From 7f7b50ed9d4ffdd6b23e0faa56b068a049da67f7 Mon Sep 17 00:00:00 2001 From: Aaron Davidson Date: Thu, 16 Oct 2014 18:58:18 -0700 Subject: [SPARK-3923] Increase Akka heartbeat pause above heartbeat interval Something about the 2.3.4 upgrade seems to have made the issue manifest where all the services disconnect from each other after exactly 1000 seconds (which is the heartbeat interval). [This post](https://groups.google.com/forum/#!topic/akka-user/X3xzpTCbEFs) suggests that heartbeat pause should be greater than heartbeat interval, and increasing the pause from 600s to 6000s seems to have rectified the issue. My current cluster has now exceeded 1400s of uptime without failure! I do not know why this fixed it, because the threshold we have set for the failure detector is the exponent of a timeout, and 300 is extremely large. Perhaps the default failure detector changed in 2.3.4 and now ignores threshold. Author: Aaron Davidson Closes #2784 from aarondav/fix-timeout and squashes the following commits: bd1151a [Aaron Davidson] Increase pause, don't decrease interval 9cb0372 [Aaron Davidson] [SPARK-3923] Decrease Akka heartbeat interval below heartbeat pause --- core/src/main/scala/org/apache/spark/util/AkkaUtils.scala | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'core/src') diff --git a/core/src/main/scala/org/apache/spark/util/AkkaUtils.scala b/core/src/main/scala/org/apache/spark/util/AkkaUtils.scala index e2d32c859b..f41c8d0315 100644 --- a/core/src/main/scala/org/apache/spark/util/AkkaUtils.scala +++ b/core/src/main/scala/org/apache/spark/util/AkkaUtils.scala @@ -77,7 +77,7 @@ private[spark] object AkkaUtils extends Logging { val logAkkaConfig = if (conf.getBoolean("spark.akka.logAkkaConfig", false)) "on" else "off" - val akkaHeartBeatPauses = conf.getInt("spark.akka.heartbeat.pauses", 600) + val akkaHeartBeatPauses = conf.getInt("spark.akka.heartbeat.pauses", 6000) val akkaFailureDetector = conf.getDouble("spark.akka.failure-detector.threshold", 300.0) val akkaHeartBeatInterval = conf.getInt("spark.akka.heartbeat.interval", 1000) -- cgit v1.2.3