Since the disk replacement, I/O performance has stabilized significantly. We are therefore closing this incident.
Thank you for your patience.
Posted Jan 27, 2026 - 13:41 CET
Update
We are continuing to monitor for any further issues.
Posted Jan 24, 2026 - 10:42 CET
Monitoring
The NVMe drives were replaced yesterday at 18:36 CET. Since then, system performance has been stable and no further anomalies have been observed.
Please contact support if you notice any remaining issues.
Posted Jan 24, 2026 - 10:41 CET
Update
We will restart node epyc01 in approximately 10 minutes in order to replace a potentially defective NVMe drive.
Further updates will follow once the maintenance has been completed.
Posted Jan 23, 2026 - 18:04 CET
Identified
We currently suspect that one of the installed NVMe drives may be the source of the I/O issues. This is under active investigation, and we will provide an update here as soon as further findings are available.
Posted Jan 22, 2026 - 13:49 CET
Investigating
We are currently investigating I/O-related performance issues on node epyc01. Our engineering team is analyzing disk and storage performance metrics to identify the root cause.
Further updates will be provided as soon as more information becomes available.
Posted Jan 22, 2026 - 13:49 CET
This incident affected: firstcolo FRA4 (EPYC KVM Server Cluster).