On 21 December, 2021 at 15:05 UTC, some of our customers with Block Storage volumes in our Frankfurt data center experienced degraded throughput.
Linode's Incident Response Procedures were immediately engaged, and we identified a combination of underlying configuration issues and hardware failures.
These issues led to the loss of functionality with multiple nodes within our Block Storage cluster during routine maintenance. While the design of our system allows for the loss of multiple nodes without loss of data, this specific failure mode led to performance issues until the misconfiguration issues were resolved.
At 16:00 UTC, Linode solved the underlying issues and customer performance issues were resolved.
We have completed a review of our internal processes surrounding this event, and steps have been taken to help ensure that this incident does not happen again.