We are seeing consistent healthy connections from the same affected IPs -- function should be fully restored. Performance is improving but we're still working toward full recovery. We'll continue monitoring and post another update once we're fully stable.

We're currently investigating the problem and will be back with updates