Around 16:30 PT we received a report of a customer that could not connect to it's Metabase instance. Our engineering team identified an issue on the application database which showed that the clusters weren't available in our cloud provider (CSP). After further investigation they found out that a cluster resize didn't apply correctly due to a configuration mismatch between the configuration scripts and the CSP offering resulting in downtime. The configuration script was modified to be compatible with the CSP and the database instances came back online.