Autoscaling support for Helm release

[mg-i-mat-bap]

Hello everyone,

I need to configure autoscaling for Infinispan using Kubernetes’ HorizontalPodAutoscaling (HPA). Looking through Infinispan’s documentation, I noticed that support for this feature is provided for releases made through the operator, but not for releases made through Helm Chart. Considering the operational nature of Infinispan, which is an in-memory caching system, having the ability to autoscale replicas based on Memory usage/average value would be very useful.

At the current state, this result is not achievable through the use of an external Autoscaler, as in the values.yaml the default value of the Replicas field is set to 1 by default.

Given these circumstances, is it possible to enhance the Helm Chart to support autoscaling feature? This feature would greatly improve the operational efficiency of Infinispan.

Thank you for your attention to this matter. I look forward to hearing your thoughts on this proposed enhancement.

[ryanemerson]

Hi,

I need to configure autoscaling for Infinispan using Kubernetes’ HorizontalPodAutoscaling (HPA). Looking through Infinispan’s documentation, I noticed that support for this feature is provided for releases made through the operator, but not for releases made through Helm Chart.

The Operator doesn't currently support HPA, although this something we have discussed adding in the past. See infinispan/infinispan-operator#749 for more details.

At the current state, this result is not achievable through the use of an external Autoscaler, as in the values.yaml the default value of the Replicas field is set to 1 by default.

Given these circumstances, is it possible to enhance the Helm Chart to support autoscaling feature? This feature would greatly improve the operational efficiency of Infinispan.

The Helm Chart provisions a StatefulSet so it should be possible to create a HorizontalPodAutoscaler CR for the StatefulSet kubernetes/kubernetes#44033