Between 2026-11-04 00:02 and 00:50 UTC there were elevated errors on Supavisor Shared Pooler in the us-east-2 region. Errors have subsided and things are back to normal.

Example errors were

ClientHandler: (ECHECKOUTFAILED) checkout failed due to %{"C" => "08006", "M" => "Failed to connect to database: {:error, :timeout}

DbHandler: Connection failed {:error, :timeout} to ~c".......us-east-2.compute.internal"