shahidki31 commented on a change in pull request #26378: [SPARK-29724][SPARK-29726][WEBUI][SQL] Support JDBC/ODBC tab for HistoryServer WebUI URL: https://github.com/apache/spark/pull/26378#discussion_r351052205
########## File path: sql/hive-thriftserver/src/main/scala/org/apache/spark/sql/hive/thriftserver/ui/HiveThriftServer2Listener.scala ########## @@ -0,0 +1,318 @@ +/* + * 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.sql.hive.thriftserver.ui + +import java.util.concurrent.ConcurrentHashMap + +import scala.collection.JavaConverters._ +import scala.collection.mutable.ArrayBuffer + +import org.apache.hive.service.server.HiveServer2 + +import org.apache.spark.{SparkConf, SparkContext} +import org.apache.spark.internal.config.Status.LIVE_ENTITY_UPDATE_PERIOD +import org.apache.spark.scheduler._ +import org.apache.spark.sql.hive.thriftserver.HiveThriftServer2.ExecutionState +import org.apache.spark.sql.internal.SQLConf +import org.apache.spark.status.{ElementTrackingStore, KVUtils, LiveEntity} + +/** + * An inner sparkListener called in sc.stop to clean up the HiveThriftServer2 + */ +private[thriftserver] class HiveThriftServer2Listener( + kvstore: ElementTrackingStore, + sparkConf: SparkConf, + server: Option[HiveServer2], + live: Boolean = true) extends SparkListener { + + private val sessionList = new ConcurrentHashMap[String, LiveSessionData]() + private val executionList = new ConcurrentHashMap[String, LiveExecutionData]() + + private val (retainedStatements: Int, retainedSessions: Int) = { + (sparkConf.get(SQLConf.THRIFTSERVER_UI_STATEMENT_LIMIT), + sparkConf.get(SQLConf.THRIFTSERVER_UI_SESSION_LIMIT)) + } + + // How often to update live entities. -1 means "never update" when replaying applications, + // meaning only the last write will happen. For live applications, this avoids a few + // operations that we can live without when rapidly processing incoming task events. + private val liveUpdatePeriodNs = if (live) sparkConf.get(LIVE_ENTITY_UPDATE_PERIOD) else -1L + + // Returns true if this listener has no live data. Exposed for tests only. + private[thriftserver] def noLiveData(): Boolean = { + sessionList.isEmpty && executionList.isEmpty + } + + kvstore.addTrigger(classOf[SessionInfo], retainedSessions) { count => + cleanupSession(count) + } + + kvstore.addTrigger(classOf[ExecutionInfo], retainedStatements) { count => + cleanupExecutions(count) + } + + kvstore.onFlush { + if (!live) { + flush(updateStore(_, trigger = true)) + } + } + + override def onApplicationEnd(applicationEnd: SparkListenerApplicationEnd): Unit = { + if (live) { + server.foreach(_.stop()) + } + } + + override def onJobStart(jobStart: SparkListenerJobStart): Unit = { + val properties = jobStart.properties + if (properties != null) { + val groupId = properties.getProperty(SparkContext.SPARK_JOB_GROUP_ID) + if (groupId != null) { + updateJobDetails(jobStart.jobId.toString, groupId) + } + } + } + + /** + * This method is to handle out of order events. ie. if Job event come after execution end event. + * @param jobId + * @param groupId + */ + private def updateJobDetails(jobId: String, groupId: String): Unit = { + val execList = executionList.values().asScala.filter(_.groupId == groupId).toSeq + if (execList.nonEmpty) { + execList.foreach { exec => + exec.jobId += jobId.toString + updateLiveStore(exec) + } + } else { + // Here will come only if JobStart event comes after Execution End event. + val storeExecInfo = kvstore.view(classOf[ExecutionInfo]).asScala.filter(_.groupId == groupId) + storeExecInfo.foreach { exec => + val liveExec = getOrCreateExecution(exec.execId, exec.statement, exec.sessionId, + exec.startTimestamp, exec.userName) + liveExec.jobId += jobId.toString + updateStore(liveExec, trigger = true) + executionList.remove(liveExec.execId) + } + } + } + + override def onOtherEvent(event: SparkListenerEvent): Unit = { + event match { + case e: SparkListenerThriftServerSessionCreated => onSessionCreated(e) + case e: SparkListenerSessionClosed => onSessionClosed(e) + case e: SparkListenerThriftServerOperationStart => onOperationStart(e) + case e: SparkListenerThriftServerOperationParsed => onOperationParsed(e) + case e: SparkListenerThriftServerOperationCanceled => onOperationCanceled(e) + case e: SparkListenerThriftServerOperationError => onOperationError(e) + case e: SparkListenerThriftServerOperationFinish => onOperationFinished(e) + case e: SparkListenerThriftServerOperationClosed => onOperationClosed(e) + case _ => // Ignore + } + } + + private def onSessionCreated(e: SparkListenerThriftServerSessionCreated): Unit = { + val session = getOrCreateSession(e.sessionId, e.startTime, e.ip, e.userName) + sessionList.put(e.sessionId, session) + updateLiveStore(session) + } + + private def onSessionClosed(e: SparkListenerSessionClosed): Unit = { + val session = sessionList.get(e.sessionId) + session.finishTimestamp = e.finishTime + updateStore(session, trigger = true) + sessionList.remove(e.sessionId) + } + + private def onOperationStart(e: SparkListenerThriftServerOperationStart): Unit = { + val info = getOrCreateExecution( + e.id, + e.statement, + e.sessionId, + e.startTime, + e.userName) + + info.state = ExecutionState.STARTED + executionList.put(e.id, info) + sessionList.get(e.sessionId).totalExecution += 1 + executionList.get(e.id).groupId = e.groupId + updateLiveStore(executionList.get(e.id)) + updateLiveStore(sessionList.get(e.sessionId)) + } + + private def onOperationParsed(e: SparkListenerThriftServerOperationParsed): Unit = { + executionList.get(e.id).executePlan = e.executionPlan + executionList.get(e.id).state = ExecutionState.COMPILED + updateLiveStore(executionList.get(e.id)) + } + + private def onOperationCanceled(e: SparkListenerThriftServerOperationCanceled): Unit = { + executionList.get(e.id).finishTimestamp = e.finishTime + executionList.get(e.id).state = ExecutionState.CANCELED + updateLiveStore(executionList.get(e.id)) + } + + private def onOperationError(e: SparkListenerThriftServerOperationError): Unit = { + executionList.get(e.id).finishTimestamp = e.finishTime + executionList.get(e.id).detail = e.errorMsg + executionList.get(e.id).state = ExecutionState.FAILED + updateLiveStore(executionList.get(e.id)) + } + + private def onOperationFinished(e: SparkListenerThriftServerOperationFinish): Unit = { + executionList.get(e.id).finishTimestamp = e.finishTime + executionList.get(e.id).state = ExecutionState.FINISHED + updateLiveStore(executionList.get(e.id)) + } + + private def onOperationClosed(e: SparkListenerThriftServerOperationClosed): Unit = { + executionList.get(e.id).closeTimestamp = e.closeTime + executionList.get(e.id).state = ExecutionState.CLOSED + updateStore(executionList.get(e.id), trigger = true) + executionList.remove(e.id) + } + + // Update both live and history stores. If trigger is enabled, it will cleanup + // entity which exceeds the threshold. + def updateStore(entity: LiveEntity, trigger: Boolean = false): Unit = { Review comment: Updated ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
