jihoonson commented on a change in pull request #6901: Introduce published segment cache in broker URL: https://github.com/apache/incubator-druid/pull/6901#discussion_r253155145
########## File path: sql/src/main/java/org/apache/druid/sql/calcite/schema/MetadataSegmentView.java ########## @@ -0,0 +1,221 @@ +/* + * 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.sql.calcite.schema; + +import com.fasterxml.jackson.core.type.TypeReference; +import com.fasterxml.jackson.databind.JavaType; +import com.fasterxml.jackson.databind.ObjectMapper; +import com.google.common.base.Preconditions; +import com.google.common.util.concurrent.ListenableFuture; +import com.google.inject.Inject; +import org.apache.druid.client.BrokerSegmentWatcherConfig; +import org.apache.druid.client.DataSegmentInterner; +import org.apache.druid.client.JsonParserIterator; +import org.apache.druid.client.coordinator.Coordinator; +import org.apache.druid.discovery.DruidLeaderClient; +import org.apache.druid.guice.ManageLifecycle; +import org.apache.druid.java.util.common.DateTimes; +import org.apache.druid.java.util.common.StringUtils; +import org.apache.druid.java.util.common.concurrent.Execs; +import org.apache.druid.java.util.common.lifecycle.LifecycleStart; +import org.apache.druid.java.util.common.lifecycle.LifecycleStop; +import org.apache.druid.java.util.emitter.EmittingLogger; +import org.apache.druid.java.util.http.client.Request; +import org.apache.druid.server.coordinator.BytesAccumulatingResponseHandler; +import org.apache.druid.sql.calcite.planner.PlannerConfig; +import org.apache.druid.timeline.DataSegment; +import org.jboss.netty.handler.codec.http.HttpMethod; +import org.joda.time.DateTime; + +import java.io.IOException; +import java.io.InputStream; +import java.util.Iterator; +import java.util.Set; +import java.util.concurrent.ConcurrentHashMap; +import java.util.concurrent.ConcurrentMap; +import java.util.concurrent.ScheduledExecutorService; +import java.util.concurrent.TimeUnit; + +/** + * This class polls the coordinator in background to keep the latest published segments. + * Provides {@link #getPublishedSegments()} for others to get segments in metadata store. + */ +@ManageLifecycle +public class MetadataSegmentView +{ + + private static final long POLL_PERIOD_IN_MS = 60000; + private static final EmittingLogger log = new EmittingLogger(MetadataSegmentView.class); + + private final DruidLeaderClient coordinatorDruidLeaderClient; + private final ObjectMapper jsonMapper; + private final BytesAccumulatingResponseHandler responseHandler; + private final BrokerSegmentWatcherConfig segmentWatcherConfig; + + private final ConcurrentMap<DataSegment, DateTime> publishedSegments = new ConcurrentHashMap<>(1000); + private ScheduledExecutorService scheduledExec; + final PlannerConfig plannerConfig; + + @Inject + public MetadataSegmentView( + final @Coordinator DruidLeaderClient druidLeaderClient, + final ObjectMapper jsonMapper, + final BytesAccumulatingResponseHandler responseHandler, + final BrokerSegmentWatcherConfig segmentWatcherConfig, + final PlannerConfig plannerConfig + ) + { + this.plannerConfig = Preconditions.checkNotNull(plannerConfig, "plannerConfig"); + this.coordinatorDruidLeaderClient = druidLeaderClient; + this.jsonMapper = jsonMapper; + this.responseHandler = responseHandler; + this.segmentWatcherConfig = segmentWatcherConfig; + } + + @LifecycleStart + public void start() + { + if (plannerConfig.isMetadataSegmentCacheEnable()) { + scheduledExec = Execs.scheduledSingleThreaded("MetadataSegmentView-Cache--%d"); + scheduledExec.schedule(new PollTask(), 0, TimeUnit.MILLISECONDS); + } + } + + @LifecycleStop + public void stop() + { + scheduledExec.shutdownNow(); Review comment: Ah, I think those locks are required for some modules which can be started and stopped repeatedly, like whenever getting/losing the zk leadership. For them, start() and stop() can be called simultaneously at any time, so they need a lock for coordination. However, `MetadataSegmentView` has the same lifecycle with the broker. Once the broker is stopped, `MetadataSegmentView.start()` would never get called again. So, I think the only case we should consider is when stop() is called while starting. `DruidSchema` is also similar to this and doesn't have a lock. What do you think? ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@druid.apache.org For additional commands, e-mail: commits-h...@druid.apache.org
