When software engineer Ademayowa Adeniyi needed to reboot the VPS hosting his Kubernetes cluster, he knew his applications and internal monitoring tools would go offline with it.
Before starting the reboot, he set up UptimeRobot to monitor his backend service and health endpoint. That gave him a way to measure the downtime from outside the cluster, even while Grafana was unable to collect metrics.
UptimeRobot recorded 15 minutes and 14 seconds of downtime. In our latest Community Spotlight, Ademayowa walks Eleftheria through his setup, explains what he learned from the reboot, and discusses how shorter monitoring intervals could give him a more precise picture of recovery.
He also shares what he’d change next time, including adding a status page to his application.
Watch the interview below, or read Ademayowa’s original LinkedIn post for more details.