Github user tgravescs commented on a diff in the pull request: https://github.com/apache/spark/pull/14079#discussion_r77890828 --- Diff: core/src/main/scala/org/apache/spark/scheduler/BlacklistTracker.scala --- @@ -0,0 +1,393 @@ +/* + * 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.scheduler + +import java.util.concurrent.atomic.AtomicReference + +import scala.collection.mutable.{ArrayBuffer, HashMap, HashSet} + +import org.apache.spark.SparkConf +import org.apache.spark.internal.Logging +import org.apache.spark.internal.config +import org.apache.spark.util.{Clock, SystemClock, Utils} + +/** + * BlacklistTracker is designed to track problematic executors and nodes. It supports blacklisting + * executors and nodes across an entire application (with a periodic expiry). TaskSetManagers add + * additional blacklisting of executors and nodes for individual tasks and stages which works in + * concert with the blacklisting here. + * + * The tracker needs to deal with a variety of workloads, eg.: + * + * * bad user code -- this may lead to many task failures, but that should not count against + * individual executors + * * many small stages -- this may prevent a bad executor for having many failures within one + * stage, but still many failures over the entire application + * * "flaky" executors -- they don't fail every task, but are still faulty enough to merit + * blacklisting + * + * See the design doc on SPARK-8425 for a more in-depth discussion. + * + * THREADING: As with most helpers of TaskSchedulerImpl, this is not thread-safe. Though it is + * called by multiple threads, callers must already have a lock on the TaskSchedulerImpl. The + * one exception is [[nodeBlacklist()]], which can be called without holding a lock. + */ +private[scheduler] class BlacklistTracker ( + conf: SparkConf, + clock: Clock = new SystemClock()) extends Logging { + + BlacklistTracker.validateBlacklistConfs(conf) + private val MAX_FAILURES_PER_EXEC = conf.get(config.MAX_FAILURES_PER_EXEC) + private val MAX_FAILED_EXEC_PER_NODE = conf.get(config.MAX_FAILED_EXEC_PER_NODE) + val BLACKLIST_TIMEOUT_MILLIS = BlacklistTracker.getBlacklistTimeout(conf) + + /** + * A map from executorId to information on task failures. Tracks the time of each task failure, + * so that we can avoid blacklisting executors due to failures that are very far apart. We do not + * actively remove from this as soon as tasks hit their timeouts, to avoid the time it would take + * to do so. But it will not grow too large, because as soon as an executor gets too many + * failures, we blacklist the executor and remove its entry here. + */ + private[scheduler] val executorIdToFailureList: HashMap[String, ExecutorFailureList] = + new HashMap() + val executorIdToBlacklistStatus: HashMap[String, BlacklistedExecutor] = new HashMap() + val nodeIdToBlacklistExpiryTime: HashMap[String, Long] = new HashMap() + /** + * An immutable copy of the set of nodes that are currently blacklisted. Kept in an + * AtomicReference to make [[nodeBlacklist()]] thread-safe. + */ + private val _nodeBlacklist: AtomicReference[Set[String]] = new AtomicReference(Set()) + /** + * Time when the next blacklist will expire. Used as a + * shortcut to avoid iterating over all entries in the blacklist when none will have expired. + */ + private[scheduler] var nextExpiryTime: Long = Long.MaxValue + /** + * Mapping from nodes to all of the executors that have been blacklisted on that node. We do *not* + * remove from this when executors are removed from spark, so we can track when we get multiple + * successive blacklisted executors on one node. Nonetheless, it will not grow too large because + * there cannot be many blacklisted executors on one node, before we stop requesting more + * executors on that node, and we periodically clean up the list of blacklisted executors. + */ + val nodeToFailedExecs: HashMap[String, HashSet[String]] = new HashMap() + + def applyBlacklistTimeout(): Unit = { + val now = clock.getTimeMillis() + // quickly check if we've got anything to expire from blacklist -- if not, avoid doing any work + if (now > nextExpiryTime) { + // Apply the timeout to blacklisted nodes and executors + val execsToUnblacklist = executorIdToBlacklistStatus.filter(_._2.expiryTime < now).keys + if (execsToUnblacklist.nonEmpty) { + // Un-blacklist any executors that have