Kafka source offset-based deduplication.

Author: tomstepp

runners/google-cloud-dataflow-java/worker/src/main/java/org/apache/beam/runners/dataflow/worker/OrderedCode.java

@@ -17,6 +17,8 @@
  */
 package org.apache.beam.sdk.io.kafka;
 
+import static org.apache.beam.vendor.guava.v32_1_2_jre.com.google.common.base.Preconditions.checkState;
+
 import java.io.Serializable;
 import java.util.List;
 import java.util.Optional;
@@ -42,6 +44,8 @@ public class KafkaCheckpointMark implements UnboundedSource.CheckpointMark {
   @SuppressWarnings("initialization") // Avro will set the fields by breaking abstraction
   private KafkaCheckpointMark() {} // for Avro
 
+  private static final long OFFSET_DEDUP_PARTITIONS_PER_SPLIT = 1;
+
   public KafkaCheckpointMark(
       List<PartitionMark> partitions, Optional<KafkaUnboundedReader<?, ?>> reader) {
     this.partitions = partitions;
@@ -66,6 +70,23 @@ public String toString() {
     return "KafkaCheckpointMark{partitions=" + Joiner.on(",").join(partitions) + '}';
   }
 
+  @Override
+  public byte[] getOffsetLimit() {
+    if (!reader.isPresent()) {
+      throw new RuntimeException(
+          "KafkaCheckpointMark reader is not present while calling getOffsetLimit().");
+    }
+    if (!reader.get().offsetBasedDeduplicationSupported()) {
+      throw new RuntimeException(
+          "Unexpected getOffsetLimit() called while KafkaUnboundedReader not configured for offset deduplication.");
+    }
+
+    // KafkaUnboundedSource.split() must produce a 1:1 partition to split ratio.
+    checkState(partitions.size() == OFFSET_DEDUP_PARTITIONS_PER_SPLIT);
+    PartitionMark partition = partitions.get(/* index= */ 0);
+    return KafkaIOUtils.OffsetBasedDeduplication.encodeOffset(partition.getNextOffset());
+  }
+
   /**
    * A tuple to hold topic, partition, and offset that comprise the checkpoint for a single
    * partition.

runners/google-cloud-dataflow-java/worker/src/test/java/org/apache/beam/runners/dataflow/worker/OrderedCodeTest.java

@@ -610,6 +610,7 @@ public static <K, V> Read<K, V> read() {
         .setRedistributed(false)
         .setAllowDuplicates(false)
         .setRedistributeNumKeys(0)
+        .setOffsetDeduplication(false)
         .build();
   }
 
@@ -717,6 +718,9 @@ public abstract static class Read<K, V>
     @Pure
     public abstract int getRedistributeNumKeys();
 
+    @Pure
+    public abstract boolean isOffsetDeduplication();
+
     @Pure
     public abstract @Nullable Duration getWatchTopicPartitionDuration();
 
@@ -782,6 +786,8 @@ abstract Builder<K, V> setConsumerFactoryFn(
 
       abstract Builder<K, V> setRedistributeNumKeys(int redistributeNumKeys);
 
+      abstract Builder<K, V> setOffsetDeduplication(boolean offsetDeduplication);
+
       abstract Builder<K, V> setTimestampPolicyFactory(
           TimestampPolicyFactory<K, V> timestampPolicyFactory);
 
@@ -892,6 +898,10 @@ static <K, V> void setupExternalBuilder(
           builder.setRedistributeNumKeys(0);
           builder.setAllowDuplicates(false);
         }
+        // TODO(tomstepp): Auto-enable offset deduplication if: redistributed and !allowDuplicates.
+        if (config.offsetDeduplication != null) {
+          builder.setOffsetDeduplication(config.offsetDeduplication);
+        }
       }
 
       private static <T> Coder<T> resolveCoder(Class<Deserializer<T>> deserializer) {
@@ -959,6 +969,7 @@ public static class Configuration {
         private Integer redistributeNumKeys;
         private Boolean redistribute;
         private Boolean allowDuplicates;
+        private Boolean offsetDeduplication;
 
