<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=521127644762074&amp;ev=PageView&amp;noscript=1">

Kubernetes Resource Usage: Estimate Workload Cost with Goldilocks Open Source

If you are looking for help on how to set Kubernetes resource limits and requests, you’ve come to the right place. Goldilocks is an open source tool that helps users optimize their resources by setting proper CPU/Memory. This helps engineers avoid a lot of trial/error guesswork. The changes could help you understand if you are under or over provisioning workloads and set budgets accordingly. 

Watch a video on Kubernetes Cost Management with Goldilocks

New Report Available: Estimate Workload Costs

Goldilocks open source users can now upgrade to the latest version (v4.5.0) to access new functionality that helps to estimate the cost of the workload and the cost impact of applying our recommendations.

Instead of having to look through a cloud bill to find out rates, punch in CPU hours or memory and calculate spend, Goldilocks does all the time consuming work for you. Users save time and can more accurately configure Kubernetes workloads to impact cloud consumption. 

How it Works

After updating Goldilocks, users will be presented with an option to enter an email address to unlock the cost estimates. 

Image: Screenshot of the Goldilocks UI with sign up box at the top. 

After entering your email address, you will receive an API token via email that will be added to the dashboard. Note: Anyone working on a different computer will need to enter their email address, even if they're looking at the same dashboard. 

Once the API key is entered, the prompt will be replaced by a Cost Settings dialog. 

Screenshot of Goldilocks UI with cost setting dropdown menu options

The user can choose an AWS or GCP instance type, or set their costs manually. When the user scrolls down, they'll see each workload annotated with cost data. Goldilocks automatically calculates how much a gigabyte of RAM and CPU costs per hour. 

This can be correlated back to each workload and get recommendations to increase or decrease CPU or memory.

Image: Goldilocks UI screenshot of the cost to run Goldilocks and recommendations. 

In this example above, you can see that the cost to run this demo application is $.0747/hour. Goldilocks actually suggests changes to CPU and memory that will help to reduce cost by $.0729/hour. These numbers may seem small, but they can add up across many pods and over long periods of time. As users run larger workloads in Kubernetes, the recommendations become more significant. 

Goldilocks is a great option for teams running a small number of Kubernetes clusters. When more indepth Kubernetes cost visibility and optimization is required, users should look at Fairwinds Insights, our Kubernetes governance platform.

You can use Fairwinds Insights for free, forever. Get it here.

Insights provides one centralized view into Kubernetes cost across multiple clusters providing consistency and alignment across teams, and the ability to eliminate cloud cost waste.

Resources: 

See how Fairwinds Insights reduces your Kubernetes risk!