Runtime failure on function (now resolved) - webhooks lost, deploy blocked, credits wasted - requesting refund
Status: Open · Asked by Bianca Flood on · 0 views
For roughly 24 hours from 21 Aug 2026, workflow invocations on one
function in my account were returning HTTP 500 with a MongoDB
connection error from your infrastructure.
Error: localhost:27017: [Errno 111] Connection refused
(socketTimeoutMS 20000, connectTimeoutMS 20000)
Three symptoms, same apparent root cause:
1. Direct invoke to the function's endpoint returned 500. No run
was logged in Run History.
2. Incoming webhooks from an external service were silently lost -
correctly registered, delivered, but no run created.
3. Deploying this workflow failed with the same Mongo error.
Last successful run before the failure: 21 Aug 2026 16:09:07. A
webhook at 17:03:17 produced nothing, and the function stayed
broken until 22 Aug, when it resolved on its own without any
action from me.
I also had trouble creating a new workflow on 21 Aug (a "server
offline" message) - that's resolved too now. I can't confirm
whether it's the same root cause, but the timing lines up and I'd
like to know if they're related, since it suggests the issue may
have been broader than just the one function's runtime.
Could you confirm what caused this and whether it's fully resolved
on your end?
I spent a great deal of time and approximately $20 in credits diagnosing this - retries, a
duplicate workflow test, and workaround attempts - all caused by
the failure described above, not by any error on my part.
Requesting a credit refund for the wasted spend.
Happy to share the exact org and function IDs privately - can you
reply here or by email so I don't post them in a public thread?
Hello Bianca,
Thank you for reaching out to us.
Kindly share the further details regarding the issue at support@pabbly.com so that our technical team can review the concern in detail and assist you accordingly.