been blacklisted longer than the blacklist timeout. + logInfo(s"Removing executors $execsToUnblacklist from blacklist because the blacklist " + + s"has timed out") + execsToUnblacklist.foreach { exec => + val status = executorIdToBlacklistStatus.remove(exec).get + val failedExecsOnNode = nodeToFailedExecs(status.node) + failedExecsOnNode.remove(exec) + if (failedExecsOnNode.isEmpty) { + nodeToFailedExecs.remove(status.node) + } + } + } + val nodesToUnblacklist = nodeIdToBlacklistExpiryTime.filter(_._2 < now).keys + if (nodesToUnblacklist.nonEmpty) { + // Un-blacklist any nodes that have been blacklisted longer than the blacklist timeout. + logInfo(s"Removing nodes $nodesToUnblacklist from blacklist because the blacklist " + + s"has timed out") + nodesToUnblacklist.foreach { node => nodeIdToBlacklistExpiryTime.remove(node) } + _nodeBlacklist.set(nodeIdToBlacklistExpiryTime.keySet.toSet) + } + updateNextExpiryTime() + } + } + + private def updateNextExpiryTime(): Unit = { + if (executorIdToBlacklistStatus.nonEmpty) { + nextExpiryTime = executorIdToBlacklistStatus.map{_._2.expiryTime}.min + } else { + nextExpiryTime = Long.MaxValue + } + } + + + def updateBlacklistForSuccessfulTaskSet( + stageId: Int, + stageAttemptId: Int, + failuresByExec: HashMap[String, ExecutorFailuresInTaskSet]): Unit = { + // if any tasks failed, we count them towards the overall failure count for the executor at + // this point. + val now = clock.getTimeMillis() + val expiryTime = now + BLACKLIST_TIMEOUT_MILLIS + failuresByExec.foreach { case (exec, failuresInTaskSet) => + val allFailuresOnOneExecutor = + executorIdToFailureList.getOrElseUpdate(exec, new ExecutorFailureList) + // Apply the timeout to individual tasks. This is to prevent one-off failures that are very + // spread out in time (and likely have nothing to do with problems on the executor) from + // triggering blacklisting. However, note that we do *not* remove executors and nodes from + // the blacklist as we expire individual task failures -- each have their own timeout. Eg., + // suppose: + // * timeout = 10, maxFailuresPerExec = 2 + // * Task 1 fails on exec 1 at time 0 + // * Task 2 fails on exec 1 at time 5 + // --> exec 1 is blacklisted from time 5 - 15. + // This is to simplify the implementation, as well as keep the behavior easier to understand + // for the end user. + allFailuresOnOneExecutor.dropFailuresWithTimeoutBefore(now) + allFailuresOnOneExecutor.addFailures(stageId, stageAttemptId, failuresInTaskSet) + val newTotal = allFailuresOnOneExecutor.numUniqueTaskFailures + + if (newTotal >= MAX_FAILURES_PER_EXEC) { + logInfo(s"Blacklisting executor id: $exec because it has $newTotal" + + s" task failures in successful task sets") + val node = failuresInTaskSet.node + executorIdToBlacklistStatus.put(exec, BlacklistedExecutor(node, expiryTime)) + executorIdToFailureList.remove(exec) + updateNextExpiryTime() + + // In addition to blacklisting the executor, we also update the data for failures on the + // node, and potentially put the entire node into a blacklist as well. + val blacklistedExecsOnNode = nodeToFailedExecs.getOrElseUpdate(node, HashSet[String]()) + blacklistedExecsOnNode += exec + if (blacklistedExecsOnNode.size >= MAX_FAILED_EXEC_PER_NODE) { + logInfo(s"Blacklisting node $node because it has ${blacklistedExecsOnNode.size} " + + s"executors blacklisted: ${blacklistedExecsOnNode}") + nodeIdToBlacklistExpiryTime.put(node, expiryTime) + _nodeBlacklist.set(nodeIdToBlacklistExpiryTime.keySet.toSet) + } + } + } + } + + def isExecutorBlacklisted(executorId: String): Boolean = { + executorIdToBlacklistStatus.contains(executorId) + } + + /** + * Get the full set of nodes that are blacklisted. Unlike other methods in this class, this *IS* + * thread-safe -- no lock required on a taskScheduler. + */ + def nodeBlacklist(): Set[String] = { + _nodeBlacklist.get() + } + + def isNodeBlacklisted(node: String): Boolean = { + nodeIdToBlacklistExpiryTime.contains(node) + } + + def handleRemovedExecutor(executorId: String): Unit = { + // We intentionally do not clean up executors that are already blacklisted in nodeToFailedExecs, + // so that if another executor on the same node gets blacklisted, we can blacklist the entire + // node. We also can't clean up executorIdToBlacklistStatus, so we can eventually remove + // the executor after the