feat(helm): add filesystem savepoint storage mode

Dockerfile

@@ -1,4 +1,4 @@
-FROM public.ecr.aws/docker/library/golang:1.23.4-bookworm AS build
+FROM public.ecr.aws/docker/library/golang:1.24.1-bookworm AS build
 
 ARG upx_version=4.2.4
 

Dockerfile.flink

@@ -27,6 +27,7 @@ RUN wget -q https://repo1.maven.org/maven2/org/apache/flink/flink-jdbc_2.12/1.10
 RUN wget -q https://repo1.maven.org/maven2/com/aventrix/jnanoid/jnanoid/2.0.0/jnanoid-2.0.0.jar -P /opt/flink/lib/
 RUN wget -q https://repo1.maven.org/maven2/org/apache/flink/flink-connector-jdbc/3.2.0-1.19/flink-connector-jdbc-3.2.0-1.19.jar -P /opt/flink/lib/ 
 RUN wget -q https://repo1.maven.org/maven2/org/apache/flink/flink-s3-fs-presto/1.20.1/flink-s3-fs-presto-1.20.1.jar -P /opt/flink/lib/ 
+RUN wget -q https://repo1.maven.org/maven2/com/github/luben/zstd-jni/1.5.7-2/zstd-jni-1.5.7-2.jar -P /opt/flink/lib/ 
 
 # Command to start Flink JobManager and TaskManager in a mini-cluster setup
 CMD ["bin/start-cluster.sh"]

go.sum

@@ -64,6 +64,8 @@ github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
 github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/logi-camp/go-flink-client v0.2.0 h1:PIyfJq7FjW28bnvemReCicIuQD7JzVgJDk2xPTZUS2s=
 github.com/logi-camp/go-flink-client v0.2.0/go.mod h1:A79abedX6wGQI0FoICdZI7SRoGHj15QwMwWowgsKYFI=
+github.com/logi-camp/go-flink-client v0.2.1 h1:STfKamFm9+2SxxfZO3ysdFsb5MViQdThB4UHbnkUOE8=
+github.com/logi-camp/go-flink-client v0.2.1/go.mod h1:A79abedX6wGQI0FoICdZI7SRoGHj15QwMwWowgsKYFI=
 github.com/mailru/easyjson v0.7.7 h1:UGYAvKxe3sBsEDzO8ZeWOSlIQfWFlxbzLZe7hwFURr0=
 github.com/mailru/easyjson v0.7.7/go.mod h1:xzfreul335JAWq5oZzymOObrkdz5UnU4kGfJJLY9Nlc=
 github.com/mattn/go-colorable v0.1.13 h1:fFA4WZxdEF4tXPZVKMLwD8oUnCTTo08duU7wxecdEvA=

helm/chart/Chart.yaml

@@ -2,9 +2,5 @@ apiVersion: v2
 name: flink-kube-operator
 description: Helm chart for flink kube operator
 type: application
-version: 1.0.0
+version: 1.2.0
 appVersion: "0.1.1"
-dependencies:
-  - name: minio
-    repository: https://charts.bitnami.com/bitnami
-    version: 16.0.2 

helm/chart/templates/flink/checkpoint-pvc.yaml

@@ -0,0 +1,12 @@
+{{- if eq .Values.flink.state.checkpoint.storageType "filesystem" }}
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: {{ .Release.Name }}-flink-checkpoint-pvc
+spec:
+  accessModes:
+    - ReadWriteMany
+  resources:
+    requests:
+      storage: {{ .Values.flink.state.checkpoint.size }}
+{{- end }}

helm/chart/templates/flink/config.yaml

@@ -18,14 +18,24 @@
     high-availability.type: kubernetes
     kubernetes.namespace: {{ .Release.Namespace }}
     kubernetes.cluster-id: {{ .Values.clusterId | default (print .Release.Name "-cluster") }}
-    execution.checkpointing.interval: {{ .Values.flink.checkpoint.interval }}
-    execution.checkpointing.mode: {{ .Values.flink.checkpoint.mode }}
+    execution.checkpointing.interval: {{ .Values.flink.state.checkpoint.interval }}
+    execution.checkpointing.mode: {{ .Values.flink.state.checkpoint.mode }}
+    {{- if eq .Values.flink.state.checkpoint.storageType "filesystem" }}
+    state.checkpoints.dir: file:///opt/flink/checkpoints/
+    {{- else if eq .Values.flink.state.checkpoint.storageType "s3" }}
     state.checkpoints.dir: s3://flink/checkpoints/
+    {{- end }}
     state.backend.rocksdb.localdir: /opt/flink/rocksdb
     high-availability.storageDir: /opt/flink/ha
+    {{- if eq .Values.flink.state.savepoint.storageType "filesystem" }}
+    state.savepoints.dir: file:///opt/flink/checkpoints/
+    {{- else if eq .Values.flink.state.savepoint.storageType "s3" }}
     state.savepoints.dir: s3://flink/savepoints/
+    {{- end }}
     state.backend.incremental: {{ .Values.flink.state.incremental }}
     rest.profiling.enabled: true
     s3.endpoint: http://{{ .Release.Name }}-minio:9000
     s3.path.style.access: true
+    {{- toYaml .Values.flink.properties | default "" | nindent 4 }}
+    
 {{- end }}

