abstractdog commented on code in PR #5613: URL: https://github.com/apache/hive/pull/5613#discussion_r1937613702
########## ql/src/java/org/apache/hadoop/hive/ql/queryhistory/QueryHistoryService.java: ########## @@ -0,0 +1,238 @@ +/* + * 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.hadoop.hive.ql.queryhistory; + +import java.util.Queue; +import java.util.concurrent.CompletableFuture; +import java.util.concurrent.ExecutorService; +import java.util.concurrent.Executors; +import java.util.concurrent.LinkedBlockingQueue; +import java.util.concurrent.ScheduledExecutorService; +import java.util.concurrent.TimeUnit; + +import com.google.common.annotations.VisibleForTesting; +import org.apache.hadoop.hive.conf.HiveConf; +import org.apache.hadoop.hive.ql.DriverContext; +import org.apache.hadoop.hive.ql.ServiceContext; +import org.apache.hadoop.hive.ql.queryhistory.repository.QueryHistoryRepository; +import org.apache.hadoop.hive.ql.queryhistory.schema.Record; +import org.apache.hadoop.hive.ql.queryhistory.schema.Schema; +import org.apache.hadoop.hive.ql.session.SessionState; +import org.apache.hadoop.util.Time; +import org.apache.hive.common.util.ReflectionUtil; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import com.google.common.util.concurrent.ThreadFactoryBuilder; + +public class QueryHistoryService { + private static final Logger LOG = LoggerFactory.getLogger(QueryHistoryService.class); + + private static QueryHistoryService INSTANCE = null; + + protected final Queue<Record> queryHistoryQueue = new LinkedBlockingQueue<>(); + + private final ExecutorService flushExecutor = Executors.newFixedThreadPool(1, + new ThreadFactoryBuilder().setNameFormat("QueryHistoryService repository thread").setDaemon(true).build()); + private final ScheduledExecutorService periodicFlushExecutor = Executors.newScheduledThreadPool(1, + new ThreadFactoryBuilder().setNameFormat("QueryHistoryService periodic flush check").setDaemon(true).build()); + + private final HiveConf conf; + private final ServiceContext serviceContext; + + private final int batchSize; + private final int maxQueueSizeInMemoryBytes; + private final int flushInterval; + private final QueryHistoryRepository repository; + + public QueryHistoryService(HiveConf inputConf, ServiceContext serviceContext) { + LOG.info("Creating QueryHistoryService instance"); + + this.conf = new HiveConf(inputConf); + this.serviceContext = serviceContext; + + batchSize = HiveConf.getIntVar(conf, HiveConf.ConfVars.HIVE_QUERY_HISTORY_BATCH_SIZE); + maxQueueSizeInMemoryBytes = HiveConf.getIntVar(conf, HiveConf.ConfVars.HIVE_QUERY_HISTORY_MAX_MEMORY_BYTES); + flushInterval = HiveConf.getIntVar(conf, HiveConf.ConfVars.HIVE_QUERY_HISTORY_FLUSH_INTERVAL_SECONDS); + repository = createRepository(conf); + + printConfigInformation(); + + LOG.info("QueryHistoryService created [batchSize: {}, maxQueueSizeInMemoryBytes: {}]", + batchSize, maxQueueSizeInMemoryBytes); + + // keeping a global, static instance is usually considered an antipattern, even in case of using the Singleton + // design, this time the constraint was the ability to hook into the query processing (Driver) from somewhere the + // outer service (HiveServer2, QTestUtil, etc.), which otherwise would have needed passing a service instance + // through many layers, making the code and method signatures more complicated + INSTANCE = this; + } + + /** + * Start method actually starts the service in a sense that it initializes the repository (which might involve + * database operations) and starts the executor that's responsible for flushing records. + */ + public void start() { + repository.init(conf, new Schema()); + + try { + flushExecutor.submit(() -> { + // this session won't need tez, let's not mess with starting a TezSessionState + conf.setBoolVar(HiveConf.ConfVars.HIVE_CLI_TEZ_INITIALIZE_SESSION, false); + + // localize this conf for the thread that will be running the repository, + // as iceberg integration heavily relies on the session state conf + SessionState session = SessionState.start(conf); + LOG.info("Session for QueryHistoryService started, sessionId: {}", session.getSessionId()); + }); + } catch (Exception e) { + throw new RuntimeException("Failed to start QueryHistoryService", e); + } + + if (flushInterval > 0) { + // only flush, don't wait, this is a background task + periodicFlushExecutor.scheduleAtFixedRate(this::doFlush, 0, flushInterval, TimeUnit.SECONDS); + } + } + + public static QueryHistoryService getInstance() { + if (INSTANCE == null) { + throw new RuntimeException("QueryHistoryService is not present when asked for the singleton instance"); + } + return INSTANCE; + } + + private void printConfigInformation() { + if (batchSize == 0) { + LOG.info("Query History batch size is 0: every record will be flushed synchronously after the query" + + " (not recommended)"); + } else { + LOG.info("Query History batch size is {}: records will be flushed after every {} queries", + batchSize, batchSize); + } + if (maxQueueSizeInMemoryBytes == 0) { + LOG.info("Query History max queue memory byte size is 0: the service won't check the consumed memory to make " + + "flush decision (not recommended)"); + } else { + LOG.info("Query History max queue memory byte size is {}: records will be flushed when the queue consumes " + + "more than {} bytes in the memory", + maxQueueSizeInMemoryBytes, maxQueueSizeInMemoryBytes); + } + + if (flushInterval == 0) { + LOG.info("Query History flush interval size is 0: the service won't check for records to flush periodically"); + } else { + LOG.info("Query History flush interval is {}s: service will attempt to flush records in every {}s", + flushInterval, flushInterval); + } + } + + private QueryHistoryRepository createRepository(HiveConf conf) { + try { + return (QueryHistoryRepository) ReflectionUtil.newInstance( + conf.getClassByName(conf.get(HiveConf.ConfVars.HIVE_QUERY_HISTORY_REPOSITORY_CLASS.varname)), conf); + } catch (ClassNotFoundException e) { + throw new RuntimeException(e); + } + } + + /** + * Creates a Query History record from a given DriverContext and adds it to the queue. + * This is done asynchronously to completely decouple from the threads that run the query to make this service work + * silently in the background. + * @param driverContext a DriverContext containing all the needed information for a record to be created + */ + public void logQuery(DriverContext driverContext) { + long recordCreateStart = Time.monotonicNow(); + Record record = createRecord(driverContext); + long recordCreateEnd = Time.monotonicNow(); + queryHistoryQueue.add(record); + LOG.info("Created query history record in {}ms, current queue size: {}", + recordCreateEnd - recordCreateStart, queryHistoryQueue.size()); + + CompletableFuture<?> flushFuture = flushRecordsIfNeeded(); + maybeWaitForRecordsToBeFlushed(batchSize == 0, flushFuture); + } + + private CompletableFuture<?> flushRecordsIfNeeded() { + if (needToFlush()) { + return doFlush(); + } else { + LOG.debug("Not flushing history records yet, current queue size: {}, batchSize: {}", + queryHistoryQueue.size(), batchSize); + return CompletableFuture.completedFuture(null); + } + } + + private void maybeWaitForRecordsToBeFlushed(boolean shouldWait, CompletableFuture<?> flushFuture) { Review Comment: you mean waitForRecordsToBeFlushed? the rest of the name is ok? -- 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. To unsubscribe, e-mail: gitbox-unsubscr...@hive.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: gitbox-unsubscr...@hive.apache.org For additional commands, e-mail: gitbox-h...@hive.apache.org
