- Newest
- Most votes
- Most comments
No, cost allocation tags do not work as expected when Split Cost Allocation Data (SCAD) is enabled for Amazon ECS and EKS. SCAD breaks down costs into finer granularity, but this new data format isn't directly compatible with cost allocation tags, making it difficult to aggregate costs using tags.
Reasoning:
1/SCAD creates new cost lines: SCAD generates new cost lines in the Cost and Usage Report (CUR) that are based on granular usage data (e.g., per-pod in EKS). These new cost lines don't have the same cost information (e.g., unblended costs) as the original cost lines.
2/ Tagging limitations: Cost allocation tags rely on the cost information present in the CUR. Since SCAD changes the cost structure, the existing cost allocation tags may not be applicable to these new cost lines.
3/ Impact on cost management tools: Services and features that rely on cost allocation tags, like AWS Cost Explorer and custom billing views, may not be able to correctly identify and aggregate costs when SCAD is enabled. While SCAD provides detailed cost data for container workloads, it introduces complexities when trying to analyse costs using traditional cost allocation methods.
Possible workarounds:
1/ Cost Categories: Cost Categories can be used as an alternative to cost allocation tags for grouping costs based on user-defined rules, but they need to be configured separately from cost allocation tags.
2/ CUR queries: You can write specific CUR queries to aggregate costs based on SCAD data, but it requires a deeper understanding of the SCAD data structure.
3/ AWS Support: AWS recommends consulting their documentation and support channels for guidance on using SCAD with cost allocation tags
Pls. refer - https://docs.aws.amazon.com/cur/latest/userguide/split-cost-allocation-data.html
answered a year ago
Thanks for this AI generated response.
#1 Unfortunately Cost Categories does not work either (for the same reasons that the tags do not work).
#2 CUR queries : As described in the other thread below - there does not seem to be a good way to query CUR either https://repost.aws/questions/QUoIO4XMgxTQugekFJnMDKGQ/cur-query-for-using-split-cost-allocation-data
This is an analysis by Google Gemini after reviewing: https://medium.com/@sree-chalasani/why-cost-allocation-tags-are-broken-if-split-cost-allocation-data-scad-is-enabled-57c22fba48d7
The "Double Fracture" in AWS Billing Data AWS Split Cost Allocation Data (SCAD) introduces a fundamental schema schism that breaks standard FinOps reporting. By creating a distinct data namespace for containers, AWS forces a disconnect on two critical dimensions:
The Tagging Fracture (The "Where"): AWS effectively creates two parallel tagging universes. A tag like Project: Alpha lands in the standard resource_tags columns for standard resources (S3, RDS) but shifts to entirely new split_line_item columns for pods. You cannot group by a single "Project" column; you must coalesce disparate columns to stitch the view back together.
The Metric Fracture (The "What"): Even worse, the cost data itself is fractured. Standard resources report value in universal columns like line_item_unblended_cost, while SCAD pods report value in unique columns like split_line_item_split_cost.
The Result: A standard SUM(cost) query returns $0 for your containers. To calculate the Total Cost of Ownership (TCO) for a single application, FinOps teams must write complex, brittle ETL logic to normalize these incompatible data sets—effectively repairing a broken schema that should have been unified by design.
answered 9 months ago
Hello , The Split Cost Allocation Data is only available in the Cost and Usage Reports and not in Cost Explorer. Here is the documentation on Split Cost Allocation: https://aws.amazon.com/blogs/aws-cloud-financial-management/improve-cost-visibility-of-amazon-eks-with-aws-split-cost-allocation-data/ Here is the documentation on Cost and Usage Reporting: https://docs.aws.amazon.com/cur/latest/userguide/what-is-cur.html?nc2=type_a
Hmm - This article does not seem to say anything about these tags in cost management products. If they are not usable in these other CE/Budgets / etc. - why do they appear under Cost Allocation Tags in these products ? Why are these called as Cost Allocation Tags if they dont behave like tags in any product ? Confusing! Is the only reason these are called tags, the fact that they appear under tags header in CUR ?
Also, user-defined tags do not work as expected. try it out. (1) create a ecs cluster with say 3 ec2 instances (2) deploy 2 tasks tagged application:ms1 and application:ms2 in cluster (3) task1 also uses a dynamoDB table that is also tagged application:ms1; and task2 uses another s3 bucket tagged application:ms2. Set replication to 5 for each task . (4) Enable scad. If I try to report the cost of task1 using tag application:ms1, I expect that it should add the dynamoDB cost and container runtime costs . Similarly for task2, tag application:ms2, should add the s3 bucket and container runtime costs. This does not seem to work. However , in step (1) if using fargate, and not ec2 auto scaling group, then everything works fine without scad.
Relevant content
asked 2 years ago
asked 2 years ago
- AWS OFFICIALUpdated 3 years ago
- AWS OFFICIALUpdated 3 years ago
Also the recent launch of k8s labels treated as tags : https://aws.amazon.com/about-aws/whats-new/2025/10/split-cost-allocation-data-amazon-eks-kubernetes-labels/
why are these tags not showing up in Cost explorer?