docker image tag
-  tag: 1.10.7 # FLYTEAGENT_TAG
+  tag: 1.10.8b4 # FLYTEAGENT_TAG
   # -- Docker image pull policy
   pullPolicy: IfNotPresent
 ports:
@@ -56,6 +56,13 @@ serviceAccount:
   imagePullSecrets: []
 # -- Security context for pod
 podSecurityContext: {}
+# -- Readiness probe for flyteagent. Use readinessProbe: {} if agent doesn't implement grpc-health-checking service.
+# -- https://kubernetes.io/blog/2022/05/13/grpc-probes-now-in-beta/#trying-the-feature-out
+readinessProbe:
+  grpc:
+    port: 8000
+  initialDelaySeconds: 1
+  periodSeconds: 3
 # -- Security context for container
 securityContext:
   allowPrivilegeEscalation: false

diff --git a/deployment/agent/flyte_agent_helm_generated.yaml b/deployment/agent/flyte_agent_helm_generated.yaml
@@ -78,7 +78,7 @@ spec:
         - pyflyte
         - serve
         - agent
-        image: "ghcr.io/flyteorg/flyteagent:1.10.7"
+        image: "ghcr.io/flyteorg/flyteagent:1.10.8b4"
         imagePullPolicy: "IfNotPresent"
         name: flyteagent
         volumeMounts:
@@ -87,6 +87,11 @@ spec:
         ports:
         - containerPort: 8000
           name: agent-grpc
+        readinessProbe:
+          grpc:
+            port: 8000
+          initialDelaySeconds: 1
+          periodSeconds: 3
         securityContext:
             allowPrivilegeEscalation: false
         resources:

diff --git a/docker/sandbox-bundled/manifests/complete-agent.yaml b/docker/sandbox-bundled/manifests/complete-agent.yaml
@@ -816,7 +816,7 @@ type: Opaque
 ---
 apiVersion: v1
 data:
-  haSharedSecret: NTRkQThQbW1FNDlVQ1B4Rw==
+  haSharedSecret: UVdVTnB4cXBMVXMyRjhGUw==
   proxyPassword: ""
   proxyUsername: ""
 kind: Secret
@@ -1412,7 +1412,7 @@ spec:
     metadata:
       annotations:
         checksum/config: 8f50e768255a87f078ba8b9879a0c174c3e045ffb46ac8723d2eedbe293c8d81
-        checksum/secret: 7ce58af0efd945f657619718c017f528d7055b8069cab01aed85e29639bd6d6a
+        checksum/secret: 7dd7be244652f0c2d4d5651db7e9d879fb74fb3a4407a80c48cba3dcd7b74e53
       labels:
         app: docker-registry
         release: flyte-sandbox
@@ -1755,12 +1755,17 @@ spec:
           value: minio
         - name: FLYTE_AWS_SECRET_ACCESS_KEY
           value: miniostorage
-        image: ghcr.io/flyteorg/flyteagent:1.10.7
+        image: ghcr.io/flyteorg/flyteagent:1.10.8b4
         imagePullPolicy: IfNotPresent
         name: flyteagent
         ports:
         - containerPort: 8000
           name: agent-grpc
+        readinessProbe:
+          grpc:
+            port: 8000
+          initialDelaySeconds: 1
+          periodSeconds: 3
         resources:
           limits:
             cpu: 500m

diff --git a/docker/sandbox-bundled/manifests/complete.yaml b/docker/sandbox-bundled/manifests/complete.yaml
@@ -796,7 +796,7 @@ type: Opaque
 ---
 apiVersion: v1
 data:
-  haSharedSecret: c21mSkU1U0c0S0J0Q2lRRg==
+  haSharedSecret: cnBScDd2Y3Y0a2JlcHlzVQ==
   proxyPassword: ""
   proxyUsername: ""
 kind: Secret
@@ -1360,7 +1360,7 @@ spec:
     metadata:
       annotations:
         checksum/config: 8f50e768255a87f078ba8b9879a0c174c3e045ffb46ac8723d2eedbe293c8d81
-        checksum/secret: 67f0b60e6520f3ae46a9f396411386d473441e7733c2544f6bface0f3b8475d9
+        checksum/secret: 459fe43b90e7be7fce10bc94020ba94037c1a21501b61d4e7121c84251f1b198
       labels:
         app: docker-registry
         release: flyte-sandbox

diff --git a/docker/sandbox-bundled/manifests/dev.yaml b/docker/sandbox-bundled/manifests/dev.yaml
@@ -499,7 +499,7 @@ metadata:
 ---
 apiVersion: v1
 data:
-  haSharedSecret: YjFqbVE2UE05QnpwMVVVeg==
+  haSharedSecret: NlJkVzYyZ1MzZmZIRE1nOA==
   proxyPassword: ""
   proxyUsername: ""
 kind: Secret
@@ -934,7 +934,7 @@ spec:
     metadata:
       annotations:
         checksum/config: 8f50e768255a87f078ba8b9879a0c174c3e045ffb46ac8723d2eedbe293c8d81
-        checksum/secret: ec699a25c688a15358899da82f9fa9bb199fe085f200f00facc09b448ad80369
+        checksum/secret: 2e417b7dd337346d064d1cd4edf6b5624d335fce92ccfe22acd20485be450ec9
       labels:
         app: docker-registry
         release: flyte-sandbox