helm/chart/templates/flink/ha.pvc.yaml

@@ -1,7 +1,7 @@
 apiVersion: v1
 kind: PersistentVolumeClaim
 metadata:
-  name: {{ .Release.Name }}-{{ .Values.flink.state.ha.pvcName }}
+  name: {{ .Release.Name }}-flink-ha-pvc
 spec:
   accessModes:
     - ReadWriteOnce

helm/chart/templates/flink/job-manager-deploy.yaml

@@ -64,12 +64,28 @@ spec:
           volumeMounts:
             - name: flink-ha
               mountPath: {{ .Values.flink.state.ha.dir }}
-
+            {{- if eq .Values.flink.state.checkpoint.storageType "filesystem" }}
+            - name: flink-checkpoint
+              mountPath: /opt/flink/checkpoints
+            {{- end }}
+            {{- if eq .Values.flink.state.savepoint.storageType "filesystem" }}
+            - name: flink-savepoint
+              mountPath: /opt/flink/savepoint
+            {{- end }}
       volumes:
         - name: flink-ha
           persistentVolumeClaim:
-            claimName: {{ .Release.Name }}-{{ .Values.flink.state.ha.pvcName }}
-
+            claimName: {{ .Release.Name }}-flink-ha-pvc
+        {{- if eq .Values.flink.state.checkpoint.storageType "filesystem" }}
+        - name: flink-checkpoint
+          persistentVolumeClaim:
+            claimName: {{ .Release.Name }}-flink-checkpoint-pvc
+        {{- end }}
+        {{- if eq .Values.flink.state.savepoint.storageType "filesystem" }}
+        - name: flink-savepoint
+          persistentVolumeClaim:
+            claimName: {{ .Release.Name }}-flink-savepoint-pvc
+        {{- end }}
       {{- with .Values.nodeSelector }}
       nodeSelector:
         {{- toYaml . | nindent 8 }}

helm/chart/templates/flink/savepoint-pvc.yaml

@@ -0,0 +1,12 @@
+{{- if eq .Values.flink.state.savepoint.storageType "filesystem" }}
+apiVersion: v1
+kind: PersistentVolumeClaim
+metadata:
+  name: {{ .Release.Name }}-flink-savepoint-pvc
+spec:
+  accessModes:
+    - ReadWriteMany
+  resources:
+    requests:
+      storage: {{ .Values.flink.state.savepoint.size }}
+{{- end }}

helm/chart/templates/flink/task-manager-statefulset.yaml

@@ -43,11 +43,25 @@ spec:
             secretKeyRef:
               name: {{ .Release.Name }}-flink-secrets
               key: minio_secret_key
+        - name: FLINK_TASKMANAGER_HOST
+          valueFrom:
+            fieldRef:
+              fieldPath: spec.hostname
         volumeMounts:
         - name: rocksdb-storage
           mountPath: /opt/flink/rocksdb
+        {{- if eq .Values.flink.state.checkpoint.storageType "filesystem" }}
+        - name: flink-checkpoint
+          mountPath: /opt/flink/checkpoints
+        {{- end }}
         resources:
           {{- toYaml .Values.flink.taskManager.resources | nindent 10 }}
+      {{- if eq .Values.flink.state.checkpoint.storageType "filesystem" }}
+      volumes:
+        - name: flink-checkpoint
+          persistentVolumeClaim:
+            claimName: {{ .Release.Name }}-flink-checkpoint-pvc
+      {{- end }}
   volumeClaimTemplates:
   - metadata:
       name: rocksdb-storage

helm/chart/values.yaml

@@ -38,8 +38,7 @@ podAnnotations: {}
 # For more information checkout: https://kubernetes.io/docs/concepts/overview/working-with-objects/labels/
 podLabels: {}
 
-podSecurityContext: {}
-  # fsGroup: 2000
+podSecurityContext: {} # fsGroup: 2000
 
 securityContext: {}
   # capabilities:
@@ -106,7 +105,6 @@ autoscaling:
 config:
   flinkApiUrl: flink:8081
 
-
 nodeSelector: {}
 
 tolerations: []
@@ -122,10 +120,6 @@ flink:
   parallelism:
     default: 1 # Default parallelism for Flink jobs
 
