Exam AWS Certified Developer Associate topic 1 question 62 discussion

It's A https://docs.aws.amazon.com/step-functions/latest/dg/connect-to-resource.html#connect-wait-example

A) Correct - AWS Step Functions supports the callback pattern using the .waitForTaskToken feature. This approach allows the workflow to pause and wait for an external process or service to send a callback signal before continuing.

This is from AWS : "MessageBody": { "Message": "Hello from Step Functions!", "TaskToken.$": "$$.Task.Token" } The token is being passed in the message body

B. Use the .waitForTaskToken option in the Lambda function task state. Pass the invocation request. Here's why this is the appropriate choice: The .waitForTaskToken option allows you to pause the execution of the Step Functions workflow until a specific task token is returned by the external service. In this scenario, when the Lambda function is invoked to call the external service, it can provide the task token as part of its invocation request. The external service can then take the required time to validate the data, and when it's ready to respond, it should include the task token in its response. When the Step Functions workflow receives the response with the task token, it will be able to correlate the response with the specific task in the workflow and resume its execution accordingly.

Can someone explain to me why we have to use the .wait ForTaskToken? Can be the C correct? Imagine that there are two tasks A and B, if the task A call synchronously a lamda function, then B needs to wait A to finish for definition. And this because A and B are two consecutive tasks and non parallel

The .waitForTaskToken option allows Step Functions to pause the workflow until a specific task is completed. In this case, when the Lambda function is invoked, it can generate a token using the TaskToken field provided by Step Functions. This token is then passed in the response body of the Lambda function, indicating that the task is not complete yet. Step Functions will then wait for the task to complete and provide the token back to the Lambda function in the input of the next state, allowing it to resume execution.

The best option to meet the requirement of pausing the Step Functions workflow and waiting for the response from the external service is B. Use the .waitForTaskToken option in the Lambda function task state and pass the invocation request. Option A is incorrect because the .waitForTaskToken option should be used instead of .wait for the Lambda function task state. The .waitForTaskToken option allows Step Functions to provide a task token to the Lambda function, and the Lambda function can use this token to notify Step Functions when it's done processing.

The answer is A and I am 100% sure. The task token is passed in the body. Check this page to see for yourself: https://docs.aws.amazon.com/step-functions/latest/dg/connect-to-resource.html#connect-wait-example

C: lambda will pass request to third party service and return to step. step will continue before that third party service finish. D: step will send request to lambda in asynchronous and will wait. but it can not know for haw much time it will wait because processing time is unknown so we can not use Wait state. A is correct.

A is correct

A :https://docs.aws.amazon.com/step-functions/latest/dg/connect-to-resource.html#connect-wait-example

Just curious to know, wouldn't the Lambda time out? as the question says "the external service will take a few hours" and Lambda cannot wait that long. I was leaning more towards Option D. Can someone explain what is wrong with option D?

i would go for A https://instil.co/blog/step-functions-and-task-tokens/

I think it's B... https://docs.aws.amazon.com/step-functions/latest/dg/connect-lambda.html

Maybe B. Token is get accessed from context object for that reason A is not accurate for me. The other ones not seems be the correct.