Retry options for REST API Integrations
213
Views
2
Comments
New
Data & Integrations
  • What do you need?
    • Retry options for REST integrations. They should include
      • The number of times a retry should be performed before failing.
      • A duration of time (in seconds) to wait between each retry.
      • Specify which status codes to perform retries on. For instance, you may want to fail immediately on 400 (Bad Request) or 401 (Unauthorized) but would like to retry on codes such as 408 (Request Timeout), 429 (Too Many Requests), 503 (Service Unavailable), 504 (Gateway Timeout) etc. This should be a list of status codes that can be specified at the developer's discretion.
      • Potential Stretch Option - Callback for OnBeforeRetryAttempt to allow the request to fix a potential expected failure or log additional details for troubleshooting.
  • What could it improve?
    • It would increase the stability of REST communication and reduce application errors caused by transient errors on a REST API that the OutSystems developer may not have control over.
  • Who would benefit from that?
    • Anyone that uses REST integrations could benefit from this additional control.

Yes, this is a great idea. No API Producer is 100% perfect and there are always unexpected failures. Adding resilience natively in the platform helps developers avoid spending considerable time coding this retry logic.   

Something similar that I used to do in .NET world using polly it would be great to have this in Outsystems!