Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
/*
* 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.executiongraph.failover;
import org.apache.flink.annotation.VisibleForTesting;
import org.apache.flink.runtime.executiongraph.ExecutionGraph;
import org.apache.flink.runtime.executiongraph.ExecutionVertex;
import org.apache.flink.runtime.executiongraph.IntermediateResultPartition;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.ArrayDeque;
import java.util.IdentityHashMap;
import java.util.Optional;
import java.util.Queue;
import java.util.Set;
import java.util.concurrent.Executor;
/**
* A strict version of {@link RestartPipelinedRegionStrategy} which ensures data consistency
* even if there are non-deterministic shuffles(rebalance, rescale..).
*/
public class StrictRestartPipelinedRegionStrategy extends RestartPipelinedRegionStrategy {
/** The log object used for debugging. */
private static final Logger LOG = LoggerFactory.getLogger(StrictRestartPipelinedRegionStrategy.class);
/**
* Creates a new failover strategy to restart pipelined regions that works on the given
* execution graph and uses the execution graph's future executor to call restart actions.
*
* @param executionGraph The execution graph on which this FailoverStrategy will work
* @param regionFailLimit The max number a region can fail
*/
public StrictRestartPipelinedRegionStrategy(ExecutionGraph executionGraph, int regionFailLimit) {
this(executionGraph, executionGraph.getFutureExecutor(), regionFailLimit);
}
/**
* Creates a new failover strategy to restart pipelined regions that works on the given
* execution graph and uses the given executor to call restart actions.
*
* @param executionGraph The execution graph on which this FailoverStrategy will work
* @param executor The executor used for future actions
* @param regionFailLimit The max number a region can fail
*/
public StrictRestartPipelinedRegionStrategy(ExecutionGraph executionGraph, Executor executor, int regionFailLimit) {
this(executionGraph, executor, executionGraph.getResultPartitionFailureChecker(), regionFailLimit);
}
@VisibleForTesting
public StrictRestartPipelinedRegionStrategy(
ExecutionGraph executionGraph,
Executor executor,
ResultPartitionFailureChecker resultPartitionFailureChecker,
int regionFailLimit) {
super(executionGraph, executor, resultPartitionFailureChecker, regionFailLimit);
}
// ------------------------------------------------------------------------
// failover implementation
// ------------------------------------------------------------------------
/**
* All 'involved' regions are proposed to be restarted.
* The 'involved' regions are calculated with rules below:
* 1. The region containing the failed task is always involved
* 2. If an input result partition of an involved region is not available, i.e. Missing or Corrupted,
* the region containing the partition producer task is involved
* 3. If a region is involved, all of its consumer regions are involved
*/
@Override
@VisibleForTesting
public Set getRegionsToRestart(ExecutionVertex failedVertex, Throwable cause) {
IdentityHashMap regionsToRestart = new IdentityHashMap<>();
IdentityHashMap visitedRegions = new IdentityHashMap<>();
Queue regionsToVisit = new ArrayDeque<>();
FailoverRegion failedRegion = getFailoverRegion(failedVertex);
regionsToVisit.add(failedRegion);
visitedRegions.put(failedRegion, null);
// get root failed region.
Optional rootFailedRegion = getRootFailedRegion(failedVertex, cause);
if (rootFailedRegion.isPresent() && rootFailedRegion.get() != failedRegion) {
regionsToVisit.add(rootFailedRegion.get());
visitedRegions.put(rootFailedRegion.get(), null);
}
// start from the failed region to visit all involved regions
while (!regionsToVisit.isEmpty()) {
FailoverRegion regionToRestart = regionsToVisit.poll();
// an involved region should be restarted
regionsToRestart.put(regionToRestart, null);
// if a needed input result partition is not available, its producer region is involved
for (ExecutionVertex vertex : regionToRestart.getAllExecutionVertices()) {
for (int i = 0; i < vertex.getNumberOfInputs(); ++i) {
for (IntermediateResultPartition consumedPartition : vertex.getConsumedPartitions(i)) {
if (resultPartitionFailureChecker.isFailed(consumedPartition.getPartitionId())) {
FailoverRegion producerRegion = getFailoverRegion(consumedPartition.getProducer());
if (!visitedRegions.containsKey(producerRegion)) {
visitedRegions.put(producerRegion, null);
regionsToVisit.add(producerRegion);
}
}
}
}
}
// all consumer regions of an involved region should be involved
for (ExecutionVertex vertex : regionToRestart.getAllExecutionVertices()) {
for (IntermediateResultPartition resultPartition : vertex.getProducedPartitions().values()) {
for (ExecutionVertex consumerVertex : resultPartition.getConsumers()) {
FailoverRegion consumerRegion = getFailoverRegion(consumerVertex);
if (!visitedRegions.containsKey(consumerRegion)) {
visitedRegions.put(consumerRegion, null);
regionsToVisit.add(consumerRegion);
}
}
}
}
}
return regionsToRestart.keySet();
}
@Override
public String getStrategyName() {
return "Strict Pipelined Region Failover";
}
// ------------------------------------------------------------------------
// factory
// ------------------------------------------------------------------------
/**
* Factory that instantiates the StrictRestartPipelinedRegionStrategy.
*/
public static class Factory implements FailoverStrategy.Factory {
private int regionFailLimit = 100;
@Override
public FailoverStrategy create(ExecutionGraph executionGraph) {
return new StrictRestartPipelinedRegionStrategy(executionGraph, regionFailLimit);
}
public void setRegionFailLimit(int regionFailLimit) {
this.regionFailLimit = regionFailLimit;
}
}
}
