'HPA spawns too many pods

This is my hpa configuration.

apiVersion: autoscaling/v2beta2
kind: HorizontalPodAutoscaler
metadata:
  name: customer-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: customer
  minReplicas: 2
  maxReplicas: 8
  metrics:
    - type: Resource
      resource:
        name: memory
        target:
          type: Utilization
          averageUtilization: 70

The assigned memory and cpu limits for the deployment are like below.

resources:
  requests:
    memory: 660Mi
    cpu: 100m
  limits:
    memory: 860Mi
    cpu: 500m

The number of replicas of the deployment is like below.

replicas: 2

When I listed the hpa, it is showed like below.

NAME                     REFERENCE                           TARGETS         MINPODS   MAXPODS   REPLICAS   AGE
customer-service-hpa     Deployment/customer-service         46%/70%         2         8         4          36d

When I deploy the deployment and run

kubectl top pods

the output is like below.

customer-9fddb6649-24r88                   7m           294Mi
customer-9fddb6649-6x97x                   7m           305Mi
customer-9fddb6649-cpg6g                   56m          309Mi
customer-9fddb6649-ldbb2                   12m          314Mi

Eventhough the load is low, initially pod count is 4. But the given minimum pod is 2.

Is there any issue in deployment files?

kuberneteshpa


Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution Source

Related Questions