@guardian/cdk
    Preparing search index...

    Interface StreamErrorHandlingProps

    In order to prevent your lambda from continuously retrying if it encounters errors whilst processing a batch, use retryBehaviour to give up on a record once it reaches a certain age, or after a specified number of attempts. See [[StreamRetry]] for more details.

    In order to isolate bad records as part of the retry process, you may also want to bisect problematic batches using bisectBatchOnError. For example:

    const errorHandlingProps: ErrorHandlingProps = {
    bisectBatchOnError: true,
    retryBehaviour: StreamRetry.maxAttempts(5),
    }

    Records which could not be processed successfully can (optionally) be sent to a dead letter queue via deadLetterQueueForSkippedRecords.

    interface StreamErrorHandlingProps {
        bisectBatchOnError: boolean;
        blockProcessingAndRetryIndefinitely?: false;
        deadLetterQueueForSkippedRecords?: IEventSourceDlq;
        retryBehaviour: StreamRetry;
    }
    Index

    Properties

    bisectBatchOnError: boolean
    blockProcessingAndRetryIndefinitely?: false
    deadLetterQueueForSkippedRecords?: IEventSourceDlq
    retryBehaviour: StreamRetry