For those who deploy your Phoenix app on Kubernetes: what metric(s) are you using to trigger horizontal pod autoscaling? In my case where most of the traffic is serving GraphQL requests, it looks like memory is a good proxy for load.
For those who deploy your Phoenix app on Kubernetes: what metric(s) are you using to trigger horizontal pod autoscaling? In my case where most of the traffic is serving GraphQL requests, it looks like memory is a good proxy for load.
© Copyright Elixir Forum | Terms | Privacy & Cookies
