How Lambda processes records from Amazon Kinesis Data Streams - Amazon Lambda
Polling and batching streams Example event
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

How Lambda processes records from Amazon Kinesis Data Streams

You can use a Lambda function to process records in an Amazon Kinesis data stream. You can map a Lambda function to a Kinesis Data Streams shared-throughput consumer (standard iterator), or to a dedicated-throughput consumer with enhanced fan-out. For standard iterators, Lambda polls each shard in your Kinesis stream for records using HTTP protocol. The event source mapping shares read throughput with other consumers of the shard.

For details about Kinesis data streams, see Reading Data from Amazon Kinesis Data Streams.

Note

Kinesis charges for each shard and, for enhanced fan-out, data read from the stream. For pricing details, see Amazon Kinesis pricing.

Topics

Polling and batching streams

Lambda reads records from the data stream and invokes your function synchronously with an event that contains stream records. Lambda reads records in batches and invokes your function to process records from the batch. Each batch contains records from a single shard/data stream.

By default, Lambda invokes your function as soon as records are available. If the batch that Lambda reads from the event source has only one record in it, Lambda sends only one record to the function. To avoid invoking the function with a small number of records, you can tell the event source to buffer records for up to 5 minutes by configuring a batching window. Before invoking the function, Lambda continues to read records from the event source until it has gathered a full batch, the batching window expires, or the batch reaches the payload limit of 6 MB. For more information, see Batching behavior.

Warning

Lambda event source mappings process each event at least once, and duplicate processing of records can occur. To avoid potential issues related to duplicate events, we strongly recommend that you make your function code idempotent. To learn more, see How do I make my Lambda function idempotent in the Amazon Knowledge Center.

Configure the ParallelizationFactor setting to process one shard of a Kinesis data stream with more than one Lambda invocation simultaneously. You can specify the number of concurrent batches that Lambda polls from a shard via a parallelization factor from 1 (default) to 10. For example, when you set ParallelizationFactor to 2, you can have 200 concurrent Lambda invocations at maximum to process 100 Kinesis data shards (though in practice, you may see different values for the ConcurrentExecutions metric). This helps scale up the processing throughput when the data volume is volatile and the IteratorAge is high. When you increase the number of concurrent batches per shard, Lambda still ensures in-order processing at the partition-key level.

You can also use ParallelizationFactor with Kinesis aggregation. The behavior of the event source mapping depends on whether you're using enhanced fan-out:

  • Without enhanced fan-out: All of the events inside an aggregated event must have the same partition key. The partition key must also match that of the aggregated event. If the events inside the aggregated event have different partition keys, Lambda cannot guarantee in-order processing of the events by partition key.

  • With enhanced fan-out: First, Lambda decodes the aggregated event into its individual events. The aggregated event can have a different partition key than events it contains. However, events that don't correspond to the partition key are dropped and lost. Lambda doesn't process these events, and doesn't send them to a configured failure destination.

Example event

{
    "Records": [
        {
            "kinesis": {
                "kinesisSchemaVersion": "1.0",
                "partitionKey": "1",
                "sequenceNumber": "49590338271490256608559692538361571095921575989136588898",
                "data": "SGVsbG8sIHRoaXMgaXMgYSB0ZXN0Lg==",
                "approximateArrivalTimestamp": 1545084650.987
            },
            "eventSource": "aws:kinesis",
            "eventVersion": "1.0",
            "eventID": "shardId-000000000006:49590338271490256608559692538361571095921575989136588898",
            "eventName": "aws:kinesis:record",
            "invokeIdentityArn": "arn:aws-cn:iam::123456789012:role/lambda-role",
            "awsRegion": "us-east-2",
            "eventSourceARN": "arn:aws-cn:kinesis:us-west-2:123456789012:stream/lambda-stream"
        },
        {
            "kinesis": {
                "kinesisSchemaVersion": "1.0",
                "partitionKey": "1",
                "sequenceNumber": "49590338271490256608559692540925702759324208523137515618",
                "data": "VGhpcyBpcyBvbmx5IGEgdGVzdC4=",
                "approximateArrivalTimestamp": 1545084711.166
            },
            "eventSource": "aws:kinesis",
            "eventVersion": "1.0",
            "eventID": "shardId-000000000006:49590338271490256608559692540925702759324208523137515618",
            "eventName": "aws:kinesis:record",
            "invokeIdentityArn": "arn:aws-cn:iam::123456789012:role/lambda-role",
            "awsRegion": "us-east-2",
            "eventSourceARN": "arn:aws-cn:kinesis:us-west-2:123456789012:stream/lambda-stream"
        }
    ]
}