Abstract
Kubernetes clusters don’t end at the data centre. As workloads push to the edge — IoT gateways, retail stores, remote sites — observability becomes both more critical and harder to achieve. Network constraints, limited compute, and intermittent connectivity all conspire against traditional monitoring approaches.
This talk explores what observability actually looks like at the edge of a cluster: what to collect, what to leave behind, and how to get signals back to a central platform without overwhelming your uplink or your budget.
Recording not yet available
Slides not yet available