The reality of our users is applications are made of:

application = X OpenShift projects/pods + Y virtual machines + Z cloud services + T external costs

When X is projects, we can give them the cost information. When X is pods, COST-3874 (to be implemented) should solve the requirement.

When Y is virtual machines running on OpenShift (i. e. KubeVirt), they look like projects in Cost Management and all is fine. Otherwise, COST-3042 (to be implemented) should solve requirement.

Problem with Z is we do not itemize the cloud costs: as of today, Cost Management has AWS, Azure, Google Cloud and Oracle Cloud pages but we just give subtotals: "you have spent this much on AWS EC2 in April", "you have spent that much on AWS RDS in May" and that's it. That's not actionable for users, since they need to know "AWS RDS instance F is part of my business application BA", which we are not providing. This can be mitigated with proper tagging but that still makes difficult to know what part of the cost is being attributed and what not.

This feature is about itemizing the cloud costs:

• AWS
  • AWS EC2
    • AWS EC2 instance foobar3
    • AWS EC2 instance baztar7
    • ....
  • AWS RDS
    • AWS RDS instance 1boingboing
    • AWS RDS instace fsck34
    • ...
  • AWS S3
  • AWS Route 53
  • AWS DynamoDB
  • ...
• Azure
• GCP
• OCI

We should also provide a "deep dive" view, where users start from an aggregate (like we have today) and can dive deeper using multiple filters or group-by's (not limited to 1 or 2 group by). E. g.
1. Aggregated AWS cost
2. Aggregated AWS cost per account
3. Aggregated AWS cost per service per account: EC2, RDS, Route 53, etc
4. Aggregated AWS EC2 cost per instance type for one account
5. ...
6. Itemized cost per instance name for AWS EC2 for that account

Document for feature discussion