Help Center> Cloud Container Engine> User Guide> Old Console> Auto Scaling> Scaling a Workload> Switching from AOM to HPA/CustomHPA for Auto Scaling
View PDF

Switching from AOM to HPA/CustomHPA for Auto Scaling

CCE clusters of v1.15 or earlier support workload scaling based on AOM monitoring data. This function is no longer supported in CCE clusters of v1.17 or later.

If you have configured auto scaling based on AOM, you can switch to HPA/CustomHPA after your cluster is upgraded to v1.17. Note the following differences during the switchover:

  • In AOM-based auto scaling, resource usage rate is calculated based on the limit of a workload, from 0% to 100%.
  • In HPA/CustomHPA-based auto scaling, the usage rate is calculated based on the request of a workload. The range can be greater than 100%.

For example, if the memory request of a workload is 2 GB and the memory limit is 16 GB, a scale-out is triggered as long as the memory usage reaches 50% of the limit (8 GB) in AOM-based auto scaling. In HPA/CustomHPA-based scaling, you need to set the memory usage rate to 400% (16 x 50%/2) to trigger the same scaling.

Parent topic: Scaling a Workload