Performance Degradation on epyc7543-3 / epyc7543-4

Update timeline

1. investigating Nov 19, 2025, 07:23 AM UTC

   Our internal monitoring system has detected a significant performance degradation on the nodes epyc7543-3 and epyc7543-4. We are actively investigating the root cause of this issue to restore normal operation as quickly as possible. At this time, we are experiencing an unprecedented surge in support tickets related to this matter. Due to this high volume, we may not be able to respond individually to every ticket promptly. We kindly ask that you subscribe to the newsletter thread for this incident to receive continuous updates, rather than opening new tickets. Thank you for your understanding and patience as we work diligently to resolve this.

2. monitoring Nov 19, 2025, 09:11 PM UTC

   We have managed to temporarily distribute the excess load on our EPYC servers, ensuring stable performance across all affected systems. Over the next few days, we are planning to deploy a new EPYC-based server which will enable us to rebalance resources more effectively and further improve overall performance.

3. resolved Dec 28, 2025, 10:55 AM UTC

   Overall performance has stabilized to some extent. To further improve capacity and resilience, a new node will be provisioned on 10 January 2026 at the new FirstColo FRA4 location. Further updates will be provided once the node is online.