On February 24, 2023, at 05:12 UTC, configuration changes to networking hardware in the Newark data center triggered a routing failure. The routing failure impacted a subset of hosts, and delayed any commands issued through the Cloud Manager or the Linode API until connectivity was restored. At 05:42 UTC connectivity between the legacy network and API was restored, and we considered this incident resolved as of 08:59 UTC.
Upon review of this incident, the issue was related to a human error when making the configuration changes as we retire the legacy network. We plan to review and refine our change management policies to ensure such a human error does not occur in the future.