since a week or two we are experiencing more and more connection issues with the database.
We reached a point where we have these issues several times a day, and I don’t understand where it can come from.
The typicla issues are
502 Bad gateway
504 Gateway Time-Out
To other studios people: do you experience such issues?
To Shotgun people: could you please check everything is ok on your side?
We are seeing this maybe once every couple of days or so. Not to the point of being an obstruction, but it is definitely there.
In my experience this is usually Cloudfront or some other cloud service. Also, probably kinda difficult to diagnose/debug.
A fix has been deployed this morning. Our preliminary testing show that 502s errors are down by 95%. The automatic retries in the Shotgun Python API should handle the rest. We are monitoring the errors rate and take further actions if needed.