Aggregated Selection

The AggregatedSelection is an advanced feature designed to handle complex filtering logic that goes beyond simple ResourceSelector matches. Instead of just selecting individual resources, it allows you to analyze and make decisions based on a group of related resources.

Use the AggregatedSelection when your selection criteria depend on the state or relationship between multiple resources of different kinds. A great use case is when we need to enforce a policy or automate a task based on the interaction of various components within your application.

Example: Enforcing HPA-Deployment Match

The provided example is a perfect illustration of this. It's not just checking for the presence of a Deployment or an HorizontalPodAutoscaler (HPA). Instead, it looks at both types of resources together to find an inconsistency: an HPA that exists without a matching Deployment.

The AggregatedSelection Lua function allows you to:

1. Gather all relevant resources: The function receives an array of both Deployment and HorizontalPodAutoscaler resources that were selected by the resourceSelectors.
2. Perform complex logic: The script can iterate through all the HPAs and check if each one corresponds to an existing Deployment in the list. This isn't possible with a simple ResourceSelector that only evaluates one resource at a time.
3. Return the filtered result: The function returns only the HPAs that do not have a corresponding Deployment. This allows the EventTrigger to act specifically on those mismatched resources, perhaps to trigger a notification or a cleanup action.

This level of inter-resource analysis is the key differentiator of AggregatedSelection. It moves Sveltos from simple, per-resource matching to a powerful, group-based automation engine.

The Lua function must return a struct with:

• resources field: slice of matching resorces;
• message field: (optional) message.

---
apiVersion: lib.projectsveltos.io/v1beta1
kind: EventSource
metadata:
  name: sveltos-service
spec:
  collectResources: true
  resourceSelectors:
  - group: "apps"
    version: "v1"
    kind: "Deployment"
  - kind: HorizontalPodAutoscaler
    group: "autoscaling"
    version: v2
  aggregatedSelection: |
      function getKey(namespace, name)
        return namespace .. ":" .. name
      end

      function evaluate()
        local hs = {}
        hs.message = ""

        local deployments = {}
        local autoscalers = {}
        local deploymentsWithNoAutoscaler = {}

        for _, resource in ipairs(resources) do
          local kind = resource.kind
          if kind == "Deployment" then
            key = getKey(resource.metadata.namespace, resource.metadata.name)
            deployments[key] = true
          elseif kind == "HorizontalPodAutoscaler" then
            table.insert(autoscalers, resource)
          end
        end

        -- Check for each horizontalPodAutoscaler if there is a matching Deployment
        for _,hpa in ipairs(autoscalers) do
            key = getKey(hpa.metadata.namespace, hpa.spec.scaleTargetRef.name)
            if hpa.spec.scaleTargetRef.kind == "Deployment" then
              if not deployments[key] then
                table.insert(unusedAutoscalers, hpa)
              end
            end
        end

        if #unusedAutoscalers > 0 then
          hs.resources = unusedAutoscalers
        end
        return hs
      end