We have identified and fixed the underlying issue, which was caused by a database having issues. We have replaced this database with another instance and the service is back online.
To ensure such outages do not happen again, we will continue making our infrastructure more robust and fail-safe.
