EPYC-501: Refurbishment & 14-Day Credit

Yesterday, our scheduled maintenance on EPYC-501 didn’t go as planned—and we’d like to address this openly.

The planned motherboard replacement was intended to resolve the recurring crashes of the past few weeks and was scheduled to take one hour of downtime. However, despite a BIOS update, the replacement board failed to recognize the CPU, and during the subsequent rollback, it became apparent that the original board was also no longer capable of stable operation. We subsequently migrated all affected VMs to our 3rd-generation EPYC systems. Instead of the planned one hour, the servers were unavailable for about four hours.

We would like to sincerely apologize for the significantly longer downtime than announced, for the frequent crashes, and for the general instability on EPYC-501 during the period leading up to this. Furthermore, the fact that all 5th-gen EPYC servers are currently running temporarily on 3rd-gen systems rather than on the booked 5th-gen hardware does not meet the standards we set for ourselves.

We are working around the clock to restore stable operation of the 5th-Gen system as quickly as possible—as of now, we plan to have this completed by the end of August. We’ll announce the migration back to the 5th Gen systems in a timely manner as a separate maintenance window. Until then, all VMs will run stably on the 3rd-Gen systems; data and configurations will be fully preserved.

All customers with an EPYC 5th-gen server will receive a 14-day service credit. This will be automatically applied to every affected server—you don’t need to take any further action.

Thank you for your patience over the past few days, and especially yesterday.