Class UsePartitionTimeOnInvalidTimestamp

java.lang.Object
org.apache.kafka.streams.processor.UsePartitionTimeOnInvalidTimestamp
All Implemented Interfaces:
TimestampExtractor

public class UsePartitionTimeOnInvalidTimestamp extends Object
Retrieves embedded metadata timestamps from Kafka messages. If a record has a negative (invalid) timestamp, a new timestamp will be inferred from the current stream-time.

Embedded metadata timestamp was introduced in "KIP-32: Add timestamps to Kafka message" for the new 0.10+ Kafka message format.

Here, "embedded metadata" refers to the fact that compatible Kafka producer clients automatically and transparently embed such timestamps into message metadata they send to Kafka, which can then be retrieved via this timestamp extractor.

If the embedded metadata timestamp represents CreateTime (cf. Kafka broker setting message.timestamp.type and Kafka topic setting log.message.timestamp.type), this extractor effectively provides event-time semantics. If LogAppendTime is used as broker/topic setting to define the embedded metadata timestamps, using this extractor effectively provides ingestion-time semantics.

If you need processing-time semantics, use WallclockTimestampExtractor.

See Also:
  • Constructor Details

    • UsePartitionTimeOnInvalidTimestamp

      public UsePartitionTimeOnInvalidTimestamp()
  • Method Details

    • onInvalidTimestamp

      public long onInvalidTimestamp(ConsumerRecord<Object,Object> record, long recordTimestamp, long partitionTime) throws StreamsException
      Returns the current stream-time as new timestamp for the record.
      Parameters:
      record - a data record
      recordTimestamp - the timestamp extractor from the record
      partitionTime - the highest extracted valid timestamp of the current record's partition˙ (could be -1 if unknown)
      Returns:
      the provided highest extracted valid timestamp as new timestamp for the record
      Throws:
      StreamsException - if highest extracted valid timestamp is unknown
    • extract

      public long extract(ConsumerRecord<Object,Object> record, long partitionTime)
      Extracts the embedded metadata timestamp from the given ConsumerRecord.
      Specified by:
      extract in interface TimestampExtractor
      Parameters:
      record - a data record
      partitionTime - the highest extracted valid timestamp of the current record's partition˙ (could be -1 if unknown)
      Returns:
      the embedded metadata timestamp of the given ConsumerRecord