We’re excited to announce the release of Kubecost 2.8, which delivers major usability and performance improvements. This version simplifies how teams manage custom pricing across environments and introduces a new backend architecture—powered by ClickHouse—that significantly boosts performance and reduces compute overhead.

Revamped Custom CSV Pricing: Centralized, Transparent, and UI-Validated

Kubecost 2.8 makes it dramatically easier to apply custom pricing models across your infrastructure. In previous versions, using custom pricing required maintaining a separate CSV file per cluster—a clunky and error-prone setup for teams managing multiple environments.

With this release, you can now define a single centralized pricing sheet that applies across all your clusters.

This enhancement is especially valuable for on-prem and hybrid environments. While Kubecost automatically pulls accurate cloud billing data—including negotiated discounts—from your provider’s APIs, on-prem clusters like OpenShift don’t have a cloud bill. With a custom CSV file, you can define internal cost models for things like hardware, energy, licensing, and support.

Now, with one CSV file, you can:

• Standardize pricing across your entire infrastructure
• Define internal costs for self-managed environments
• Apply pricing retroactively
• Compare on-prem and cloud workloads side-by-side in the Kubecost UI

This means teams get a unified view of Kubernetes costs across AWS, Azure, GCP, and private datacenters—all in one place.

Built-In Error Validation

Custom pricing is still configured via your Helm values file, but Kubecost now provides real-time validation of your pricing sheet in the UI. If your CSV contains formatting issues or invalid keys, Kubecost will surface those errors directly on the Settings page. This makes it faster and easier to catch misconfigurations before they impact your cost data.

ClickHouse Beta: A New Backend for Scale and Speed

Kubecost 2.8 also includes the beta release of our new ClickHouse backend, a foundational upgrade for teams running at scale.

Kubecost has historically used DuckDB to power its cost allocation engine. While reliable for small to mid-sized clusters, DuckDB can become a bottleneck at large scale or with high query volumes.

With the move to ClickHouse, we’re seeing:

• Massive performance improvements on large datasets and long-range queries
• Significant reductions in CPU and memory usage, lowering the cost to run Kubecost
• Faster UI responsiveness and snappier dashboards
• A more scalable architecture for long-term data retention and future features

The ClickHouse backend is available to all users in beta and runs alongside the current engine, making it easy to evaluate the upgrade without disrupting your existing workflow.

Upgrade to 2.8 Today

To get started with the latest release, upgrade Kubecost via Helm:

helm upgrade kubecost cost-analyzer \
--repo https://kubecost.github.io/cost-analyzer/ \
--namespace kubecost

Read the full release notes

We’re especially interested in feedback on the ClickHouse beta—how it performs, what it unlocks for your team, and how we can make it even better. Drop us a note in Slack and share your experience!

Kubecost 2.8 is a major step forward—whether you’re managing cloud-native infrastructure, operating in hybrid environments, or scaling to thousands of nodes. We’re building a cost monitoring platform that’s accurate, powerful, and ready for wherever your infrastructure runs.

Stay tuned—more improvements are on the way!