Frictionless Kubernetes cost visibility & optimization — even cheaper and easier to deploy

The nOps EKS agent gives you full visibility into your workloads: pod-level usage, labels, and resource consumption — enabling 100% accurate cost allocation and container rightsizing. 

Today, we’re excited to share a significant evolution of the nOps EKS agent designed to reduce complexity, improve performance, and streamline your experience.

What's New

We have completely redesigned the nOps EKS agent architecture. The agent is now purely focused on collecting and transmitting metrics and metadata — no heavy processing or storage inside your cluster.

We’ve removed Prometheus and OpenCost servers from your clusters. Metrics and cost data are now securely ingested via Amazon Managed Service for Prometheus (AMP) and processed on the nOps side — reducing cluster resource usage, operational overhead, and ongoing maintenance.

We’ve also eliminated all local storage requirements. The latest version streams usage data directly to nOps—without EBS, S3, or provisioning. That means lower costs and ultra-easy installs. In sum: 

  • One-line Helm install – quick deployment with no custom config
  • Graviton support – cheaper runtime cost on ARM-based nodes
  • No EBS volume – runs fully in memory; no storage provisioning required
  • No more Datadog by default — Datadog is now optional and used only for advanced debugging

Why the nOps EKS Agent?

Once installed, the nOps agent collects real-time container-level usage data directly from your workloads, for:

Total Kubernetes Visibility

Usage data is enriched with Kubernetes context—like labels, namespaces, and workloads—giving you precise cost allocation across teams, services, and environments.

  • Full visibility into Kubernetes costs across all accounts and clusters
  • Accurate mapping of spend to pods, namespaces, teams, or customers
  • Identification of overprovisioned workloads and excess node capacity
  • Cost tracking with built-in waste detection and actionable rightsizing insights

Automated Container Rightsizing

The nOps EKS agent powers precise container rightsizing based on real-time CPU and memory usage. Historical utilization data—p95, p98, p99, or max—is analyzed to generate intelligent recommendations for each workload. You can tune these recommendations based on your cost or performance goals, and optionally auto-rightsize new workloads as they’re deployed.

How It Works

  1. Deploy the nOps EKS agent: Install the agent using a single Helm command. No custom configuration is required. The agent runs on each EC2 instance in your EKS cluster.

  2. Collect runtime data every minute: The agent captures CPU, memory, and transaction-level data for every container on the node. This data is collected at one-minute intervals throughout the hour.

  3. Stream data directly to nOps At the end of each hour, the agent aggregates the collected data and streams it directly to nOps. There are no EBS volumes or S3 buckets—everything runs in memory and transmits over the network in real time.

  4. Process and enrich data inside nOps. Once received, the data is processed and enriched with Kubernetes context (e.g., labels, namespaces, workloads).

If you’re already on nOps…

Have questions about the nOps EKS agent? Need help getting started? Simply reach out to your Customer Success Manager or visit our Help Center. If you’re not sure who your CSM is, send our Support Team a message.

If you’re new to nOps…

nOps was recently ranked #1 with five stars in G2’s cloud cost management category, and we optimize $2+ billion in cloud spend for our customers.

Join our customers using nOps to understand your cloud costs and leverage automation with complete confidence by booking a demo with one of our AWS experts.