Video | Tableau | Tool strategy | Data engineering

Stale content admin view improvements & the Metadata API: Tableau Server & Tableau Online

Tagging stale content and querying it through the Metadata API is the combination every admin should be using to stop content getting out of control.

Part ofTableau Server
Watch on YouTube
  • The stale content report lives under the site status icon, and you can manually set the stale threshold (default 45 days) to suit your instance via the hamburger menu filters.
  • From 2020.3 you can multi-select workbooks and data sources in the stale content report and tag them, then find and act on them via the Tag filter on the Explore page's workbooks and data sources views.
  • Stale content matters because it consumes server resources through unnecessary extract refreshes and takes up storage space.
  • The Metadata API uses a GraphQL interface (under External Assets) where you incrementally build queries, adding fields like tag name, workbooks, published data sources, owners and downstream workbooks to enrich the results.
  • You can filter the Metadata API query by tag name using filter notation so that only your stale content tag is returned rather than every tag on the server.

Tagging stale content on Tableau Server/Online lets admins act on unused workbooks and data sources instead of just viewing them, and the Metadata API lets you query that tagged content with GraphQL to find owners, projects and dependencies. Together they stop unused content quietly eating server resources and storage.

Aimed at Tableau Server/Online admins dealing with old, unused workbooks and data sources clogging up the site. Uses the built-in stale content report and, for Data Management add-on users, the Metadata API's GraphQL interface.

The Breakdown
  1. Open the stale content report 0:34

    Go to the site status icon at the bottom left to reach Tableau's pre-built admin dashboards, then select the stale content report, which queries the server's database in real time to show unused content.

  2. Set your own stale threshold 1:25

    Use the hamburger menu filters to change the default 45-day stale threshold to whatever timeframe actually suits your instance and usage patterns.

  3. Understand why it matters 2:01

    Stale content isn't just clutter — it keeps triggering unnecessary extract refreshes and consumes storage, so isolating it lets you reach out to owners and archive what's not needed.

  4. Multi-select and tag stale items 2:32

    From 2020.3 you can select multiple workbooks and data sources directly in the stale content report and tag them in one action — a capability that didn't exist before.

  5. Find tagged content via Explore 3:33

    Tagging alone changes nothing visible on the report page; go to the Explore page's workbooks or data sources views, open the filter panel and filter by the Tag field to surface everything you tagged.

  6. Query stale content with the Metadata API 4:53

    For richer, actionable detail than the Explore interface offers, Data Management add-on users can open External Assets and use the built-in GraphQL query tool to interrogate the server's metadata directly.

  7. Build a GraphQL query incrementally 6:09

    Start from a blank query and build it up field by field: query 'tags' for the tag ID, add 'name' to see the tag label, then nest 'workbooks' and 'published data sources' to pull back the tagged assets themselves. This must be built in this layered order since each new field depends on the object it's nested inside.

  8. Enrich with owners and downstream assets, then filter by tag 8:43

    Add fields like owner and downstream workbooks to see who's responsible and what depends on a data source, but don't over-load one query since each added field slows it down. If you have multiple tags on the server, add a filter block specifying the tag name so the query returns only your stale content tag rather than everything.

Worth Knowing
  • The GraphQL interface in External Assets is meant for exploring what's queryable, not as a day-to-day tool — the real value comes from taking the query further via the REST API.
  • Querying the Metadata API requires the Data Management add-on, so it's not available to every admin.
  • Loading too many nested fields in one query noticeably slows down the response, so only query the specific information you need.
  • Tagging content changes nothing visible on the stale content report itself — you have to go to the Explore page to see and act on tagged items.
Use It When

Reach for this when you're an admin trying to claw back server resources and storage from workbooks and data sources nobody's using, and you want a repeatable way to identify owners and dependencies before archiving anything.

How this Rollup was made provenance & method

A Rollup is drafted by AI from the video's transcript, then reviewed and edited by Tim. Everything used to produce this one is listed below — the model, the exact prompt, and the source video — so the process is transparent and reproducible.

Transcription
On-device — NVIDIA Parakeet v3 for recent videos, OpenAI Whisper large-v3 for earlier ones. The transcript never leaves the machine or gets published.
Drafting
Claude Sonnet 5 in the cloud, from that transcript.
Prompt
The exact Rollup prompt (v2) — the full system prompt, unedited.
Source video
Watch on YouTube
Drafted
5 July 2026 at 09:38
Reviewed & edited
5 July 2026 at 09:41 · by Tim Ngwena

Model + prompt + video is everything you'd need to recreate a Rollup like this yourself. The one thing we don't share is the transcript.

Rights. The video and its transcript are the property of TN Media Ltd. Unauthorised use or download is prohibited. © TN Media Ltd.