How do I troubleshoot connection timeout issues in DynamoDB?

Short description

When you use AWS SDK to interact with Amazon DynamoDB, you might get an error similar to the following:

"com.amazonaws.SdkClientException: Unable to execute HTTP request: Request did not complete before the request timeout configuration."

This error occurs when the application doesn't get a response from DynamoDB within a set period. DynamoDB might not respond within the time frame for the following reasons:

• Network congestion
• Hardware or software issues
• Communication problems between devices.

Resolution

To troubleshoot DynamoDB connection timeout issues, complete the following tasks.

Update your SDK HTTP timeout settings

Update the following HTTP timeout settings:

• ConnectionTimeout: The default value for ConnectionTimeout in AWS Java SDK is 10 seconds. If it takes longer than the default time to establish a TCP and TLS socket, then you might have network issues. To resolve this issue, increase the ConnectionTimeout value. Or, review your network stability.
• RequestTimeout: RequestTimeout is the time that it takes for the client to complete a single HTTP request. This time includes from the moment a DynamoDB API call is made until a response is received. Set this timeout to a lower setting than the ClientExecutionTimeout.
• ClientExecutionTimeout: ClientExecutionTimeout is the maximum total time allowed for an end-to-end operation, including retries. This setting is turned off by default. If you turn on this option, then set it to a value greater than RequestTimeout to manage the maximum wait time for a DynamoDB response.
• SocketTimeout: SocketTimeout defines the maximum time that the HTTP client waits to read data from an established TCP connection. Set this value to an amount lesser than the RequestTimeout value. For operations like BatchWriteItem and BatchGetItem, set SocketTimeout to a larger value. This larger value makes sure that records are returned as UnprocessedItems if an issue occurs at the service end.

Use exponential backoff with jitter

To handle transient issues effectively in DynamoDB, use exponential backoff with jitter for retries. For more information, see Error retries and exponential backoff.

Keep internal caches warm

To reduce latency and avoid timeout errors, it's a best practice to maintain constant traffic or reuse connections. For more information, see Configuring and HTTP client.

When you aren't making requests, send test data traffic to a DynamoDB table. Also, reuse client connections or use connection pooling to keep internal caches warm.

Review your network stability

Check for network stability issues in your infrastructure. View your Amazon Virtual Private Cloud (Amazon VPC) flow logs to check if there was incoming traffic to DynamoDB during the timeframe when you got the error. To monitor application latency, use AWS X-Ray.

Related information

Tuning AWS Java SDK HTTP request settings for latency-aware Amazon DynamoDB applications

HTTP transport configuration