Cost & Node Analysis

Overview

When you need a closer look into how your Kubernetes clusters are driving cost and efficiency, the Cost & Node Analysis section brings everything into focus. The analysis helps you break down cost drivers at the nodes level and evaluate how effectively your nodes are being utilized.

The page consists of two tabs: Cost Analysis and Node Analysis, each offering a distinct lens into your infrastructure.

Cost Analysis

In the Cost Analysis tab, you're presented with a breakdown of costs specific to the cluster you're viewing. It carries the same intuitive layout as the broader Cost Analyzer in Kubernetes Visibility, but here, your focus is narrowed to a single cluster.

You can see how costs are split across compute, storage, and networking within the cluster. Use familiar groupings and filters to drill into trends and identify shifts in cluster-level spend. You get to stay proactive in managing cost surges before they become expensive problems.

Node Analysis

The Node Analysis tab is where you evaluate how well your underlying compute nodes are performing—both financially and operationally. Whether you’re trying to find underutilized nodes or validate the impact of your spot instance strategy, here you get all the data you need in one place.

Key Metrics

Right at the top, you’ll find essential metrics that give you a snapshot of your node performance:

• Total Compute Cost – See what you're spending to run all nodes within the cluster.

• Node Efficiency – Understand how well you're utilizing provisioned CPU and memory cost.

• Unused Cost – Identify the portion of your spend that went to idle resources.

• Spot Usage Cost – Track the cost saved through your spot instance strategy.

  

Node Overview Table

Dig into each node’s performance and cost with the overview table. It’s designed to help you compare and group nodes effortlessly.

List of Columns

Column	Description
Group by Node Group	Toggle this to organize the data by node groups for clearer analysis.
Node Type	Displays the instance type (e.g., m5.large, c6g.xlarge) used by the node.
Compute Hours	Shows the total number of hours each node has been active.
Cost	Reflects the actual cost incurred by each node.
Efficiency	Highlights how effectively the node's resources are being utilized.
Spot Cost	Indicates the cost specifically incurred from using spot instances.
Spot Coverage %	Represents the percentage of the node's runtime that was covered by spot usage.

You can use this table to identify over-provisioned nodes, validate the balance between spot and on-demand usage, and plan more efficient resource distribution.

Utilization Trend

The Utilization Trend section gives you visual insights into how CPU and memory are used over time—enabling quick identification of inefficiencies.

CPU Utilization

You can compare CPU utilized, CPU unused, and p99 CPU utilization against the total CPU hours. This makes it easier to identify whether certain nodes are being over- or under-utilized, and where right-sizing could cut costs.

Memory Utilization

In a similar view, track your used vs unused memory, helping you spot memory over-allocation early.

Spot vs On-Demand Usage Cost Trend

The following trend chart shows how your costs are split between spot and on-demand nodes over time. By following this graph, you can evaluate how well you're balancing cost savings with workload reliability. If spot coverage is too low, this view makes it clear—prompting you to consider opportunities for optimization.

What’s Next?

Once you've explored how your nodes contribute to cost and performance, you can move to the Workload Drilldown page to see how individual namespaces and deployments are utilizing those resources.