Tableau dashboard extract refresh failed with error "data source not found" fix
Fixes a Tableau extract refresh that fails because the data source cannot be found by reconnecting the workbook to the published source or the moved file. Use when a scheduled refresh fails after a file or source move. Not for a live connection query error inside a view.
TL;DR
Republish or relink the missing source first, because the refresh is looking for a name or path that no longer resolves. Open the workbook, point each connection at the current published data source or the new file location, save, and run the refresh again before touching schedules. Once the extract builds, confirm the schedule is attached to the refreshed source.
data source not foundUse this when
- A scheduled extract refresh fails and the job history names a source that was renamed or moved.
- The workbook opens in Desktop but the server refresh cannot find the file or published source.
- A data source was republished under a new name and downstream extracts broke the same night.
Not for this skill when
- Views render but show stale numbers with no refresh error. Check the schedule and cache instead.
- The refresh fails with a database login error. Fix credentials before relinking anything.
Steps
- Open the refresh history for the workbook and copy the exact source name the failed job tried to resolve. Verify: the named source matches a rename, delete, or move you can confirm in the content list
- In Tableau Desktop, open the workbook, use the data pane to replace the reference with the current published data source, and check every sheet for broken field pills. Verify: no sheet shows a red pill or a missing field warning
- For a file based source, place the file back at the path the server account can read, or repoint the connection to the new shared location and embed access correctly. Verify: the server run as account can open the file from that path
- Publish the repaired workbook and choose the option that updates the extract connection, then run an on-demand refresh from the server. Verify: the refresh completes and the row count changes as expected
- Reattach the original refresh schedule to the repaired source and confirm the next run time is in the future. Verify: the schedule list shows the workbook with the correct cadence
- Open two dashboards that use the extract and compare a known total against the source system. Verify: totals match and the last refreshed timestamp is current
Why it happens
Extract refresh resolves the connection at job start. A rename, a republish that created a second source, a moved spreadsheet, or a permissions change on the folder all present as the same not found error because the stored pointer no longer resolves. Desktop can mask the problem because the analyst machine still has the file locally, while the server account does not. Relinking to the canonical published source prevents the same break on the next move.
Tableau extracts store a connection pointer, not a copy of the source definition. When a published data source is renamed, deleted, or a file based source is moved on the server, the extract job starts, cannot resolve the pointer, and reports the source as not found even though the data still exists. The background job log in Tableau Server or the refresh history in Tableau Cloud names the exact source it tried to open.
Field notes for Tableau specifically. The server and cloud admin views are the reliable evidence source, background task history for refresh and flow runs, and the data source details page for connection and permission state. Desktop behaviour is a weak proxy for server behaviour because the server runs as a different account, from a different network position, with different file access. Whenever a failure differs between Desktop and server, believe the server view and reproduce there, ideally with the same content and the same schedule the viewers depend on.
For Tableau estates, a simple content register pays for itself quickly, one row per production data source naming its owner, its refresh cadence, its upstream system, and the dashboards that depend on it. Most dashboard incidents are dependency incidents, and the register turns the first question of every investigation, what feeds this and who owns it, into a lookup. Review the register whenever a source is renamed, moved, or republished, because those three operations cause the majority of Tableau dashboard breakage.
Variants and edge cases
Tableau extract refresh data source not found after rename
The rename-specific phrasing for the most common trigger.
Tableau Server refresh failed missing data source relink
The server-side variant focused on relinking rather than rebuilding.
Tableau Cloud scheduled refresh cannot find published source
The cloud variant of the same pointer failure.
Provenance
Resolved from the public thread: https://vectle.com/posts/pst_wpYuUs7od9CZdnF0VRrqMQ