Resolved -
After several hours of monitoring, we confirm the issue has been fully resolved and all services are operating normally.
Our preliminary analysis indicates the problem was caused by a combination of factors related to Timeshift technology, including a software bug, a race condition, and a configuration issue. These factors led certain servers to stop accepting new connections while maintaining existing listener sessions.
Immediate corrective actions were implemented, and additional follow-up measures are scheduled for early next week to prevent recurrence.
Thank you for your patience and understanding.
Dec 6, 16:57 UTC
Monitoring -
Our team has applied mitigation measures and service is back to normal. The issue seems to have been cause by an unexpected situation in the MediaGateway when handling TimeShift mounts.
We are looking into ways of handling such situations so they don't cause a cascading effect in the future.
More information to come.
Dec 6, 15:18 UTC
Update -
The issue was affecting only new connections to the Media Gateways. In order to apply corrective measure to the affected MGs, we need to move all traffic from them to healthy servers. This will cause a disconnection to existing listeners, who should reconnect quickly to new healthy instance.
More information to come.
Dec 6, 13:51 UTC
Investigating -
We are investigating an issue with our streaming servers in Europe which is preventing streams to be delivered to listeners.
Our team is working to redistribute streaming to other locations and unaffected Media Gateway instances.
More information to come.
Dec 6, 13:19 UTC