Dear All
i have one question about HPA (Horizon Pod Autoscaler)
i have one HorizontalPodAutoscaler manifest and one POD deployment with resource tag. how this deployment link to HPA manifest and use Utilization in HPA manifest?
BR
HPA’s leverage whatever metrics you tell it to use to scale whatever resource you target in your config.
To leverage HPA’s effectively, you will probably need to learn about the metrics pipeline. Getting the prometheus charts installed manually and tinkering with custom metrics might help in being “hands on” with the api to understand it.
I think type: Utilization and type: AverageValue are leveraged by this math that is immediately illustrated in the algorithm details.
desiredReplicas = ceil[currentReplicas * ( currentMetricValue / desiredMetricValue )]
And I think type: Value does not end up being victim to the replica count, but is acted upon based on it’s actual value.