How to manage cloud costs with the Infinity data source for Grafana and FOCUS

With the Infinity data source for Grafana and the FOCUS standard, you can analyze cloud costs across multiple providers and platforms.

Wilfried Roset is an engineering manager at OVHcloud where he leads an SRE team, and is also a Grafana Champion. He focuses on prioritizing sustainability, resilience, and industrialization to guarantee customer satisfaction.

Managing cloud billing data is a complex task. Each cloud provider uses its own format and terminology, making it difficult to compare costs across providers. Billing data is also often incomplete or inaccurate, which further limits visibility. Without a standardized format, organizations struggle to compare costs across providers, identify areas for optimization, or make informed decisions about resource allocation.

In this post, we’ll explore how to use the Infinity data source plugin for Grafana and the FOCUS standard to analyze your cloud costs no matter where you’re running your workloads — whether that’s AWS, GCP, or Azure.

Standardized billing data with FOCUS

FOCUS (FinOps Open Cost and Usage Specification) is an open standard for cloud billing data. Supported by the FinOps Foundation, FOCUS is designed to give organizations a consistent, vendor-neutral way to understand and analyze their cloud costs.

The benefits of using FOCUS, which has been adopted by AWS, Azure, and GCP, include:

• Simplified cloud cost management
• Improved transparency and visibility into cloud spend
• Enhanced ability to compare costs across different cloud providers and services

For example, suppose an organization uses both AWS and Azure for their cloud infrastructure. Without FOCUS, the billing data from each provider would come in different formats, making it difficult to compare and analyze costs. With FOCUS, the organization can standardize their billing data, making it easier to compare costs and identify areas for optimization.

Note: Grafana Labs also uses the FOCUS standard to give Grafana Cloud users access to standardized usage and cost data.

Visualizing cloud costs with the Infinity data source for Grafana

The Infinity data source is a powerful tool for visualizing data from various sources, including CSV files, in Grafana. It provides a flexible, scalable way to manage and analyze large datasets, making it ideal for cloud cost management and optimization. You can also use the Infinity data source and Grafana to combine multiple FOCUS files and create multi-cloud billing dashboards.

To demonstrate cloud cost management with the Infinity data source, we’ll set up a sandbox environment using Docker. Follow the steps in https://github.com/wilfriedroset/grafana-focus — this repository lets you reproduce the same results from this post.

These steps will start the Grafana and Python web server containers in detached mode. You can then visit http://localhost:3000, start exploring your data, and build your custom dashboards.

Creating custom dashboards with the Infinity data source

There are two main ways to work with your data using the Infinity data source for Grafana: the simple CSV option or the more powerful UQL parser.

The simplest way to get started with your data is to create a table showing cost usage data, as you would in a regular CSV editor:

Alternatively, you could use a bar chart panel to visualize the cost break down per a given dimension, such as region ID:

For more advanced use cases, such as cost flow analysis, consider the sankey panel plugin. Here, we show the cost per service category per sub-account name and region ID:

These visualizations can be combined to create custom dashboards that surface real insights into cloud spend.

Here’s an example from the infrastructure my organization uses: the screenshot below shows the cost breakdown of multiple Mimir OSS clusters over the last six months — clusters hosting nearly 300 million active series at 8 million samples per second. The data tells a clear story: object storage is significantly cheaper than cloud compute. If you’re looking to cut infrastructure costs, start by optimizing the read and write paths before targeting object storage. That single insight helped shape our overall cost optimization strategy.

Wrapping up

A few pitfalls to keep in mind: data processing happens in the browser, which can cause dashboards to feel laggy with large datasets. The UQL parser is also powerful but has a learning curve. Lastly, the sandbox environment we used here may not be suitable for large or complex organizations — further customization may be needed, such as switching to a dedicated FinOps solution or a general-purpose data platform.

That said, using Grafana, FOCUS, and the Infinity data source plugin to visualize cloud costs is a powerful way to gain insights into cloud spend and make data-driven decisions about resource allocation.