Beginning yesterday we had a rash of connection and server problems.
The cause of the downtime and interruptions turned out to be two unrelated heavy accesses operations that caused excessive load on the servers. We believe we’ve already addressed the first access pattern in the upcoming release so that it won’t be an issue once deployed, and we’ve identified what caused the second problem and will be writing a patch to handle the situation better in the future.
Terribly sorry for the downtime, we think things are “Back to normal” for the time being though.
- anoek