[Bug] fix job silent status can't convert to lost status bugs

streampark-common/src/main/scala/org/apache/streampark/common/conf/K8sFlinkConfig.scala

@@ -28,6 +28,13 @@ object K8sFlinkConfig {
     classType = classOf[java.lang.Long],
     description = "run timeout seconds of single flink-k8s metrics tracking task")
 
+  @deprecated
+  val jobStatusTrackCacheTimeoutSec: InternalOption = InternalOption(
+    key = "streampark.flink-k8s.tracking.cache-timeout-sec.job-status",
+    defaultValue = 300,
+    classType = classOf[java.lang.Integer],
+    description = "status cache timeout seconds of single flink-k8s job status tracking task")
+
   @deprecated
   val metricTrackTaskTimeoutSec: InternalOption = InternalOption(
     key = "streampark.flink-k8s.tracking.polling-task-timeout-sec.cluster-metric",

streampark-flink/streampark-flink-kubernetes/src/main/scala/org/apache/streampark/flink/kubernetes/DefaultFlinkK8sWatcher.scala

@@ -33,7 +33,7 @@ class DefaultFlinkK8sWatcher(conf: FlinkTrackConfig = FlinkTrackConfig.defaultCo
 
   // cache pool for storage tracking result
   implicit val watchController: FlinkK8sWatchController =
-    new FlinkK8sWatchController()
+    new FlinkK8sWatchController(conf.jobStatusWatcherConf)
 
   // eventBus for change event
   implicit lazy val eventBus: ChangeEventBus = {

streampark-flink/streampark-flink-kubernetes/src/main/scala/org/apache/streampark/flink/kubernetes/FlinkK8sWatchController.scala

@@ -27,7 +27,7 @@ import java.util.Objects
 import java.util.concurrent.TimeUnit
 
 /** Tracking info cache pool on flink kubernetes mode. */
-class FlinkK8sWatchController extends Logger with AutoCloseable {
+class FlinkK8sWatchController(conf: JobStatusWatcherConfig = JobStatusWatcherConfig.defaultConf) extends Logger with AutoCloseable {
 
   // cache for tracking identifiers
   lazy val trackIds: TrackIdCache = TrackIdCache.build()
@@ -38,7 +38,7 @@ class FlinkK8sWatchController extends Logger with AutoCloseable {
   lazy val endpoints: EndpointCache = EndpointCache.build()
 
   // cache for tracking flink job status
-  lazy val jobStatuses: JobStatusCache = JobStatusCache.build()
+  lazy val jobStatuses: JobStatusCache = JobStatusCache.build(conf.jobStatusCacheTimeOutSec)
 
   // cache for tracking kubernetes events with Deployment kind
   lazy val k8sDeploymentEvents: K8sDeploymentEventCache =
@@ -156,10 +156,10 @@ object TrackIdCache {
   }
 }
 
-class JobStatusCache {
+class JobStatusCache(timeout: Int) {
 
   private[this] lazy val cache: Cache[CacheKey, JobStatusCV] =
-    Caffeine.newBuilder.expireAfterWrite(20, TimeUnit.SECONDS).build()
+    Caffeine.newBuilder.expireAfterWrite(timeout, TimeUnit.SECONDS).build()
 
   def putAll(kvs: Map[TrackId, JobStatusCV]): Unit =
     cache.putAll(kvs.map(t => (CacheKey(t._1.appId), t._2)))
@@ -183,7 +183,7 @@ class JobStatusCache {
 
 object JobStatusCache {
 
-  def build(): JobStatusCache = new JobStatusCache()
+  def build(timeout: Int): JobStatusCache = new JobStatusCache(timeout)
 
 }
 

streampark-flink/streampark-flink-kubernetes/src/main/scala/org/apache/streampark/flink/kubernetes/TrackConfig.scala

@@ -48,11 +48,14 @@ case class MetricWatcherConfig(requestTimeoutSec: Long, requestIntervalSec: Long
  *   interval seconds between two single tracking task
  * @param silentStateJobKeepTrackingSec
  *   retained tracking time for SILENT state flink tasks
+ * @param jobStatusCacheTimeOutSec
+ *   job status cache time out of single tracking task, must bigger than silentStateJobKeepTrackingSec
  */
 case class JobStatusWatcherConfig(
     requestTimeoutSec: Long,
     requestIntervalSec: Long,
-    silentStateJobKeepTrackingSec: Int)
+    silentStateJobKeepTrackingSec: Int,
+    jobStatusCacheTimeOutSec: Int)
 
 object FlinkTrackConfig {
   def defaultConf: FlinkTrackConfig =
@@ -66,7 +69,8 @@ object FlinkTrackConfig {
     JobStatusWatcherConfig(
       InternalConfigHolder.get(K8sFlinkConfig.jobStatusTrackTaskTimeoutSec),
       InternalConfigHolder.get(K8sFlinkConfig.jobStatueTrackTaskIntervalSec),
-      InternalConfigHolder.get(K8sFlinkConfig.silentStateJobKeepTrackingSec)),
+      InternalConfigHolder.get(K8sFlinkConfig.silentStateJobKeepTrackingSec),
+      InternalConfigHolder.get(K8sFlinkConfig.jobStatusTrackCacheTimeoutSec)),
     MetricWatcherConfig(
       InternalConfigHolder.get(K8sFlinkConfig.metricTrackTaskTimeoutSec),
       InternalConfigHolder.get(K8sFlinkConfig.metricTrackTaskIntervalSec)))
@@ -77,12 +81,14 @@ object JobStatusWatcherConfig {
   def defaultConf: JobStatusWatcherConfig = JobStatusWatcherConfig(
     requestTimeoutSec = 120,
     requestIntervalSec = 5,
-    silentStateJobKeepTrackingSec = 60)
+    silentStateJobKeepTrackingSec = 60,
+    jobStatusCacheTimeOutSec = 300)
 
   def debugConf: JobStatusWatcherConfig = JobStatusWatcherConfig(
     requestTimeoutSec = 120,
     requestIntervalSec = 2,
-    silentStateJobKeepTrackingSec = 5)
+    silentStateJobKeepTrackingSec = 5,
+    jobStatusCacheTimeOutSec = 30)
 }
 
 object MetricWatcherConfig {

streampark-flink/streampark-flink-kubernetes/src/main/scala/org/apache/streampark/flink/kubernetes/watcher/FlinkJobStatusWatcher.scala

@@ -128,11 +128,7 @@ class FlinkJobStatusWatcher(conf: JobStatusWatcherConfig = JobStatusWatcherConfi
                 case _ =>
                   touchSessionJob(trackId) match {
                     case Some(state) =>
-                      if (FlinkJobState.isEndState(state.jobState)) {
-                        // can't find that job in the k8s cluster.
-                        watchController.unWatching(trackId)
-                      }
-                      eventBus.postSync(FlinkJobStatusChangeEvent(trackId, state))
+                      updateState(trackId, state)
                     case _ =>
                   }
               }