Opsani Pilot

Pilot enables you to experience the autonomous optimization of one of your cloud-native applications.

See how our machine learning SaaS can improve the performance and reliability of your applications, while reducing your spend simultaneously. Keep reading to see the application/service requirements of Pilot.

Opsani Pilot Pre-Requisites

  • The target cluster must be Kubernetes cluster version 1.16+
  • The cluster must be able to access Docker Hub images and api.opsani.com
  • The cluster needs headroom of approximately 2 times the size of one workload pod.
  • You must have full access to the namespace where the workload resides.
  • The application must be deployed as a Kubernetes Deployment workload resource. This is the most used workload type.
  • The application pods must be stateless and adding an extra pod and removing it periodically should be supported. This is typically supported by most cloud native applications that use a Deployment resource for high availability and scaling.

  • The application pods must have CPU and memory resources explicitly defined. At least one of requests or limits must be specified for each resource type (cpu and memory).
  • The application deployment should not have a vertical pod autoscaler configured. To optimize such an application, you must be able to temporarily disable the vertical pod autoscaler.
  • The application must process HTTP requests over a TCP port exposed on the container. 
  • The application must have at least 2-3 pods and can be scaled up and down.
  • The application receives reasonably sustained traffic during the optimization process (at least 1 request per second).
  • Additional instructions will be made available prior to launching Pilot.