Today (Sep 20th) at 15:09 UTC we had an issue that dragged the performance of the application to near zero, rendering it unable to process new requests. Our team was quickly notified and acted to mitigate the problem, and our services were back online at 15:23 UTC. The underlying issue has also been solved.
The team was focused on recovering the application as quickly as possible, so we had no time to properly notify our customers about the problem until it was solved. Our apologies for that.