Workload Drilldown

Overview

The Workloads tab gives you a focused lens into how individual workloads are consuming resources and contributing to your Kubernetes costs. From spotting inefficient deployments to uncovering potential savings, this view helps you take targeted action where it matters most.

Navigating the View

When you land on the Workloads, you're looking at cost and efficiency data with filtering and time-range controls available at the top. You have:

• Group by Namespace To group your workloads by namespace to align insights with how your teams or applications are structured.

• Filters To refine your view using filters like workload type, efficiency levels, or potential savings to zero in on specific problem areas.

• Date Range To set a custom date range to analyze trends over time. This affects all usage metrics and cost graphs in the view.

Key Metrics

At a glance, the top-level metrics give you quick insight into your environment:

• Namespaces Shows how many namespaces are currently running workloads during your selected period.

• Potential Savings Tells you how much cost you could potentially save by right-sizing your workloads based on current usage patterns.

• Workload Efficiency Provides an overall percentage representing how well your workloads are utilizing their allocated resources cost.

Exploring the Workload List

Next up, you'll find a list of all workloads within the scope of your filters. Each row is a workload that you can explore further.

Here’s what each column tells you:

Column	Description
Namespace	The Kubernetes namespace where the workload is running. Click the name to open a namespace-level breakdown.
Workload Name	The name of the workload
Workload Type	Indicates if it’s a Deployment, StatefulSet, DaemonSet, etc.
Efficiency	Calculated based on CPU and memory utilization vs. requested values.
Total Cost	Total spend associated with the workload during the selected timeframe.
Potential Savings	Estimated savings from optimizing CPU or memory allocations.
CPU Utilization (p99)	Peak CPU usage observed 99% of the time (in millicores).
Memory Utilization (p99)	Peak memory usage observed 99% of the time (in MB).

Use this table to spot inefficient workloads, high spenders, or those with savings opportunities. When a workload catches your eye, click its row to dive into the details.

Diving Into a Specific Workload

Once you click into a workload name, you’ll see everything you need to understand about its resource behavior and cost impact.

You’ll start with key identifiers:

• Workload Name

• Workload Type

• Namespace

• Labels (Automatically pulled from your Kubernetes metadata.)

  

Overview Tab

This section summarizes how the workload is built and how it's performing.

• Workload Specification See the number of replicas currently deployed for this workload.

• Resource Utilization (Last 30 Days)

  CPU and memory usage are visualized over time, compared against the requested and limit values. You have the option to select from the following statistical views.

• Based on the choice, the graphs for both CPU (in millicores) and memory (in MB) will be updated, helping you understand how resource allocation compares to actual usage.

Statistical Views Available

Option	Description
p99	99th percentile – the usage value below which 99% of data points fall. Useful for understanding consistent high usage without being affected by rare spikes.
p95	95th percentile – balances typical usage and peaks, ideal for sizing buffers.
Max	Shows the absolute highest value observed.
Min	Displays the lowest recorded value, highlighting underuse.
Average	Provides the overall mean usage over time.

• Daily Cost You can track how a workload’s cost changes over time using adjustable date ranges. Compare recent trends with historical patterns, identify any cost spikes, and evaluate whether optimizations have had an impact.

  The data is available both as a visual trend graph and in a detailed breakdown table.

Container-Level Insights

Every workload is made up of containers — and here’s where you get to see their individual impact.

For each container tab, you’ll find:

• Resource Utilization (Last 30 Days)

  • CPU usage in millicores

  • Memory usage in MB

• Daily Cost

  • Grouped by resource type option

  • Visualized with total and average cost graphs

  • Backed by a detailed table for container-by-container breakdown

With this view, you can identify which containers are over-provisioned or which are driving cost spikes — giving you clear direction for optimization.