We have now identified and resolved the issue. Bad configuration was pushed to one of our applications, which caused it to consume all available connections to the database. We have pushed a new version of the config that will prevent this issue from happening in the future.
Posted over 3 years ago. Mar 07, 2016 - 14:52 GMT
We had a brief service outage from 11:04 to 11:06, we are investigating now.