Explore Your Costs
Once your data sources are connected, KOSTRA gives you a normalized view of spend across every provider. This guide shows how to explore costs with Cost Explorer and the Cost Map so you can answer “where is the money going?” quickly.Open Cost Explorer
- In the left navigation, open Cost Explorer.
- Choose the time period you want to analyze.
- Review the total spend and the breakdown for the period.
Group and filter
The power of Cost Explorer is in slicing spend along the dimension you care about. You can break costs down and filter by:- Cloud / data source — compare spend across providers, or isolate a single Huawei Cloud account.
- Pool — see spend by the team, project, or business-unit structure you have modeled.
- Service — for example, compute, block storage, or object storage on any provider.
- Region — where spend is concentrated geographically.
- Resource type — the kind of resource driving cost.
- Owner — the employee associated with a resource.
Read the Cost Map
The Cost Map answers a different question: where in the world is my spend? It plots cost by region on a map, so you can see geographic concentration at a glance — useful for data-residency reviews and for understanding where your footprint is concentrated.- In the left navigation, open Cost Map.
- Review the regions where spend is concentrated; larger markers indicate higher cost.
- Select a region to see its detail.
From spend to action
Cost Explorer tells you what you spend; the rest of KOSTRA tells you what to do about it:- High spend on a service or region is often a cue to check Recommendations for optimization opportunities there.
- Spend you cannot yet attribute to an owner is a cue to refine your Pools and Allocation Rules.
- A workload that is expensive because of how it is architected is a cue to run a Modernization scan.
Tips for a useful cost review
- Start broad (all clouds, this month) and drill down, rather than starting from a single resource.
- Group by pool first if your goal is accountability; group by service first if your goal is optimization.
- Use period comparison to separate a real trend from a one-off spike.