         public void setConsumerConfig(Map<String, String> consumerConfig) {
           this.consumerConfig = consumerConfig;
@@ -1015,6 +1026,10 @@ public void setRedistribute(Boolean redistribute) {
         public void setAllowDuplicates(Boolean allowDuplicates) {
           this.allowDuplicates = allowDuplicates;
         }
+
+        public void setOffsetDeduplication(Boolean offsetDeduplication) {
+          this.offsetDeduplication = offsetDeduplication;
+        }
       }
     }
 
@@ -1086,6 +1101,10 @@ public Read<K, V> withRedistributeNumKeys(int redistributeNumKeys) {
       return toBuilder().setRedistributeNumKeys(redistributeNumKeys).build();
     }
 
+    public Read<K, V> withOffsetDeduplication(boolean offsetDeduplication) {
+      return toBuilder().setOffsetDeduplication(offsetDeduplication).build();
+    }
+
     /**
      * Internally sets a {@link java.util.regex.Pattern} of topics to read from. All the partitions
      * from each of the matching topics are read.

sdks/java/io/kafka/src/main/java/org/apache/beam/sdk/io/kafka/KafkaCheckpointMark.java

@@ -137,7 +137,12 @@ Object getDefaultValue() {
         return false;
       }
     },
-    ;
+    OFFSET_DEDUPLICATION(LEGACY) {
+      @Override
+      Object getDefaultValue() {
+        return false;
+      }
+    };
 
     private final @NonNull ImmutableSet<KafkaIOReadImplementation> supportedImplementations;
 

sdks/java/io/kafka/src/main/java/org/apache/beam/sdk/io/kafka/KafkaIO.java

@@ -19,6 +19,8 @@
 
 import static org.apache.beam.vendor.guava.v32_1_2_jre.com.google.common.base.Preconditions.checkArgument;
 
+import java.nio.ByteBuffer;
+import java.nio.charset.StandardCharsets;
 import java.util.HashMap;
 import java.util.Map;
 import java.util.Random;
@@ -142,4 +144,17 @@ void update(double quantity) {
       return avg;
     }
   }
+
+  static final class OffsetBasedDeduplication {
+    private static final ByteBuffer offsetBuffer = ByteBuffer.allocate(Long.BYTES);
+
+    static byte[] encodeOffset(long offset) {
+      offsetBuffer.putLong(/* index= */ 0, offset);
+      return offsetBuffer.array();
+    }
+
+    static byte[] getUniqueId(String topic, int partition, long offset) {
+      return (topic + "-" + partition + "-" + offset).getBytes(StandardCharsets.UTF_8);
+    }
+  }
 }

sdks/java/io/kafka/src/main/java/org/apache/beam/sdk/io/kafka/KafkaIOReadImplementationCompatibility.java

@@ -214,6 +214,12 @@ public boolean advance() throws IOException {
         curTimestamp =
             pState.timestampPolicy.getTimestampForRecord(pState.mkTimestampPolicyContext(), record);
         curRecord = record;
+        if (this.offsetBasedDeduplicationSupported) {
+          curOffset = KafkaIOUtils.OffsetBasedDeduplication.encodeOffset(offset);
+          curId =
+              KafkaIOUtils.OffsetBasedDeduplication.getUniqueId(
+                  rawRecord.topic(), rawRecord.partition(), rawRecord.offset());
+        }
 
         int recordSize =
             (rawRecord.key() == null ? 0 : rawRecord.key().length)
@@ -299,6 +305,28 @@ public Instant getCurrentTimestamp() throws NoSuchElementException {
     return curTimestamp;
   }
 
