Bulk data exports: record groups and collections appear empty

[tacman]

Issue

The AWS Registry of Open Data documentation states that users can access portions of the dataset:

To pull descriptions for specific record groups or collections, the following AWS CLI commands can be used.

And provides examples like:

aws s3 sync s3://nara-national-archives-catalog/descriptions/record-groups/rg_011/ [destination] --no-sign-request

However, when I try to list or sync individual record groups, they appear to be empty:

$ aws s3 ls s3://nara-national-archives-catalog/descriptions/record-groups/ --no-sign-request
                           PRE rg_1/
                           PRE rg_10/
                           ...

$ aws s3 ls s3://nara-national-archives-catalog/descriptions/record-groups/rg_011/ --no-sign-request
# Returns nothing

The record group directories appear to exist but contain no files. Only the full zip exports in /zip/ contain actual data:

$ aws s3 ls s3://nara-national-archives-catalog/zip/ --no-sign-request
2025-04-11 08:38:07   44179771 nac_export_authorities_2025-04-08.zip    (43 MB)
2025-04-11 08:38:29 93018790202 nac_export_descriptions_2025-04-08.zip (87 GB)

Questions

1. Are the record groups and collections supposed to contain data? If so, what
2. Is there a plan to populate them?
3. Is there a way to get smaller subsets of descriptions (e.g., by record group)?

Impact

Without smaller subsets, users must download the entire 87 GB descriptions file to access any specific record group data. This is problematic for:

• Users with limited bandwidth
• Users who only need specific record groups
• Development and testing workflows

---

Tested with: AWS CLI, no-sign-request (public access)