hlteoh37 opened a new pull request, #21003: URL: https://github.com/apache/flink/pull/21003
…FanOutRecordPublisher Cherry-pick of https://github.com/apache/flink/pull/20945/files ## What is the purpose of the change The consumer fails when three conditions are met: 1. Kinesis EFO record publisher uses an AT_TIMESTAMP starting position. 2. The first record batch is not empty. 3. The deaggregated record batch is empty. This can happen when the user writes aggregated records into their Kinesis stream whilst specifying the explicitHashKey for every single user record. When resharding occurs, there is a possibility that records within the same aggregated batch fall into two different shards, since they have different explicit hash keys. When this happens, the deaggregated records are dropped. (We need to do this in Kinesis Consumer so we do not process duplicate records. See https://github.com/awslabs/kinesis-aggregation/issues/11) If all deaggregated records are dropped, the consumer receives an empty deaggregated record batch even though the record batch is not empty. ## Brief change log This is fixed by returning the AT_TIMESTAMP starting position when receiving an AT_TIMESTAMP sentinel sequence number. ## Verifying this change - Unit tests ## Does this pull request potentially affect one of the following parts: - Dependencies (does it add or upgrade a dependency): no - The public API, i.e., is any changed class annotated with `@Public(Evolving)`: no - The serializers: no - The runtime per-record code paths (performance sensitive): no - Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: no - The S3 file system connector: no ## Documentation - Does this pull request introduce a new feature? no - If yes, how is the feature documented? not applicable -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
