Proxy route: PROTOCOL_ERROR / INTERNAL_ERROR from peer

Notehub.io
1

I have a strange error that I’ve been trying for several hours to diagnose, and have tried many things without success.

I have a proxy route configured. A “web.get” request from Notecard shows this:

error routing to <url-redacted>: Get "<url-redacted>": stream error: stream ID 5; INTERNAL_ERROR; received from peer

Or sometimes I see PROTOCOL_ERROR.

The HTTP response code in notehub is 0, while Notecard shows a 503.

I have tested the URL (as printed in the error message) with curl and in a browser and both work as expected, both with HTTP/1.1 and HTTP/2.

The proxy route doesn’t do anything other than forward to the URL given. No headers, no jsonata etc. The cloud function is as simple as can be - just returns a success result with a json body. I’ve also tried with other cloud functions that are in active use via regular HTTP routes, and these too generate the same error when used as the proxy endpoint. I also see the same error when I use a url that points to a non-existent cloud function.

The problem appeared 37 hours ago at the time of writing (around 9pm PST, Nov 19.)

Thinking it might be project-related, I also tried creating a proxy route in a different notehub account and project, but same result.

Routing to the Notehub API works, but for some reason not to my cloud functions. These were working fine prior to the error appearing yesterday, and I’ve reduced as much as possible to a minimal test example.

So puzzling! I wonder if you have more details about the error in server logs etc.. Can DM any project details needed.

Thanks!

3

Hey @devElert,

As an update we’re looking into this. It seems to be related to us sending an empty object {} in the body of the GET, but we’re not entirely sure yet. I’ve got it logged and we’ll get back with you.

TJ

1 Like
4

Hey @devElert,

fyi we deployed a fix for this yesterday. If you’re still having issues let me know.

TJ

1 Like
5

Thanks for the update and for the fix! I’ve moved on to other things (and deployed a workaround for the original problem) but I’ll let you know the results when I circle back to that. I do have other proxy routes recently developed and deployed, with endpoints also in the same Firebase project, and they are working fine, which may be an indication the original issue has been fixed.

Happy New Year!

1 Like