Skip to main content
Version: v1.0.1

Savings

The Cost Saving Recommendations module analyzes configured environment and identifies actionable opportunities to reduce unnecessary spending. It uses aggregated usage metrics, resource configurations, cost patterns, and optimization signals across user's cloud providers to deliver intelligent, prioritized savings suggestions.

This feature helps user:

  • Identify idle or underutilized resources
  • Detect over-provisioned compute/storage
  • Spot unused reservations or discounts
  • Recommend right-sizing or cleanup actions
  • Estimate potential monthly and annual savings

Key Capabilities​

Automated Cost Optimization Analysis​

The system continuously scans cloud resources (compute, storage, databases, networking, etc.) to detect:

  • Low-utilization nodes or pods
  • Unattached or orphaned persistent volumes
  • Idle load balancers
  • Over-allocated persistent volumes
  • Inefficient SKUs or instance sizes
  • High-cost anomaly patterns

Unified View Across Clouds​

All recommendations are displayed in a single unified dashboard, regardless of resource providers (AKS, EKS or custom cluster)

Actionable Recommendations​

Each recommendation includes:

  • Description of the issue
  • Recommended remediation action
  • Estimated savings
  • Severity/Priority level
  • Resource metadata (cluster, namespace, tags or group)
  • Last analysis timestamp

Integration With Cost Groups & Cost Insights​

Recommendations can be filtered and organized by:

  • Cost Groups
  • Cloud Accounts
  • Projects / Resource Groups
  • Teams, business units, or tags
  • Services and resource types

This allows teams to quickly identify optimization opportunities within their own scope.

  1. Open portal
  2. From the left navigation pane, select Savings
  3. Select View Details to explore details for respective recommendation

Recommendation Types​

  • Abandoned Workloads: Pods with little or no traffic may be abandoned. Consider scaling down, deleting, or notifying owners.
  • Right Size Local Disk: Underutilized node disks can be resized to save costs. Launch new nodes with smaller disks on next turndown.
  • Right Size Container: Save money by reconfiguring the pod resource allocations based on actual usage patterns of users.
  • Unclaimed Persistent Volumes: Unused volumes not claimed by any workload. Consider deleting or moving to cheaper storage.
  • Right Size Node: Underutilized nodes can be resized to save cost.
  • Zombie Pods: Save cost by removing pods not managed by any controller to improve effective resource utilization.
  • Reserved Instances: Billing discount applied to on-demand cloud instances in exchange for a commitment to use a specific capacity for a one- or three-year term.
  • Stale Nodes: Nodes no longer actively participating in the cluster's operations but are still registered with the control plane.
  • Resource Quota: Constraints that limit aggregate resource consumption per namespace.
  • Enable Autoscaler: Automatically adjusts cluster capacity based on workload demands to optimize resource utilization and reduce costs.