Which actions should a DevOps engineer take to resolve this delay?

Explanations:

Checking theApproximateAgeOfOldestMessagemetric for the SQS queue helps identify if there is a backlog of messages that are not being processed in time. Increasing the Lambda function concurrency limit will allow more parallel executions, reducing delays in processing orders.

While checking theApproximateAgeOfOldestMessagemetric is useful, configuring a redrive policy is more appropriate for handling messages that cannot be processed (e.g., due to failures), not for addressing delays in processing.

TheNumberOfMessagesSentmetric does not directly address processing delays. Increasing the SQS queue visibility timeout is useful for preventing message duplication but not for addressing processing delays due to Lambda function concurrency or DynamoDB capacity.

Checking theWriteThrottleEventsmetric for the DynamoDB table helps identify if write capacity is being exceeded. If this is happening, increasing the maximum write capacity units (WCUs) for the table’s scaling policy can prevent throttling and delays in inserting orders into DynamoDB.

TheThrottlesmetric for Lambda indicates whether Lambda invocations are being throttled due to concurrency limits. However, increasing the Lambda function timeout does not resolve delays related to processing capacity or DynamoDB throttling.