On May 8, 2024 at 1213 UTC, Compute customers began to experience failures when using the Linode API or attempting to log into the Cloud Manager. These failures were caused by a recent configuration change that was implemented to improve accuracy of customer statistics viewable in Cloud Manager.
On May 8, 2024 at 1248 UTC, our Site Reliability team reverted the problematic configuration and confirmed the API was operational at 1255 UTC. At 1308 UTC, it was confirmed that full access to the API and Cloud Manager was restored.
Moving forward, our team will be implementing improved alerting regarding API failures which will provide notifications for downtime as well as more stringent policies regarding deployment of API affecting changes.