futurewei-cloud/alcor

[Scalability] Enable Horizontal Pod Autoscaler (HPA) for Alcor deployment

xieus opened this issue · 3 comments

xieus commented

Feature Requirements

The K8s HPA automatically scales the number of Pods in a replication controller, deployment, replica set or stateful set based on observed CPU utilization (or, with custom metrics support, on some other application-provided metrics). This naturally fits into Alcor's performance requirement to scale up the number of replicas for each Microservice when the load increases.

Detailed Requirements

  • Define HorizontalPodAutoscaler including target metrics (like CPU)
  • Tune scaling policies including ScaleDown and ScaleUp policies (like periodSecond, selectPolicy and stabilizationWindowSeconds etc.). Make recommended default behavior for each Microservice.
  • Stress testing a few key Microservices (PM for example) to trigger scale up and down policies.

Reference

K8s HPA: https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/

xieus commented

@kevin-zhonghao Checking if you can help on this issue.

kevin-zhonghao commented

@kevin-zhonghao Checking if you can help on this issue.

NP, the current question is we may find a suitable test environment.

xieus commented

@kevin-zhonghao Checking if you can help on this issue.

NP, the current question is we may find a suitable test environment.

Good point. I will see if we can bring up a few AWS VMs for testing.