timeout. Despite not clearing those structures here, we don't expect + // they will grow too big since you won't get too many executors on one node, and the timeout + // will clear it up periodically in any case. + executorIdToFailureList -= executorId + } +} + + +private[scheduler] object BlacklistTracker extends Logging { + + private val DEFAULT_TIMEOUT = "1h" + + /** + * Returns true if the blacklist is enabled, based on checking the configuration in the following + * order: + * 1. Is it specifically enabled or disabled? + * 2. Is it enabled via the legacy timeout conf? + * 3. Use the default for the spark-master: + * - off for local mode + * - on for distributed modes (including local-cluster) + */ + def isBlacklistEnabled(conf: SparkConf): Boolean = { + conf.get(config.BLACKLIST_ENABLED) match { + case Some(isEnabled) => + isEnabled + case None => + // if they've got a non-zero setting for the legacy conf, always enable the blacklist, + // otherwise, use the default based on the cluster-mode (off for local-mode, on otherwise). + val legacyKey = config.BLACKLIST_LEGACY_TIMEOUT_CONF.key + conf.get(config.BLACKLIST_LEGACY_TIMEOUT_CONF) match { + case Some(legacyTimeout) => + if (legacyTimeout == 0) { + logWarning(s"Turning off blacklisting due to legacy configuaration:" + + s" $legacyKey == 0") + false + } else { + // mostly this is necessary just for tests, since real users that want the blacklist + // will get it anyway by default + logWarning(s"Turning on blacklisting due to legacy configuration:" + + s" $legacyKey > 0") + true + } + case None => + // local-cluster is *not* considered local for these purposes, we still want the + // blacklist enabled by default + !Utils.isLocalMaster(conf) + } + } + } + + def getBlacklistTimeout(conf: SparkConf): Long = { + conf.get(config.BLACKLIST_TIMEOUT_CONF).getOrElse { + conf.get(config.BLACKLIST_LEGACY_TIMEOUT_CONF).getOrElse { + Utils.timeStringAsMs(DEFAULT_TIMEOUT) + } + } + } + + /** + * Verify that blacklist configurations are consistent; if not, throw an exception. Should only + * be called if blacklisting is enabled. + * + * The configuration for the blacklist is expected to adhere to a few invariants. Default + * values follow these rules of course, but users may unwittingly change one configuration + * without making the corresponding adjustment elsewhere. This ensures we fail-fast when + * there are such misconfigurations. + */ + def validateBlacklistConfs(conf: SparkConf): Unit = { + + def mustBePos(k: String, v: String): Unit = { + throw new IllegalArgumentException(s"$k was $v, but must be > 0.") + } + + // undocumented escape hatch for validation -- just for tests that want to run in an "unsafe" + // configuration. + if (!conf.get("spark.blacklist.testing.skipValidation", "false").toBoolean) { + + Seq( + config.MAX_TASK_ATTEMPTS_PER_EXECUTOR, + config.MAX_TASK_ATTEMPTS_PER_NODE, + config.MAX_FAILURES_PER_EXEC_STAGE, + config.MAX_FAILED_EXEC_PER_NODE_STAGE, + config.MAX_FAILURES_PER_EXEC, + config.MAX_FAILED_EXEC_PER_NODE + ).foreach { config => + val v = conf.get(config) + if (v <= 0) { + mustBePos(config.key, v.toString) + } + } + + val timeout = getBlacklistTimeout(conf) + if (timeout <= 0) { + // first, figure out where the timeout came from, to include the right conf in the message. + conf.get(config.BLACKLIST_TIMEOUT_CONF) match { + case Some(t) => + mustBePos(config.BLACKLIST_TIMEOUT_CONF.key, timeout.toString) + case None => + mustBePos(config.BLACKLIST_LEGACY_TIMEOUT_CONF.key, timeout.toString) + } + } + + val maxTaskFailures = conf.getInt("spark.task.maxFailures", 4) + val maxNodeAttempts = conf.get(config.MAX_TASK_ATTEMPTS_PER_NODE) + + if (maxNodeAttempts >= maxTaskFailures) { --- End diff -- I don't really see a reason why to do this (=) but at the same time who knows what someone will come up with. It only disables that part of the blacklisting check as it could easily hit one of the other stage limits depending on how many executors per node you get, so worst case they don't get blacklisting and it fails the job like it did before. I would think we either do the check as is or just change the entire thing to just log a warning. Because if its = or its > doesn't matter, its going to hit the task failure limit first and fail the job rather then blacklisting.
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org