-  checkpoint:
-    interval: 5min
-    mode: EXACTLY_ONCE
-
   state:
     backend: rocksdb # Use RocksDB for state backend
     incremental: true
@@ -133,15 +127,25 @@ flink:
       dir: "/opt/flink/ha" # Directory to store ha data
       pvcName: flink-ha-pvc # PVC for ha
       size: 10Gi # PVC size for ha
+    checkpoint:
+      storageType: s3 # s3 / filesystem
+      interval: 5min
+      mode: EXACTLY_ONCE
+      size: 8Gi
+    savepoint:
+      storageType: s3
+      size: 8Gi
 
   jobManager:
     processMemory: 4096m # Size of job manager process memory
 
+  properties:
+    jobmanager.rpc.timeout: 300s
 
   taskManager:
-    numberOfTaskSlots: 12  # Number of task slots for TaskManager
+    numberOfTaskSlots: 12 # Number of task slots for task manager
     processMemory: 4096m # Size of task manager process memory
-    replicas: 1
+    replicas: 1 # Number of task manager replicas
     storage:
       rocksDb:
         size: 4Gi

helm/index.yaml

@@ -1,6 +1,48 @@
 apiVersion: v1
 entries:
   flink-kube-operator:
+  - apiVersion: v2
+    appVersion: 0.1.1
+    created: "2025-05-17T12:47:25.848097207+03:30"
+    dependencies:
+    - name: minio
+      repository: https://charts.bitnami.com/bitnami
+      version: 16.0.2
+    description: Helm chart for flink kube operator
+    digest: 3458b9be97d2a4bcf8574706e44ea9f7fdeb11e83058a615566e6e094a51b920
+    name: flink-kube-operator
+    type: application
+    urls:
+    - flink-kube-operator-1.2.0.tgz
+    version: 1.2.0
+  - apiVersion: v2
+    appVersion: 0.1.1
+    created: "2025-04-15T12:06:59.425538953+03:30"
+    dependencies:
+    - name: minio
+      repository: https://charts.bitnami.com/bitnami
+      version: 16.0.2
+    description: Helm chart for flink kube operator
+    digest: 2b307a113476eebb34f58308bf1d4d0d36ca5e08fe6541f369a1c231ae0a71be
+    name: flink-kube-operator
+    type: application
+    urls:
+    - flink-kube-operator-1.1.2.tgz
+    version: 1.1.2
+  - apiVersion: v2
+    appVersion: 0.1.1
+    created: "2025-04-12T23:13:39.394371646+03:30"
+    dependencies:
+    - name: minio
+      repository: https://charts.bitnami.com/bitnami
+      version: 16.0.2
+    description: Helm chart for flink kube operator
+    digest: 14b08b443b4118cee4c279f62b498bc040b4a3e7ebafa8e195606e3d9b21810a
+    name: flink-kube-operator
+    type: application
+    urls:
+    - flink-kube-operator-1.0.1.tgz
+    version: 1.0.1
   - apiVersion: v2
     appVersion: 0.1.1
     created: "2025-04-06T01:52:09.478716316+03:30"
@@ -165,4 +207,4 @@ entries:
     urls:
     - flink-kube-operator-0.1.0.tgz
     version: 0.1.0
-generated: "2025-04-06T01:52:09.466886557+03:30"
+generated: "2025-04-15T12:06:59.397928815+03:30"

internal/managed_job/cycle.go

@@ -13,7 +13,7 @@ func (job *ManagedJob) Cycle() {
 	// pkg.Logger.Debug("[managed-job] [new] check cycle", zap.String("jobName", job.def.GetName()))
 
 	// Init job
-	if job.def.Status.LifeCycleStatus == "" && job.def.Status.JobStatus == "" {
+	if job.def.Status.LifeCycleStatus == "" && (job.def.Status.JobStatus == "" || job.def.Status.JobStatus == v1alpha1.JobStatusFinished) {
 		job.Run(false)
 		return
 	}

internal/managed_job/manager.go

@@ -117,14 +117,15 @@ func (mgr *Manager) cycle(client *api.Client, crdInstance *crd.Crd) {
 				"status": patchStatusObj,
 			})
 		} else {
-			patchStatusObj := map[string]interface{}{
-				"jobStatus":       "",
-				"lifeCycleStatus": string(v1alpha1.LifeCycleStatusFailed),
-			}
+			// TODO handle job not found status
+			// patchStatusObj := map[string]interface{}{
+			// 	"jobStatus":       "",
+			// 	"lifeCycleStatus": string(v1alpha1.LifeCycleStatusFailed),
+			// }
 
-			crdInstance.Patch(uid, map[string]interface{}{
-				"status": patchStatusObj,
-			})
+			// crdInstance.Patch(uid, map[string]interface{}{
+			// 	"status": patchStatusObj,
+			// })
 		}
 
 		managedJob.Cycle()