Application Server

Widespread recurring 502 Bad Gateway errors detected

Postmortem

We have investigated and found the issue is related to how our upstream provider, Heroku, responds to our database usage. We’re working with them to understand the cause of the issue, but are making changes that we feel will mitigate it even without their assistance. We apologize for the inconvenience.

Löst
Utreder

Over the past few days we’ve had an outage at the same time (8:20pm PT) every 48 hours which lasts 15-20 minutes, then self-resolves.