Mailing-List: contact dev-help@apex.incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@apex.incubator.apache.org
From: tweise <git@git.apache.org>
To: dev@apex.incubator.apache.org
Reply-To: dev@apex.incubator.apache.org
References: <git-pr-145-incubator-apex-malhar@git.apache.org>
In-Reply-To: <git-pr-145-incubator-apex-malhar@git.apache.org>
Subject: [GitHub] incubator-apex-malhar pull request: [For Review Only]
 MLHR-1897 #c...
Content-Type: text/plain
Message-Id: <20151228233429.646FAE0948@git1-us-west.apache.org>
Date: Mon, 28 Dec 2015 23:34:29 +0000 (UTC)

Github user tweise commented on a diff in the pull request:

    https://github.com/apache/incubator-apex-malhar/pull/145#discussion_r48513206
  
    --- Diff: library/src/main/java/com/datatorrent/lib/state/managed/AbstractManagedStateImpl.java ---
    @@ -0,0 +1,487 @@
    +/**
    + * 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 com.datatorrent.lib.state.managed;
    +
    +import java.io.IOException;
    +import java.lang.reflect.Array;
    +import java.util.Comparator;
    +import java.util.Map;
    +import java.util.Timer;
    +import java.util.TimerTask;
    +import java.util.concurrent.Callable;
    +import java.util.concurrent.ExecutorService;
    +import java.util.concurrent.Executors;
    +import java.util.concurrent.Future;
    +import java.util.concurrent.atomic.AtomicReference;
    +
    +import javax.validation.constraints.Min;
    +import javax.validation.constraints.NotNull;
    +
    +import org.joda.time.Duration;
    +import org.slf4j.Logger;
    +import org.slf4j.LoggerFactory;
    +
    +import com.esotericsoftware.kryo.serializers.FieldSerializer;
    +import com.esotericsoftware.kryo.serializers.JavaSerializer;
    +import com.google.common.base.Preconditions;
    +import com.google.common.collect.Maps;
    +import com.google.common.collect.MinMaxPriorityQueue;
    +import com.google.common.util.concurrent.Futures;
    +
    +import com.datatorrent.api.Component;
    +import com.datatorrent.api.Context.DAGContext;
    +import com.datatorrent.api.Context.OperatorContext;
    +import com.datatorrent.api.Operator;
    +import com.datatorrent.api.annotation.Stateless;
    +import com.datatorrent.common.util.NameableThreadFactory;
    +import com.datatorrent.lib.fileaccess.FileAccess;
    +import com.datatorrent.lib.fileaccess.TFileImpl;
    +import com.datatorrent.lib.util.comparator.SliceComparator;
    +import com.datatorrent.netlet.util.DTThrowable;
    +import com.datatorrent.netlet.util.Slice;
    +
    +/**
    + * Not thread safe.
    + */
    +public abstract class AbstractManagedStateImpl
    +    implements ManagedState, Operator, Operator.CheckpointListener, ManagedStateContext
    +{
    +  private long maxCacheSize;
    +
    +  protected int numBuckets;
    +
    +  private int incrementalCheckpointWindowCount = DAGContext.CHECKPOINT_WINDOW_COUNT.defaultValue;
    +
    +  @NotNull
    +  private FileAccess fileAccess = new TFileImpl.DTFileImpl();
    +  @NotNull
    +  protected TimeBucketAssigner timeBucketAssigner = new TimeBucketAssigner();
    +
    +  protected Bucket[] buckets;
    +
    +  private StateTracker tracker;
    +
    +  @Min(1)
    +  private int numReaders = 10;
    +  @NotNull
    +  protected transient ExecutorService readerService;
    +
    +  @NotNull
    +  private final BucketsDataManager dataManager = new BucketsDataManager(this);
    +
    +  private final BucketsMetaDataManager bucketsMetaDataManager = new BucketsMetaDataManager(this);
    +
    +  private transient int operatorId;
    +  private transient long windowId;
    +
    +  private transient int windowCount;
    +
    +  private transient long largestRecoveryWindow;
    +  protected transient boolean replay;
    +
    +  @NotNull
    +  protected Comparator<Slice> keyComparator = new SliceComparator();
    +
    +  protected final transient AtomicReference<Throwable> throwable = new AtomicReference<>();
    +
    +  @NotNull
    +  @FieldSerializer.Bind(JavaSerializer.class)
    +  private Duration checkStateSizeInterval = Duration.millis(
    +      DAGContext.STREAMING_WINDOW_SIZE_MILLIS.defaultValue * OperatorContext.APPLICATION_WINDOW_COUNT.defaultValue);
    +
    +  private transient StateTracker stateTracker;
    +
    +  private final transient Object commitLock = new Object();
    +
    +  @SuppressWarnings("unchecked")
    +  @Override
    +  public void setup(OperatorContext context)
    +  {
    +    operatorId = context.getId();
    +    fileAccess.init();
    +    timeBucketAssigner.register(dataManager);
    +    timeBucketAssigner.setup(context);
    +
    +    numBuckets = getNumBuckets();
    +    buckets = (Bucket[])Array.newInstance(Bucket.class, numBuckets);
    +
    +    Preconditions.checkArgument(numReaders <= numBuckets, "no. of readers cannot exceed no. of buckets");
    +    //setup state data manager
    +    dataManager.setup(context);
    +
    +    //recovering data for window files to bucket
    +    try {
    +      Map<Long, Map<Slice, Bucket.BucketedValue>> recovered = dataManager.load(operatorId);
    +      if (recovered != null && !recovered.isEmpty()) {
    +
    +        for (Map.Entry<Long, Map<Slice, Bucket.BucketedValue>> entry : recovered.entrySet()) {
    +          int bucketIdx = prepareBucket(entry.getKey());
    +
    +          for (Map.Entry<Slice, Bucket.BucketedValue> dataEntry : entry.getValue().entrySet()) {
    +            buckets[bucketIdx].put(dataEntry.getKey(), dataEntry.getValue().getTimeBucket(),
    +                dataEntry.getValue().getValue());
    +          }
    +        }
    +      }
    +    } catch (IOException e) {
    +      throw new RuntimeException("recovering", e);
    +    }
    +
    +    largestRecoveryWindow = dataManager.getLargestRecoveryWindow();
    +    long activationWindow = context.getValue(OperatorContext.ACTIVATION_WINDOW_ID);
    +
    +    if (activationWindow != Stateless.WINDOW_ID && largestRecoveryWindow <= activationWindow) {
    +      replay = true;
    +    }
    +
    +    readerService = Executors.newFixedThreadPool(numReaders, new NameableThreadFactory("managedStateReaders"));
    +
    +    stateTracker = new StateTracker(this, throwable);
    +    stateTracker.setup(context);
    +  }
    +
    +  public abstract int getNumBuckets();
    +
    +  @Override
    +  public void beginWindow(long l)
    +  {
    +    if (throwable.get() != null) {
    +      throw DTThrowable.wrapIfChecked(throwable.get());
    +    }
    +
    +    windowCount++;
    +    timeBucketAssigner.beginWindow(l);
    +    if (replay && l > largestRecoveryWindow) {
    +      replay = false;
    +    }
    +  }
    +
    +  @Override
    +  public void put(long groupId, Slice key, Slice value)
    +  {
    +    if (replay) {
    +      return;
    +    }
    +    int bucketIdx = prepareBucket(groupId);
    +    long timeBucket = timeBucketAssigner.getTimeBucketFor(windowId);
    +    if (timeBucket != -1) {
    +      buckets[bucketIdx].put(key, timeBucket, value);
    +    }
    +  }
    +
    +  @Override
    +  public Slice getSync(long groupId, Slice key)
    +  {
    +    int bucketIdx = prepareBucket(groupId);
    +    return buckets[bucketIdx].get(key, -1, Bucket.ReadSource.ALL);
    +  }
    +
    +  @SuppressWarnings("SynchronizationOnLocalVariableOrMethodParameter")
    +  @Override
    +  public Future<Slice> getAsync(long groupId, Slice key)
    +  {
    +    int bucketIdx = prepareBucket(groupId);
    +    Bucket bucket = buckets[bucketIdx];
    --- End diff --
    
    Why the local variable?


---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastructure@apache.org or file a JIRA ticket
with INFRA.
---