Hi all 🙂, when submitting jobs to flink session cluster, jobs changed to finished status after changing the flinksessionjob spec (JAR file), and jobs keep in finished status forever (logs : Resource fully reconciled, nothing to do...); when we submiting the same spec change to flink operator 1.3.1(operator image version), jobs status will change to finished firstly then change to running again with the updated spec( recovery from last savepoing); Can anybody share some info on how to achieving the same results as flink operator 1.3 with the latest flink operator. key flink cluster config:

spec:

flinkConfiguration: blob.server.port: “6124” fs.local.impl: org.apache.flink.core.fs.local.LocalFileSystem high-availability.jobmanager.port: “50010" high-availability.storageDir: s3://XXX high-availability.zookeeper.client.connection-timeout: “30000” high-availability.zookeeper.client.maxSessionTimeout: “30000" high-availability.zookeeper.client.session-timeout: “30000” jobmanager.archive.fs.dir: s3://XXX jobmanager.memory.heap.size: 512mb jobmanager.memory.jvm-metaspace.size: 256mb jobmanager.memory.jvm-overhead.max: 2000mb jobmanager.memory.jvm-overhead.min: 100mb jobmanager.memory.process.size: 1024mb jobmanager.rpc.port: “6123" kubernetes.operator.jm-deployment-recovery.enabled: “false” parallelism.default: “1" queryable-state.proxy.ports: “6125” recovery.mode: zookeeper recovery.zookeeper.path.namespace: /flink_operator recovery.zookeeper.path.root: /flink_iot_operator recovery.zookeeper.quorum: zookeeper-ingest-headless:2181 s3.access-key: XXX s3.secret-key: XXX state.backend: rocksdb state.backend.async: “true” state.backend.fs.memory-threshold: “1024” state.backend.fs.write-buffer-size: “4096" state.backend.incremental: “true” state.backend.local-recovery: “false” state.backend.rocksdb.checkpoint.transfer.thread.num: “1” state.backend.rocksdb.localdir: /opt/flink/data/rocksdb state.backend.rocksdb.metrics.actual-delayed-write-rate: “true” state.backend.rocksdb.metrics.background-errors: “true” state.backend.rocksdb.options-factory: org.apache.flink.contrib.streaming.state.DefaultConfigurableOptionsFactory state.backend.rocksdb.predefined-options: DEFAULT state.backend.rocksdb.timer-service.factory: HEAP state.backend.rocksdb.ttl.compaction.filter.enabled: “false” state.checkpoints.dir: s3://XXX state.checkpoints.num-retained: “100” state.savepoints.dir: s3://XXX taskmanager.memory.jvm-metaspace.size: 256mb taskmanager.memory.jvm-overhead.min: 64mb taskmanager.memory.managed.size: 512mb taskmanager.memory.process.size: 3328mb taskmanager.memory.task.heap.size: 2048mb taskmanager.memory.task.off-heap.size: 256mb taskmanager.numberOfTaskSlots: “4" taskmanager.rpc.port: “6122” taskmanager.state.local.root-dirs: /opt/flink/data/local-recovery flinkVersion: v1_14