apex-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From davidyan74 <...@git.apache.org>
Subject [GitHub] apex-malhar pull request #345: REVIEW ONLY (WindowedOperator): splitting Win...
Date Mon, 18 Jul 2016 20:46:40 GMT
Github user davidyan74 commented on a diff in the pull request:

    https://github.com/apache/apex-malhar/pull/345#discussion_r71225979
  
    --- Diff: library/src/main/java/org/apache/apex/malhar/lib/window/impl/SpillableWindowedKeyedStorage.java
---
    @@ -0,0 +1,136 @@
    +package org.apache.apex.malhar.lib.window.impl;
    +
    +import java.util.List;
    +import java.util.Map;
    +
    +import javax.annotation.concurrent.Immutable;
    +import javax.validation.constraints.NotNull;
    +
    +import org.apache.apex.malhar.lib.state.spillable.SpillableByteArrayListMultimapImpl;
    +import org.apache.apex.malhar.lib.state.spillable.SpillableByteMapImpl;
    +import org.apache.apex.malhar.lib.state.spillable.managed.ManagedStateSpillableStateStore;
    +import org.apache.apex.malhar.lib.utils.serde.Serde;
    +import org.apache.apex.malhar.lib.window.Window;
    +import org.apache.apex.malhar.lib.window.WindowedStorage;
    +import org.apache.commons.lang3.tuple.ImmutablePair;
    +
    +import com.datatorrent.netlet.util.Slice;
    +
    +/**
    + * Created by david on 7/15/16.
    + */
    +public class SpillableWindowedKeyedStorage<K, V> implements WindowedStorage.WindowedKeyedStorage<K,
V>
    +{
    +  @NotNull
    +  private final ManagedStateSpillableStateStore store;
    +
    +  protected final SpillableByteMapImpl<ImmutablePair<Window, K>, V> internValues;
    +  protected final SpillableByteArrayListMultimapImpl<Window, K> internKeys;
    +
    +  private SpillableWindowedKeyedStorage()
    +  {
    +    // for kryo
    +    store = null;
    +    internValues = null;
    +    internKeys = null;
    +  }
    +
    +  public SpillableWindowedKeyedStorage(long bucket, byte[] identifier, Serde<Window,
Slice> serdeWindow, Serde<K, Slice> serdeKey, Serde<ImmutablePair<Window, K>,
Slice> serdeWindowKey, Serde<V, Slice> serdeValue)
    +  {
    +    store = new ManagedStateSpillableStateStore();
    +    store.getCheckpointManager().setNeedBucketFile(false);
    +    internValues = new SpillableByteMapImpl<>(store, identifier, bucket, serdeWindowKey,
serdeValue);
    +    internKeys = new SpillableByteArrayListMultimapImpl<>(store, identifier, bucket,
serdeWindow, serdeKey);
    +  }
    +
    +  @Override
    +  public boolean containsWindow(Window window)
    +  {
    +    return internKeys.containsKey(window);
    +  }
    +
    +  @Override
    +  public long size()
    +  {
    +    return internKeys.size();
    +  }
    +
    +  @Override
    +  public void remove(Window window)
    +  {
    +    List<K> keys = internKeys.get(window);
    +    for (K key : keys) {
    +      internValues.remove(new ImmutablePair<>(window, key));
    +    }
    +    internKeys.removeAll(window);
    +  }
    +
    +  @Override
    +  public void migrateWindow(Window fromWindow, Window toWindow)
    +  {
    +    List<K> keys = internKeys.get(fromWindow);
    +    internValues.remove(toWindow);
    +    for (K key : keys) {
    +      internKeys.put(toWindow, key);
    +      ImmutablePair<Window, K> oldKey = new ImmutablePair<>(fromWindow, key);
    +      ImmutablePair<Window, K> newKey = new ImmutablePair<>(toWindow, key);
    +
    +      V value = internValues.get(oldKey);
    +      internValues.remove(oldKey);
    +      internValues.put(newKey, value);
    +    }
    +    internKeys.removeAll(fromWindow);
    +  }
    +
    --- End diff --
    
    Will do. Can you explain a little bit what bucket and identifier actually do here? Is
it something that is used to name the directories in HDFS eventually?


---
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.
---

Mime
View raw message