Scaling on queue length with the HPA
WHAT IT TESTS: external-metric autoscaling. OUTLINE: expose queue length through an external metrics adapter behind the metrics API, point the HPA at that external metric with a target per pod; KEDA packages this.
WHAT IT TESTS: how the HPA consumes non-resource metrics. ANSWER OUTLINE: the HPA only reads metrics through aggregated APIs, so you run an adapter that implements the external.metrics.k8s.io API and surfaces queue length from your monitoring or the broker. Configure the HPA with an External metric and a target value per pod; it divides current by target to compute replicas. KEDA bundles scalers for common queues. RED FLAG: assuming the HPA polls the queue itself or supports only CPU and memory.
Read the original → interview
- #kubernetes
- #hpa
- #autoscaling
- #keda
- #metrics
Get five bites like this every day.
Tezvyn delivers a daily feed of 60-second tech bites with quizzes to lock in what you learn.