64% of organizations use Kubernetes HPA for autoscaling, but only 20% scale on custom metrics. CPU and memory are lagging indicators that fail to capture real demand for event-driven workers, high-throughput APIs, latency-sensitive services, database-backed services, and business-logic workloads. Custom metrics like queue
Table of contents
When standard CPU and memory resource metrics fall shortWorkload patterns for Kubernetes custom metric scalingStandard approaches to custom metric scalingScaling on custom metrics with Datadog Kubernetes AutoscalingScaling on the right signalSort: