I’ve now seen 2 large instances where multiple scenarios - all of which are triggered by an Airtable ‘Watch for New Records’ module will throw this same error over a 15-30 minute period.
I’ve flagged to Make support, but received this in an auto-reply the next day:
Curious if others are facing the same issue (and if that was what’s fueling the massive number of requests), but haven’t seen any other threads citing it 
Hi @Pat_Martinson,
Could you send some screenshots of the setup of the module that gives you this error? Also please add a screenshot of the input data for the execution that fails.
Thanks
Loic
I can certainly try! However It happens across so many different scenarios, it’s hard to pick the ‘right’ ones to screenshot. Some are very simple:
While some are more complex:
Any other specific details you’d like to see?
Thanks, did you change the structure of the table your watch modules are querying?
If adding fields ‘counts’ as changing the structure, then yes we have for some, but not for others. It’s happening across multiple tables (but I believe they’re all in the same base). Again, it’s happening on so many different automations that we haven’t been able to systematically track it 
Ok I suspect there are two things going on here:
So basically an error on top of an error.
My suggestion #1:
Try to reset the starting point for your watch module. Right click on the module, click on “choose where to start”, select your new starting point, click ok and then try to run your scenario.
If this does not work try to do this:
This will not solve the problem but you will be able to see what Airtable is saying about the issue. This may shed some light.
OK, I will give this a shot, and thanks for the detailed answer!
One thing worth noting (and I may not have mentioned before): this is happening sporadically, and it will be a small-to-medium “burst” of several scenarios throwing the error at the same time.
Then they all run properly again. So I think it will be harder to catch the error, but I’ll give it a shot.
(and as a final clarification: we’re seeing 502 or 503 timeout errors from our AT modules occasionally, but the error code is distinct from these)
How many times did this happen? Could it be related to some Airtable or Make outages? It sounds like it
We’ve had 3 instances where this happens across several different scenarios:
So at this point, we haven’t seen the error come up since Dec 2nd, and more importantly it seems like this is a corner-case issue, that hasn’t been seen widely by other users.
I think those are both good things? 
Wanted to circle back with an update from Make Support:
As I have further checked the logs with this same error, it seems that the issue is within the Airtable server as it throws error 503. I would suggest reaching out to their support team if the issue still persists
We have not seen the issue again, but if it’s tied to 503 errors, I think it means we need to cut down our Watch Records triggers.
Thanks again to @loic.wiseflow for your insights, and hope this is helpful to others if you encounter the same issue.
Heya @Pat_Martinson 
I just wanna jump in real quick to say that we really appreciate you stepping back into the community and sharing the message you got from the Support team 
This can, indeed, be super helpful to someone one day. Thank you for thinking about the other Makers 
Hi @pat_Martinson, I have the same problem 503 and 502 error but I don’t understand their meaning, how can I get an explanation of whats going on? Thansk for your help!
