Production swaps servers were restarted manually, beta servers recovered automatically.

Umbrasync database was cleared manually and measures will be taken so that automated recovery is less likely to fail in the future. The swaps API should recovery automatically in the next 5 to 10 minutes, if not, they will be restarted manually.

It is now known why the automated recovery measures failed. The recovery process involved clearing the chain snapshot, which took more than the timeout time given.

At approximately 1AM EDT, Umbrasync started having a noticeable delay in providing real-time data to our swaps solver. Automated recovery measures were attempted at 4AM EDT which failed. The issue is currently being diagnosed