Previous incidents

This downtime was caused by an edge routing issue, where some requests were being rejected due to unexpected header formatting. We deployed a fix to handle all header formats correctly, resolving the issue immediately.