SPARK-3612. Executor shouldn't quit if heartbeat message fails to reach ...

...the driver

Author: Sandy Ryza <[email protected]>

Closes apache#2487 from sryza/sandy-spark-3612 and squashes the following commits:

2b7353d [Sandy Ryza] SPARK-3612. Executor shouldn't quit if heartbeat message fails to reach the driver

Author: sryza

core/src/main/scala/org/apache/spark/executor/Executor.scala

@@ -24,6 +24,7 @@ import java.util.concurrent._
 
 import scala.collection.JavaConversions._
 import scala.collection.mutable.{ArrayBuffer, HashMap}
+import scala.util.control.NonFatal
 
 import org.apache.spark._
 import org.apache.spark.deploy.SparkHadoopUtil
@@ -375,12 +376,17 @@ private[spark] class Executor(
           }
 
           val message = Heartbeat(executorId, tasksMetrics.toArray, env.blockManager.blockManagerId)
-          val response = AkkaUtils.askWithReply[HeartbeatResponse](message, heartbeatReceiverRef,
-            retryAttempts, retryIntervalMs, timeout)
-          if (response.reregisterBlockManager) {
-            logWarning("Told to re-register on heartbeat")
-            env.blockManager.reregister()
+          try {
+            val response = AkkaUtils.askWithReply[HeartbeatResponse](message, heartbeatReceiverRef,
+              retryAttempts, retryIntervalMs, timeout)
+            if (response.reregisterBlockManager) {
+              logWarning("Told to re-register on heartbeat")
+              env.blockManager.reregister()
+            }
+          } catch {
+            case NonFatal(t) => logWarning("Issue communicating with driver in heartbeater", t)
           }
+
           Thread.sleep(interval)
         }
       }