FinOps for Kubernetes: Cost Visibility and Savings
Learn FinOps for Kubernetes: see costs by team with OpenCost, rightsize workloads, use Spot safely with Karpenter, and build showback teams trust.
What You'll Learn
Understanding Why Kubernetes Costs Are Hard to See
This is the last module in the acme-shop journey.
Understanding Unit Economics and Showback
A total bill is a poor measure on its own. Context turns it into something a team can act on.
Allocating Shared Costs Fairly
Some costs belong to no single team. Forcing exact attribution creates a false sense of precision.
Labelling Workloads for Cost Allocation
Cost data is only as good as your labels. A label on 80 percent of workloads leaves you chasing the other 20 percent every month.
Seeing Costs with OpenCost
OpenCost is the free, open source CNCF project that calculates Kubernetes allocation costs.
Rightsizing with VPA and Goldilocks
Most engineers set generous requests "to be safe" and never revisit them. Rightsizing brings requests closer to real usage.
Skills You'll Master
Curriculum Index10 topics
Understanding Why Kubernetes Costs Are Hard to See
This is the last module in the acme-shop journey.
Understanding Unit Economics and Showback
A total bill is a poor measure on its own. Context turns it into something a team can act on.
Allocating Shared Costs Fairly
Some costs belong to no single team. Forcing exact attribution creates a false sense of precision.
Labelling Workloads for Cost Allocation
Cost data is only as good as your labels.
Seeing Costs with OpenCost
OpenCost is the free, open source CNCF project that calculates Kubernetes allocation costs.
Rightsizing with VPA and Goldilocks
Most engineers set generous requests "to be safe" and never revisit them.
Using Spot and Karpenter Safely
Spot capacity is spare cloud capacity sold at a discount.
Debugging Unexpected Cost
When the numbers rise, work from the biggest picture to the smallest.
Hands-On Lab: Showback on platform-lab
📌 Remember: This lab is free. Everything runs locally on the shared platform-lab kind cluster.
Using the Quick Reference and Avoiding Common Mistakes
Quick Reference Common Mistakes Optimising before you can see.
Career Impact
Roles that use the skills in this module.
DevOps Engineer
Site Reliability Engineer
Platform Engineer
Next Modules
Related Guides
Practice on the Coding Sheet
Not a software engineer sheet. Every problem comes from real DevOps, SRE, Platform and Cloud interviews, from your first script to a system you build yourself.
Open the Coding SheetFrequently Asked Questions
FinOps is a way for engineering, finance, and product teams to work together on cloud spending. It starts with seeing who spends what, then reducing waste, then keeping the habit going. It is about making cost a normal engineering input, not about cutting costs blindly.
Many teams share the same nodes, so the cloud bill shows machines, not applications. You need to connect pods, labels, and resource requests to the bill to learn what each team costs. Tools such as OpenCost do that calculation for you.
OpenCost is the free, open source CNCF project and a good default for learning and for many teams. Kubecost is a commercial product from IBM built on the same allocation ideas, with extra features and plan limits. Check each project's current documentation before you choose.
No. Spot capacity can be taken back with short notice, so it suits stateless services with several replicas and jobs that can restart. Single-replica services and databases should stay on On-Demand capacity.