Basic Recommender

[DragonBanana]

The basic recommender is a component that recommends new resource values to all pod scales in the system.

Workflow:

• For each pod scale in the system, store the corresponding pod in a local map.
• For all the pods stored in the map, retrieves the pod metrics and recommends a new cpu and memory resource value.

Features:

• PodScale synchronization.
• Local map containing all pods that should be monitored.
• HTTP client to retrieve metrics.
• Recommend, with a naive logic, new resources.

Future works

• Forward the new resources to the contention manager.
• Other protocols to retrieve metrics.