The API was down overnight between 12:42AM and 1:26AM.

Similar to the two downtime incidents (1, 2) we experienced in February, the issue was that our authentication database failed due to a routine issue with our cloud provider. We were notified of the issue swiftly, but as the error was not on our end, we were unable to take any action but to wait for our cloud provider to fix the issue.

As mentioned in our previous downtime reports, we are working on cutting over to a new database layer in order to remove the potential risk of downtime. Unfortunately, this process is not yet complete, which is why we are still at risk of downtime when this system fails. We will continue to work on switching over to the new system as swiftly as possible in order to make the API robust against such errors in the future.

Thanks again for bearing with us as we work through these issues!