Our company is an aggregator of multiple payment services fragmented on the web. So instead of businesses integrating multiple services one by one, they will only need to integrate to our API, and we will connect to every one of the products for them. We connect by writing individual scripts for every one of these services.

Initially, all our scripts are hosted in an EC2, but there happen to be some downtime because of hardware failure, so we are considering moving to a serverless architecture where each script would be its own AWS Lambda. However, some of our scripts take longer than 30 seconds to complete and since AWS Lambda HTTP trigger is an API gateway. API gateway has a 30-second timeout limit.

Note we have custom software acting as the middle layer to run the scripts.

E.g., customer A calls our API. This software will then run rules based on the API parameters and choose the correct script to run by initiating an HTTP call to the script. Once the software receives the result from the script, it will mould the response to a consistent one and return it to customer A.

Are there any other serverless suggestions or a better way of designing this to achieve 99.9% uptime while also being easier to maintain?

  • 2
    A technicality: Why does the Gateway timeout matter? Can you make the requests asynchronous, then you are up to Lambda's timeout, which was 5 minutes, but is now 15, I read.
    – joshp
    May 23, 2021 at 6:06


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.