Database connections are stable. The root cause appears to be a flood of uncached communication to one database server. We are taking steps to mitigate the issue in the future.

Database connectivity has been restored. We're continuing to monitor the situation.

We're investigating unstable server connections to the database systems.