+  @Override
+  public byte[] getCurrentRecordId() throws NoSuchElementException {
+    if (!this.offsetBasedDeduplicationSupported) {
+      throw new RuntimeException("UnboundedSource must enable offset-based deduplication.");
+    }
+    if (curId == null) {
+      throw new NoSuchElementException("KafkaUnboundedReader's curId is null.");
+    }
+    return curId;
+  }
+
+  @Override
+  public byte[] getCurrentRecordOffset() throws NoSuchElementException {
+    if (!this.offsetBasedDeduplicationSupported) {
+      throw new RuntimeException("UnboundedSource must enable offset-based deduplication.");
+    }
+    if (curOffset == null) {
+      throw new NoSuchElementException("KafkaUnboundedReader's curOffset is null.");
+    }
+    return curOffset;
+  }
+
   @Override
   public long getSplitBacklogBytes() {
     long backlogBytes = 0;
@@ -314,6 +342,10 @@ public long getSplitBacklogBytes() {
     return backlogBytes;
   }
 
+  public boolean offsetBasedDeduplicationSupported() {
+    return this.offsetBasedDeduplicationSupported;
+  }
+
   ////////////////////////////////////////////////////////////////////////////////////////////////
 
   private static final Logger LOG = LoggerFactory.getLogger(KafkaUnboundedReader.class);
@@ -336,6 +368,10 @@ public long getSplitBacklogBytes() {
   private @Nullable Instant curTimestamp = null;
   private Iterator<PartitionState<K, V>> curBatch = Collections.emptyIterator();
 
+  private final boolean offsetBasedDeduplicationSupported;
+  private byte[] curOffset = new byte[0];
+  private byte[] curId = new byte[0];
+
   private @Nullable Deserializer<K> keyDeserializerInstance = null;
   private @Nullable Deserializer<V> valueDeserializerInstance = null;
 
@@ -507,6 +543,7 @@ Instant updateAndGetWatermark() {
       KafkaUnboundedSource<K, V> source, @Nullable KafkaCheckpointMark checkpointMark) {
     this.source = source;
     this.name = "Reader-" + source.getId();
+    this.offsetBasedDeduplicationSupported = source.offsetBasedDeduplicationSupported();
 
     List<TopicPartition> partitions =
         Preconditions.checkArgumentNotNull(source.getSpec().getTopicPartitions());

sdks/java/io/kafka/src/main/java/org/apache/beam/sdk/io/kafka/KafkaIOUtils.java

@@ -113,10 +113,20 @@ public List<KafkaUnboundedSource<K, V>> split(int desiredNumSplits, PipelineOpti
         partitions.size() > 0,
         "Could not find any partitions. Please check Kafka configuration and topic names");
 
-    int numSplits = Math.min(desiredNumSplits, partitions.size());
-    // XXX make all splits have the same # of partitions
-    while (partitions.size() % numSplits > 0) {
-      ++numSplits;
+    int numSplits;
+    if (offsetBasedDeduplicationSupported()) {
+      // Enforce 1:1 split to partition ratio for offset deduplication.
+      numSplits = partitions.size();
+      LOG.info(
+          "Offset-based deduplication is enabled for KafkaUnboundedSource. "
+              + "Forcing the number of splits to equal the number of total partitions: {}.",
+          numSplits);
+    } else {
+      numSplits = Math.min(desiredNumSplits, partitions.size());
+      // Make all splits have the same # of partitions.
+      while (partitions.size() % numSplits > 0) {
+        ++numSplits;
+      }
     }
     List<List<TopicPartition>> assignments = new ArrayList<>(numSplits);
 
@@ -177,6 +187,11 @@ public boolean requiresDeduping() {
     return false;
   }
 
+  @Override
+  public boolean offsetBasedDeduplicationSupported() {
+    return spec.isOffsetDeduplication();
+  }
+
   @Override
   public Coder<KafkaRecord<K, V>> getOutputCoder() {
     Coder<K> keyCoder = Preconditions.checkStateNotNull(spec.getKeyCoder());

sdks/java/io/kafka/src/main/java/org/apache/beam/sdk/io/kafka/KafkaUnboundedReader.java

