kfaraz commented on code in PR #14951: URL: https://github.com/apache/druid/pull/14951#discussion_r1319446863
########## server/src/main/java/org/apache/druid/server/coordinator/duty/KillStalePendingSegments.java: ########## @@ -0,0 +1,139 @@ +/* + * 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.druid.server.coordinator.duty; + +import com.google.common.collect.ImmutableList; +import com.google.inject.Inject; +import org.apache.druid.common.guava.FutureUtils; +import org.apache.druid.indexer.TaskStatusPlus; +import org.apache.druid.java.util.common.DateTimes; +import org.apache.druid.java.util.common.logger.Logger; +import org.apache.druid.rpc.indexing.OverlordClient; +import org.apache.druid.server.coordinator.DruidCoordinatorRuntimeParams; +import org.apache.druid.server.coordinator.stats.Dimension; +import org.apache.druid.server.coordinator.stats.RowKey; +import org.apache.druid.server.coordinator.stats.Stats; +import org.joda.time.DateTime; +import org.joda.time.Interval; +import org.joda.time.Period; + +import java.util.HashSet; +import java.util.List; +import java.util.Set; + +/** + * Duty to kill stale pending segments which are not needed anymore. Pending segments + * are created when appending realtime or batch tasks allocate segments to build + * incremental indexes. Under normal operation, these pending segments get committed + * when the task completes and become regular segments. But in case of task failures, + * some pending segments might be left around and cause clutter in the metadata store. + * <p> + * While cleaning up, this duty ensures that the following pending segments are + * retained for at least {@link #DURATION_TO_RETAIN}: + * <ul> + * <li>Pending segments created by any active task (across all datasources)</li> + * <li>Pending segments created by the latest completed task (across all datasources)</li> + * </ul> + */ +public class KillStalePendingSegments implements CoordinatorDuty +{ + private static final Logger log = new Logger(KillStalePendingSegments.class); + private static final Period DURATION_TO_RETAIN = new Period("P1D"); + + private final OverlordClient overlordClient; + + @Inject + public KillStalePendingSegments(OverlordClient overlordClient) + { + this.overlordClient = overlordClient; + } + + @Override + public DruidCoordinatorRuntimeParams run(DruidCoordinatorRuntimeParams params) + { + final Set<String> killDatasources = new HashSet<>( + params.getUsedSegmentsTimelinesPerDataSource().keySet() + ); + killDatasources.removeAll( + params.getCoordinatorDynamicConfig() + .getDataSourcesToNotKillStalePendingSegmentsIn() + ); + + final DateTime minCreatedTime = getMinCreatedTimeToRetain(); + for (String dataSource : killDatasources) { + int pendingSegmentsKilled = FutureUtils.getUnchecked( + overlordClient.killPendingSegments( + dataSource, + new Interval(DateTimes.MIN, minCreatedTime) + ), + true + ); + if (pendingSegmentsKilled > 0) { + log.info( + "Killed [%d] pendingSegments created before [%s] for datasource[%s].", + pendingSegmentsKilled, minCreatedTime, dataSource + ); + params.getCoordinatorStats().add( + Stats.Kill.PENDING_SEGMENTS, + RowKey.of(Dimension.DATASOURCE, dataSource), + pendingSegmentsKilled + ); + } + } + return params; + } + + /** + * Computes the minimum created time of retainable pending segments. Any pending + * segment created before this time is considered stale and can be safely deleted. + * The limit is determined to ensure that pending segments created by any active + * task and the latest completed task (across all datasources) are retained for + * at least {@link #DURATION_TO_RETAIN}. + */ + private DateTime getMinCreatedTimeToRetain() + { + // Fetch the statuses of all active tasks and the latest completed task + // (The Overlord API returns complete tasks in descending order of created_date.) + final List<TaskStatusPlus> statuses = ImmutableList.copyOf( + FutureUtils.getUnchecked(overlordClient.taskStatuses(null, null, 1), true) + ); + + DateTime earliestActiveTaskStart = DateTimes.nowUtc(); + DateTime latestCompletedTaskStart = null; + for (TaskStatusPlus status : statuses) { + if (status.getStatusCode() == null) { + // Unknown status Review Comment: Fixed in ~#14891~ #14961 -- 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: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
