Retry protocol for signature requests with errors

To ensure an efficient and resilient integration with the Sign'Stash API, it is essential to adopt a retry protocol for submitting signature requests.

This approach tries to recover from temporary errors (e.g. system unavailability), bur must also mindful of situations that cannot be automatically handled (e.g. document for signature request is corrupted. This protocol should provide for limited automatic attempts and a manual analysis mechanism for persistent failures.

Retry Strategy

The ideal integration flow should follow these steps:

1. Request Submission: The client submits the signature request via the Sign'Stash API.

2. Success Verification: If the request is successfully processed, the response is returned normally and the client business flow can move forward with the signed document.

3. Error Detection: If the API returns an error, the client system should implement an automatic retry mechanism. This will handle automatic retries that will handle transient errors on system availability

4. Automated Retries:

   • The system should retry the signature process up to X times (recommendation 3 times).

   • Each retry should occur at an interval of X minutes (recommendation 1 minute) between attempts.

5. Analysis List (Dead Queue):

   • If the error persists after reaching the maximum automated retry attempts, the request should be moved to a dead queue for human operator analysis.

   • An operator (human or AI) should analyze the request and its error cause and decide the best course of action to proceed with document signing. Typical actions:

     • Request new process retry. Again trying up to X times until dropping to the dead queue.

     • Cancel the request and correct the document or initial request that were provoking the error.

Open

Possible Error Scenarios

Below are some situations where the retry protocol is necessary:

• Corrupted Document: If the submitted file is corrupted, it can never be signed, making retries ineffective. The request will be added to the analysis list for an operator to review and replace the document if necessary.

• Read-Only Document: If the file is protected against editing, Sign'Stash will not be able to sign it. In this case, the client will need to remove the restriction before resubmitting the request.

• Temporary Errors: If the API is temporarily unavailable due to network instability or maintenance, controlled automatic attempts can resolve the issue without requiring manual intervention.

By implementing this protocol, we ensure that integration with the Sign'Stash API is reliable, efficient, and resilient to failures, allowing for a better experience for clients and end users.