DynamoDB Streams Kinesis Adapter for Java

DynamoDB Streams Kinesis Adapter implements the Amazon Kinesis interface so that your application can use KCL to consume and process data from a DynamoDB stream. You can get started in minutes using Maven.

• DynamoDB Streams Developer Guide
• Amazon Kinesis Client Library GitHub and Documentation
• DynamoDB Forum
• DynamoDB Details
• Issues

Features

• The DynamoDB Streams Kinesis Adapter for Amazon Kinesis Client Library (KCL) is the best way to ingest and process data records from DynamoDB Streams.
• The KCL is designed to process streams from Amazon Kinesis, but by adding the DynamoDB Streams Kinesis Adapter, your application can process DynamoDB Streams instead, seamlessly and efficiently.

Release Notes

Latest Release (v1.6.1)

• Upgrades AWS Java SDK to version 1.12.710.
• Adds dependency on Lombok 1.18.32.
• Falling back to Maven Central repository for DynamoDBLocal library.

Release (v1.6.0)

• Upgrades Amazon Kinesis Client Library (KCL) to version 1.14.9. Customers can now use DynamoDB Streams Adapter with KCL version 1.14.9. However, DynamoDB Streams Adapter does not inherit performance optimizations like support for child shards, shard synchronization, deferred lease clean-up available in KCL.
• Fixes the bug which was causing errors in DynamoDB Streams Adapter with KCL version 1.14.0.
• With upgrade to KCL version 1.14.9, the default shard prioritization strategy has been changed to NoOpShardPrioritization. To retain the existing behavior, DynamoDB Streams customers should explicitly update the shard prioritization strategy to ParentsFirstShardPrioritization if there was no explicit override done in the application.
• Upgrades jackson-databind to version 2.12.7.1
• This release uses Apache 2.0 license.

Release (v1.5.4)

• Upgrades AWS Java SDK to version 1.12.130
• Upgrades jackson-databind to version 2.12.6.1
• Fixes logging in DynamoDBStreamsShardSyncer to log only the problematic shardId instead of logging all the shardIds

Release (v1.5.3)

• Upgrades jackson-databind to version 2.9.10.7
• Upgrades junit to version 4.13.1
• Upgrades AWS Java SDK to version 1.11.1016

Release (v1.5.2)

• Upgrades jackson-databind to version 2.9.10.5
• Updates StreamsWorkerFactory to use KinesisClientLibConfiguration billing mode when constructing KinesisClientLeaseManager.

Release (v1.5.1)

• Restores compile compatibility with KCL 1.13.3.
• Fixes a performance issue that arised when using v1.5.0 with KCL 1.12 through 1.13.2.
• Fixes a defect where MaxLeasesForWorker configuration was not being propagated to StreamsLeaseTaker.
• Finished (SHARD_END) leases will now only be delete after at least 6 hours have passed since the shard was created. This further reduces the chances of lineage replay.

Release (v1.5.0)

• Introduces the implementation of periodic shard sync in conjunction with Amazon Kinesis Client Library v1.11.x (KCL). The default shard sync strategy is to discover new/child shards only when a consumer completes processing a shard. This default strategy constrains horizontal scaling of customer applications when consuming tables with 10,000+ partitions due to increased DescribeStream calls. Periodic shard sync guarantees that only a subset of the fleet (by default 10) will perform shard syncs, and decouples DescribeStream call volume from growth in fleet size.

• Improves inconsistency handling in DescribeStream result aggregation by fixing any parent-open-child-open cases. This ensures that shard sync does not fail due to an assertion failure in KCL on this type of inconsistency.

• Modifies finished shard lease cleanup mechanism. Leases for shards that have been completely processed are now deleted only after all their children shards have been completely processed. This will prevent shard lineage replay issues, instances of which have been reported in the past by some customers.

• Introduces StreamsLeaseTaker with improved load-balancing of leases among workers.

  • SHARD_END and non-SHARD_END check-pointed leases are balanced independently.
  • Leases are now stolen evenly from other workers instead of from only the most loaded worker. MaxLeasesToStealAtOneTime no longer needs to be specified by users. It is now determined automatically based on the number of leases held by the worker. The user-specified value for this is no longer used.

• Users should continue using factory methods from StreamsWorkerFactory to create KCL Worker as specified in the guidance of Release v1.4.x.

• We strongly recommended that you create only one worker per host in your processing fleet to get optimal performance from DynamoDB Streams service.

Release (v1.4.x)

• This release fixes an issue of high propagation delay of streams records when processing streams on small tables. This issue occurs when KCL ShardSyncer is not discovering new shards due to server side delays in shard creation or in reporting new shard creation to internal services. The code is implemented in a new implementation of IKinesisProxy interface called DynamoDBStreamsProxy which is part of the latest release.
• This release requires Kinesis Client Library version >= 1.8.10. Version 1.8.10 has changes to allow IKinesisProxy injection into the KCL Worker builder which is required by DynamoDB Streams Kinesis Adapter v1.4.x for injection of DynamoDBStreamsProxy into the KCL worker during initialization. Please refer to Kinesis Client Library release notes for 1.8.10 for more information.
• Suggested AWS Java SDK version >= 1.11.218
• It is highly recommended to configure Kinesis Client Library with MaxRecords = 1000 and IdleTimeInMillis = 500 to optimize DynamoDB Streams costs.

Guidance for injecting DynamoDBStreamsProxy into KCL worker when using DynamoDB Streams Kinesis Adapter v1.4.x.

To fix high propagation delay problems, opt-into using DynamoDBStreamsProxy (instead of the default KinesisProxy) by using the StreamsWorkerFactory factory method (shown below). This injects an instance of DynamoDBStreamsProxy into the created KCL worker.

       final Worker worker = StreamsWorkerFactory
           .createDynamoDbStreamsWorker(
               recordProcessorFactory,
               workerConfig,
               adapterClient,
               amazonDynamoDB,
               amazonCloudWatchClient);

Getting Started

1. Sign up for AWS - Before you begin, you need an AWS account. Please see the AWS Account and Credentials section of the developer guide for information about how to create an AWS account and retrieve your AWS credentials. You don’t need this if you’re using DynamoDB Local.
2. Minimum requirements - To run the SDK you will need Java 1.8+. For more information about the requirements and optimum settings for the SDK, please see the Java Development Environment section of the developer guide.
3. Install the DynamoDB Streams Kinesis Adapter - Using Maven is the recommended way to install the DynamoDB Streams Kinesis Adapter and its dependencies, including the AWS SDK for Java. To download the code from GitHub, simply clone the repository by typing: git clone https://github.com/awslabs/dynamodb-streams-kinesis-adapter.git, and run the Maven command described below in "Building From Source". You may also depend on the maven artifact com.amazonaws:dynamodb-streams-kinesis-adapter.
4. Build your first application - There is a walkthrough to help you build first application using this adapter. Please see Using the DynamoDB Streams Kinesis Adapter to Process Stream Records.

Including as a Maven dependency

Add the following to your Maven pom file:

<dependency>
    <groupId>com.amazonaws</groupId>
    <artifactId>dynamodb-streams-kinesis-adapter</artifactId>
    <version>1.5.1</version>
</dependency>

Building From Source

Once you check out the code from GitHub, you can build it using Maven: mvn clean install