We had a 5 minute outage early this morning for 5 minutes at 12:53AM Eastern time. We were doing a deployment and it triggered errors. That error was corrected as quickly as possible. We apologize for the outage.
This deployment we did was part of investigating a tricky bug that is causing 1 error per hour on a database connection. One unlucky user per hour is seeing a message that says they failed to connect to the database and to try again. There are no lost transactions, of course. The unlucky user has to click on the action button of the page they were on again and the transaction will go thru.
We are continuing to work on this issue and will report once it is resolved.
UPDATE: We have found there is an issue in Amazon AWS (our cloud provider) that is also affecting many other users. We have successfully isolated it to happening only in one AWS zone (East – 1a), and have taken that server offline so users will no longer be impacted. Here are some public links –