I have configured our workflow to POST to our API, but the workflow step doesn’t seem to be taking into account returned error statuses.
Here’s the screenshot of the workflow step details
Formerly I have seen webhook failures result in exponential retries. Now the workflow just continues to the next step and completes successfully.
Has exponential retry functionality been removed or is this a regression? Does it depend on the error status returned by the API?
Hello @DBarlow , I will add some top experts into this conversation @Mike_Eastwood @JoshG any recommendations to @DBarlow ?
Thank you very much!
Pam
Hi @DBarlow
To me that looks like the error was at the URL being called by HubSpot… do you have any error logs at the other end?
Mike
Thanks @PamCotton and @Mike_Eastwood .
Yes the error message was generated at the URL being called by HubSpot. The error was caused by a misconfiguration within Hubspot.
I’m OK with the error happening. However, I didn’t expect the workflow to keep running. Apparently that is normal. I have had help from the support team and the answer is here:
https://legacydocs.hubspot.com/docs/methods/workflows/webhook_information
Basically HTTP 4xx error codes aren’t retried (except 429 - Too Many Requests). HTTP 5xx errors are retried.
My workaround is that since I control the endpoint being requested by Hubspot, I’m probably going to raise an alert when the error condition appears.
Thanks for looking into this.
