EC2 instances are expensive on AWS and I love the auto scaling feature. I’m currently trying to configure auto scaling on v2 so baseline is 0 nodes, N num of minutes before it’s needed I send an api call to spin up an instance so the meeting is available right away.
There’s no idle time configuration that I can find. I can change the auto scaling algorithm run to run on a larger gap but that comes at the risk of it not responding to a sudden increased load.
Any ideas? Is this different in v3? We currently have a v2 pro license and I haven’t actually dug too much into it.
Hello @Falpangaea,
In OpenVidu v2, there’s no option to set a minimum uptime for nodes (if that’s what you’re referring to). However, you can disable autoscaling by setting OPENVIDU_PRO_CLUSTER_AUTOSCALING to false, allowing you to manually manage the creation and termination of media nodes, but the cluster won’t autoscale by itself, you will need to manage that.
In OpenVidu v3, the process has changed. We’ve opted to let AWS and the Autoscaling Group handle the creation and deletion of media nodes. Therefore, there should be a parameter available to control the minimum duration for which new nodes remain active.