Magalix autopilot is executing the scalability recommendations on pods and containers. As Magalix builds prediction models for the use of resources inside your pods and containers, it proactively generates recommendations. These recommendations help containers and Kubernetes cluster to keep up with the anticipated increase or decrease in workloads. This proactive scalability makes applications and the cluster always ready for spikes and performance requirements.
Autopilot is a PRO feature
You can enable the autopilot on some or all of your namespaces only when you are a PRO subscriber. As long as
Enabling the autopilot is currently done only through Magalix console. You can enable through the cluster dashboard - see below snapshot.
Autopilot enable/disable in the cluster's dashboard
You can also enable/disable the autopilot in the namespace dashboard.
Autopilot disable/enable in the namespace dashboard
You can disable the autopilot from executing scalability decisions for some of your containers or pods by adding below annotation to your controller's spec file. This annotation will instruct Magalix agent to skip the execution of any scalability decisions for this pod or container.
Skip Autopilot Annotation
Adding this annotation does not disable the analysis of that container's capacity and performance. It will only disable the execution of this pod/container's decisions even if the autopilot is enabled for its namespace.
Magalix analysis your metrics to scale your cluster every hour. If there is a change required in resources allocated to any of your containers
No, currently it works on scaling up or down resources allocated to pods and containers. We do not support yet VM management workflows.
If you turn off the Autopilot up to 5 minutes before the decision execution time, the decision won't be applied. Magalix agent gets the decisions 5 minutes before decisions execution time.