@@ -111,7 +111,8 @@ public void testConstructKafkaRead() throws Exception {
                         Field.of("consumer_polling_timeout", FieldType.INT64),
                         Field.of("redistribute_num_keys", FieldType.INT32),
                         Field.of("redistribute", FieldType.BOOLEAN),
-                        Field.of("allow_duplicates", FieldType.BOOLEAN)))
+                        Field.of("allow_duplicates", FieldType.BOOLEAN),
+                        Field.of("offset_deduplication", FieldType.BOOLEAN)))
                 .withFieldValue("topics", topics)
                 .withFieldValue("consumer_config", consumerConfig)
                 .withFieldValue("key_deserializer", keyDeserializer)
@@ -123,6 +124,7 @@ public void testConstructKafkaRead() throws Exception {
                 .withFieldValue("redistribute_num_keys", 0)
                 .withFieldValue("redistribute", false)
                 .withFieldValue("allow_duplicates", false)
+                .withFieldValue("offset_deduplication", false)
                 .build());
 
     RunnerApi.Components defaultInstance = RunnerApi.Components.getDefaultInstance();
@@ -247,7 +249,8 @@ public void testConstructKafkaReadWithoutMetadata() throws Exception {
                         Field.of("timestamp_policy", FieldType.STRING),
                         Field.of("redistribute_num_keys", FieldType.INT32),
                         Field.of("redistribute", FieldType.BOOLEAN),
-                        Field.of("allow_duplicates", FieldType.BOOLEAN)))
+                        Field.of("allow_duplicates", FieldType.BOOLEAN),
+                        Field.of("offset_deduplication", FieldType.BOOLEAN)))
                 .withFieldValue("topics", topics)
                 .withFieldValue("consumer_config", consumerConfig)
                 .withFieldValue("key_deserializer", keyDeserializer)
@@ -258,6 +261,7 @@ public void testConstructKafkaReadWithoutMetadata() throws Exception {
                 .withFieldValue("redistribute_num_keys", 0)
                 .withFieldValue("redistribute", false)
                 .withFieldValue("allow_duplicates", false)
+                .withFieldValue("offset_deduplication", false)
                 .build());
 
     RunnerApi.Components defaultInstance = RunnerApi.Components.getDefaultInstance();

sdks/java/io/kafka/src/main/java/org/apache/beam/sdk/io/kafka/KafkaUnboundedSource.java

@@ -101,6 +101,7 @@ static class KafkaIOReadWithMetadataTranslator implements TransformPayloadTransl
             .addBooleanField("redistribute")
             .addBooleanField("allows_duplicates")
             .addNullableInt32Field("redistribute_num_keys")
+            .addBooleanField("offset_deduplication")
             .addNullableLogicalTypeField("watch_topic_partition_duration", new NanosDuration())
             .addByteArrayField("timestamp_policy_factory")
             .addNullableMapField("offset_consumer_config", FieldType.STRING, FieldType.BYTES)
@@ -221,6 +222,7 @@ public Row toConfigRow(Read<?, ?> transform) {
       fieldValues.put("redistribute", transform.isRedistributed());
       fieldValues.put("redistribute_num_keys", transform.getRedistributeNumKeys());
       fieldValues.put("allows_duplicates", transform.isAllowDuplicates());
+      fieldValues.put("offset_deduplication", transform.isOffsetDeduplication());
       return Row.withSchema(schema).withFieldValues(fieldValues).build();
     }
 
@@ -349,6 +351,10 @@ public Row toConfigRow(Read<?, ?> transform) {
             }
           }
         }
+        Boolean offsetDeduplication = configRow.getValue("offset_deduplication");
+        if (offsetDeduplication != null) {
+          transform = transform.withOffsetDeduplication(offsetDeduplication);
+        }
         Duration maxReadTime = configRow.getValue("max_read_time");
         if (maxReadTime != null) {
           transform =