Cognito PreSignup lambda is triggered multiple times at once

Question

Hi all, I'm dealing with a frustrating problem with Cognito that feels like a bug to me.

Every so often, maybe every 50 signups or so, our PreSignup lambda is triggered 2 or sometimes even 3 times in a row, and each call happens within ~50 milliseconds of each other. Because of what our PreSignup trigger is doing, we cannot have 2+ signup requests for the same user running at the same time. This usually results in them running into each other -- its essentially like a race condition.

I've been searching online for a solution, the typical answer is that the lambda function is taking longer than the 5 second timeout to respond, but I'm certain this is not the issue because 1) the calls are happening within less than 5 seconds of each other and 2) my function takes a maximum of 3 seconds to respond.

I would like to take this up with AWS themselves because it feels like a bug in their systems to me, but I shouldn't have to pay $29 to open a case and report this bug!!!

Accepted Answer

Hello!

Currently, there are no service incident reports related to this. Could you please check the Lambda function logs and find `Duration` and `Init Duration` values for the function executions you describe?

As you have [read](https://docs.aws.amazon.com/cognito/latest/developerguide/cognito-user-identity-pools-working-with-aws-lambda-triggers.html#important-lambda-considerations), this happens when the function invocation fails, or the execution takes more than 5 seconds. Lambda will then retry two times.

If you have not done it already, check the sections **Optimizing static initialization** and **Memory and computing power** [here](https://docs.aws.amazon.com/lambda/latest/operatorguide/perf-optimize.html), to optimize your function execution times.

Finally, when running logic like yours, my recommendation would be to always add idempotency.

Cognito PreSignup lambda is triggered multiple times at once

相关内容