org.apache.flink.runtime.dispatcher.cleanup.DispatcherResourceCleanerFactory Maven / Gradle / Ivy
/*
* 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.flink.runtime.dispatcher.cleanup;
import org.apache.flink.annotation.VisibleForTesting;
import org.apache.flink.runtime.blob.BlobServer;
import org.apache.flink.runtime.concurrent.ComponentMainThreadExecutor;
import org.apache.flink.runtime.dispatcher.DispatcherServices;
import org.apache.flink.runtime.dispatcher.JobManagerRunnerRegistry;
import org.apache.flink.runtime.highavailability.HighAvailabilityServices;
import org.apache.flink.runtime.jobmanager.JobGraphWriter;
import org.apache.flink.runtime.metrics.groups.JobManagerMetricGroup;
import org.apache.flink.util.Preconditions;
import org.apache.flink.util.concurrent.RetryStrategy;
import java.util.concurrent.Executor;
/**
* {@code DispatcherResourceCleanerFactory} instantiates {@link ResourceCleaner} instances that
* clean cleanable resources from the {@link org.apache.flink.runtime.dispatcher.Dispatcher}.
*
* We need to handle the {@link JobManagerRunnerRegistry} differently due to a dependency between
* closing the {@link org.apache.flink.runtime.jobmaster.JobManagerRunner} and the {@link
* HighAvailabilityServices}. This is fixed in {@code FLINK-24038} using a feature flag to
* enable/disable single leader election for all the {@code JobManager} components. We can remove
* the priority cleanup logic after removing the per-component leader election.
*/
public class DispatcherResourceCleanerFactory implements ResourceCleanerFactory {
private static final String JOB_MANAGER_RUNNER_REGISTRY_LABEL = "JobManagerRunnerRegistry";
private static final String JOB_GRAPH_STORE_LABEL = "JobGraphStore";
private static final String BLOB_SERVER_LABEL = "BlobServer";
private static final String HA_SERVICES_LABEL = "HighAvailabilityServices";
private static final String JOB_MANAGER_METRIC_GROUP_LABEL = "JobManagerMetricGroup";
private final Executor cleanupExecutor;
private final RetryStrategy retryStrategy;
private final JobManagerRunnerRegistry jobManagerRunnerRegistry;
private final JobGraphWriter jobGraphWriter;
private final BlobServer blobServer;
private final HighAvailabilityServices highAvailabilityServices;
private final JobManagerMetricGroup jobManagerMetricGroup;
public DispatcherResourceCleanerFactory(
JobManagerRunnerRegistry jobManagerRunnerRegistry,
DispatcherServices dispatcherServices) {
this(
dispatcherServices.getIoExecutor(),
CleanupRetryStrategyFactory.INSTANCE.createRetryStrategy(
dispatcherServices.getConfiguration()),
jobManagerRunnerRegistry,
dispatcherServices.getJobGraphWriter(),
dispatcherServices.getBlobServer(),
dispatcherServices.getHighAvailabilityServices(),
dispatcherServices.getJobManagerMetricGroup());
}
@VisibleForTesting
public DispatcherResourceCleanerFactory(
Executor cleanupExecutor,
RetryStrategy retryStrategy,
JobManagerRunnerRegistry jobManagerRunnerRegistry,
JobGraphWriter jobGraphWriter,
BlobServer blobServer,
HighAvailabilityServices highAvailabilityServices,
JobManagerMetricGroup jobManagerMetricGroup) {
this.cleanupExecutor = Preconditions.checkNotNull(cleanupExecutor);
this.retryStrategy = retryStrategy;
this.jobManagerRunnerRegistry = Preconditions.checkNotNull(jobManagerRunnerRegistry);
this.jobGraphWriter = Preconditions.checkNotNull(jobGraphWriter);
this.blobServer = Preconditions.checkNotNull(blobServer);
this.highAvailabilityServices = Preconditions.checkNotNull(highAvailabilityServices);
this.jobManagerMetricGroup = Preconditions.checkNotNull(jobManagerMetricGroup);
}
@Override
public ResourceCleaner createLocalResourceCleaner(
ComponentMainThreadExecutor mainThreadExecutor) {
return DefaultResourceCleaner.forLocallyCleanableResources(
mainThreadExecutor, cleanupExecutor, retryStrategy)
.withPrioritizedCleanup(JOB_MANAGER_RUNNER_REGISTRY_LABEL, jobManagerRunnerRegistry)
.withRegularCleanup(JOB_GRAPH_STORE_LABEL, jobGraphWriter)
.withRegularCleanup(BLOB_SERVER_LABEL, blobServer)
.withRegularCleanup(JOB_MANAGER_METRIC_GROUP_LABEL, jobManagerMetricGroup)
.build();
}
@Override
public ResourceCleaner createGlobalResourceCleaner(
ComponentMainThreadExecutor mainThreadExecutor) {
return DefaultResourceCleaner.forGloballyCleanableResources(
mainThreadExecutor, cleanupExecutor, retryStrategy)
.withPrioritizedCleanup(
JOB_MANAGER_RUNNER_REGISTRY_LABEL,
ofLocalResource(jobManagerRunnerRegistry))
.withRegularCleanup(JOB_GRAPH_STORE_LABEL, jobGraphWriter)
.withRegularCleanup(BLOB_SERVER_LABEL, blobServer)
.withRegularCleanup(HA_SERVICES_LABEL, highAvailabilityServices)
.withRegularCleanup(
JOB_MANAGER_METRIC_GROUP_LABEL, ofLocalResource(jobManagerMetricGroup))
.build();
}
/**
* A simple wrapper for the resources that don't have any artifacts that can outlive the {@link
* org.apache.flink.runtime.dispatcher.Dispatcher}, but we still want to clean up their local
* state when we terminate globally.
*
* @param localResource Local resource that we want to clean during a global cleanup.
* @return Globally cleanable resource.
*/
private static GloballyCleanableResource ofLocalResource(
LocallyCleanableResource localResource) {
return localResource::localCleanupAsync;
}
}