How to aggregate costs using UsageTypeGroups

[Haegi]

Hi all,

I am currently trying to retrieve cost and usage data for my aws account.
I am able to retrieve the data on a very granular level using the parameters below.
I found out that it is possible to group it on a higher level using UsageTypeGroups but unfortunately it is not documented on how to use them. Adding a filter "And": { "Key": "USAGE_TYPE_GROUP", "Values": [ "EC2: Running Hours" ] } seems to not work and as far as I see only one dimension is not allowed.

{
    "TimePeriod": {
      "Start": "2022-05-14",
      "End": "2022-05-15"
    },
    "Granularity": "DAILY",
    "Filter": {
      "Dimensions": {
        "Key": "RECORD_TYPE",
        "Values": [
            "Usage"
        ]
      }
    },
    "Metrics": [
        "UnblendedCost",
        "UsageQuantity"
    ],
    "GroupBy": [
      {
        "Key": "SERVICE",
        "Type": "DIMENSION"
      },
      {
        "Key": "USAGE_TYPE",
        "Type": "DIMENSION"
      }
    ]
}

NOTE: I am calling it using aws ce get-cost-and-usage --cli-input-json file:///tmp/ce.json

[tim-finnigan]

Hi @Haegi is this the filter you're looking for?

 "Filter": {
      "Dimensions": {
        "Key": "USAGE_TYPE_GROUP",
        "Values": [
            "EC2: Running Hours"
        ]
      }
    },

[Haegi]

Hi @tim-finnigan it seems to work, thanks!
But is it possible to either use a wildcard or group by USAGE_TYPE_GROUP for being able to see the costs and usage for all USAGE_TYPE_GROUPS?
I know that it is possible to aggregate on service level but then the UsageQuantity is useless.

[tim-finnigan]

Glad that works! I'm not sure about wildcard/grouping for USAGE_TYPE_GROUP as that is already a "grouping of common usage types." That being said, it might be worth asking on https://repost.aws/ as someone with more insight into CostExplorer may have a suggestion.