如何在批处理消费者模式下将 Kinesis 标头传递给 Spring Cloud Stream 函数？

Question

I am in the process of converting our existing stream processing code that currently uses the annotation-based programming model to use Spring Cloud Function instead. With consumer batch mode enabled, how can I pass the AWS Kinesis checkpointer to my Consumer so that only after the processor has successfully processed all of the messages in the batch do I perform the manual checkpoint operation? Using the annotation-based programming model, I am easily able to pass in the checkpointer using the @Header annotation. Will this approach still work as I have depicted in the following code snippet?

@Retryable(
      include = {ServerException.class},
      maxAttemptsExpression = "#{@myProperties.getRetry().getMaxAttempts()}",
      backoff = @Backoff(delayExpression = "#{@myProperties.getRetry().getDelay()}"))
  public Consumer<List<byte[]>> processStream(
      @Header(AwsHeaders.CHECKPOINTER) final IRecordProcessorCheckpointer checkpointer) {
    return m -> {
      try {
        myProcessor.process(m);
      } catch (final Exception e) {
        log.error("An error occurred processing the log messages", e);
        throw new ServerException(HttpStatus.INTERNAL_SERVER_ERROR, e);
      }
      checkpoint(checkpointer);
    };
  }

Answer 1

No. The signature of your Consumer must be a Message<List<byte[]>> . so then you can call message.getHeaders().get(AwsHeaders.CHECKPOINTER, IRecordProcessorCheckpointer.class)