Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/21068#discussion_r185110988
--- Diff:
resource-managers/yarn/src/main/scala/org/apache/spark/deploy/yarn/YarnAllocatorBlacklistTracker.scala
---
@@ -0,0 +1,150 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.spark.deploy.yarn
+
+import scala.collection.JavaConverters._
+import scala.collection.mutable.HashMap
+
+import org.apache.hadoop.yarn.client.api.AMRMClient
+import org.apache.hadoop.yarn.client.api.AMRMClient.ContainerRequest
+
+import org.apache.spark.SparkConf
+import org.apache.spark.deploy.yarn.config._
+import org.apache.spark.internal.Logging
+import org.apache.spark.internal.config._
+import org.apache.spark.util.{Clock, SystemClock, Utils}
+
+/**
+ * YarnAllocatorBlacklistTracker is responsible for tracking the
blacklisted nodes
+ * and synchronizing these node list to YARN.
+ *
+ */
+private[spark] class YarnAllocatorBlacklistTracker(
+ sparkConf: SparkConf,
+ amClient: AMRMClient[ContainerRequest],
+ failureTracker: FailureTracker)
+ extends Logging {
+
+ private val defaultTimeout = "1h"
+
+ private val blacklistTimeoutMillis =
+
sparkConf.get(BLACKLIST_TIMEOUT_CONF).getOrElse(Utils.timeStringAsMs(defaultTimeout))
+
+ private val launchBlacklistEnabled =
+ sparkConf.get(YARN_EXECUTOR_LAUNCH_BLACKLIST_ENABLED).getOrElse(false)
+
+ private val maxFailuresPerHost = sparkConf.get(MAX_FAILED_EXEC_PER_NODE)
+
+ private val blacklistMaxNodeRatio =
+ sparkConf.get(YARN_BLACKLIST_MAX_NODE_BLACKLIST_RATIO)
+
+ private val allocatorBlacklist = new HashMap[String, Long]()
+
+ private var currentBlacklistedYarnNodes = Set.empty[String]
+
+ private var schedulerBlacklist = Map.empty[String, Long]
+
+ private var numClusterNodes = (Int.MaxValue /
blacklistMaxNodeRatio).toInt
+
+ def setNumClusterNodes(numClusterNodes: Int): Unit = {
+ this.numClusterNodes = numClusterNodes
+ }
+
+ def handleResourceAllocationFailure(hostOpt: Option[String]): Unit = {
+ hostOpt match {
+ case Some(hostname) =>
+ // failures on a already blacklisted nodes are not even tracked
+ // otherwise such failures could shutdown the application
+ // as resource requests are asynchronous
+ // and a late failure response could exceed MAX_EXECUTOR_FAILURES
+ if (!schedulerBlacklist.contains(hostname) &&
+ !allocatorBlacklist.contains(hostname)) {
+ failureTracker.registerFailureOnHost(hostname)
+ updateAllocationBlacklistedNodes(hostname)
+ }
+ case None =>
+ failureTracker.registerExecutorFailure()
+ }
+ }
+
+ private def updateAllocationBlacklistedNodes(hostname: String): Unit = {
+ if (launchBlacklistEnabled) {
+ val failuresOnHost = failureTracker.numFailuresOnHost(hostname)
+ if (failuresOnHost > maxFailuresPerHost) {
+ logInfo(s"blacklisting $hostname as YARN allocation failed
$failuresOnHost times")
+ allocatorBlacklist.put(
+ hostname,
+ failureTracker.clock.getTimeMillis() + blacklistTimeoutMillis)
+ refreshBlacklistedNodes()
+ }
+ }
+ }
+
+ def setSchedulerBlacklistedNodes(schedulerBlacklistedNodesWithExpiry:
Map[String, Long]): Unit = {
+ this.schedulerBlacklist = schedulerBlacklistedNodesWithExpiry
+ refreshBlacklistedNodes()
+ }
+
+ private def refreshBlacklistedNodes(): Unit = {
+ removeExpiredYarnBlacklistedNodes()
+ val limit = (numClusterNodes * blacklistMaxNodeRatio).toInt
+ val nodesToBlacklist =
+ if (schedulerBlacklist.size +
+ allocatorBlacklist.size > limit) {
+ mostRelevantSubsetOfBlacklistedNodes(limit)
+ } else {
+ schedulerBlacklist.keySet ++ allocatorBlacklist.keySet
+ }
+
+ synchronizeBlacklistedNodeWithYarn(nodesToBlacklist)
+ }
+
+ private def mostRelevantSubsetOfBlacklistedNodes(limit: Int) = {
+ val allBlacklist = schedulerBlacklist ++ allocatorBlacklist
+ val relevant =
+ allBlacklist.toSeq
+ .sortBy(_._2)(Ordering[Long].reverse)
+ .take(limit)
+ .map(_._1)
+ .toSet
+ logInfo(s"blacklist size limit ($limit) is reached, total count:
${allBlacklist.size}")
+ relevant
+ }
+
+ private def synchronizeBlacklistedNodeWithYarn(nodesToBlacklist:
Set[String]): Unit = {
+ // Update blacklist information to YARN ResourceManager for this
application,
+ // in order to avoid allocating new Containers on the problematic
nodes.
+ val additions = (nodesToBlacklist --
currentBlacklistedYarnNodes).toList.sorted
+ val removals = (currentBlacklistedYarnNodes --
nodesToBlacklist).toList.sorted
+ if (additions.nonEmpty) {
+ logInfo(s"adding nodes to YARN application master's blacklist:
$additions")
+ }
+ if (removals.nonEmpty) {
+ logInfo(s"removing nodes from YARN application master's blacklist:
$removals")
+ }
+ amClient.updateBlacklist(additions.asJava, removals.asJava)
+ currentBlacklistedYarnNodes = nodesToBlacklist
+ }
+
+ private def removeExpiredYarnBlacklistedNodes() = {
--- End diff --
`: Unit =`
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]