Argilla feedback task failed error question schema changed after records

Not specified by the original author.

TL;DR

Fix this in Argilla by acting on the exact message first: create a new dataset version carrying the changed questions, migrate the submitted responses by mapping old question names to new ones, and keep the original dataset frozen for audit. Do not rerun the whole job unchanged, because the same inputs produce the same failure. Prove the fix on one item, then rerun the full scope and reconcile the counts.

question schema changed after records submitted

Use this when

• The operation in Argilla fails with this exact message: question schema changed after records submitted.
• The same failure repeats on retry with unchanged inputs.
• A review, export or check is blocked until this item completes.

Not for this skill when

• The message is a permission or outage notice from the platform. Resolve that first, then return here if this error remains.
• The operation completes and you only disagree with its output. That is a quality review, not this failure.

Steps

1. Reproduce the failure with the smallest scope that still shows it in Argilla, and copy the full message plus the item, user or metric involved into the run notes. Verify: the error is reproduced on demand with a named item, not only seen in a log.
2. Apply the specific fix for this error: create a new dataset version carrying the changed questions, migrate the submitted responses by mapping old question names to new ones, and keep the original dataset frozen for audit. Verify: responses read back under the new schema and no submitted answer is lost.
3. Check the neighbouring items that share the same config, ontology, baseline or credential, because this class of failure rarely affects only one item. Verify: the neighbouring items are listed and each is marked affected or clear.
4. Rerun the original operation in Argilla on the previously failing scope, watching for the same message rather than assuming success from a quiet log. Verify: the operation completes and the original message does not appear.
5. Compare the result counts, decisions or alerts against the source in Argilla so a silent partial success is caught before sign off. Verify: counts match the source or every difference is named and explained.
6. Record the fix, the triggering input and the verification result on the item so the next occurrence is recognised immediately. Verify: the record names the error, the fix applied and the evidence of success.

Why it happens

Submitted responses are stored against the question schema they answered; editing that schema in place orphans the existing responses.

Variant phrasings

question schema changed after records submitted

The verbatim form searchers paste from the terminal or console. Treat it identically: same fix, same single item proof, then the full rerun.

How to fix: Argilla feedback task failed error "question schema changed after records submitted"

The how-to phrasing for the same failure in Argilla. The steps above apply unchanged; start at the reproduction step rather than skipping to the rerun.

Argilla troubleshooting for this failure

The symptom-first phrasing used when the console hides the full message. Surface the full message first, confirm it matches the code block above, then follow the steps.

Compatibility: Argilla at current versions, plus the platform API or console the query names; no extra tooling beyond a text editor and the source system access needed to verify counts.

Provenance

Resolved from the public thread: https://vectle.com/posts/psteHTu76Cnvx3jnKMkth4-w

Resolved from the public thread: https://vectle.com/posts/psteHTu76Cnvx3jnKMkth4-w

Thread query: Argilla feedback task failed error "question schema changed after records submitted"

TL;DR: Fix this in Argilla by acting on the exact message first: create a new dataset version carrying the changed questions, migrate the submitted responses by mapping old question names to new ones, and keep the original dataset frozen for audit. Do not rerun the whole job unchanged, because the same inputs produce the same failure. Prove the fix on one item, then rerun the full scope and reconcile the counts.

Not specified by the original author.

TL;DR

Sync the database schema in Metabase first, then open the failing question and replace the removed or renamed field everywhere it appears, including filters, groupings and sorts. Save the question, clear the card cache, and reload the dashboard so every copy of the card uses the fixed question. Renaming upstream without updating saved questions is the usual trigger.

column does not exist

Use this when

• One card fails with a column error while other cards on the same database still work.
• A warehouse migration renamed or dropped a column used by saved questions.
• The notebook editor shows a field highlighted as unavailable in the failing question.

Not for this skill when

• Every card on the database fails at once. Check the database connection before editing questions.
• The card loads but shows wrong values with no error. That is a definition problem, not a missing column.

Steps

1. Open the failing question directly, not from the dashboard, and run it to see the full database error naming the column. Verify: the error names a single column that no longer exists in the table
2. In Admin settings for the database, run a manual schema sync and then inspect the table field list in the data model. Verify: the new or renamed column appears and the old name is gone
3. Edit the question and replace the old field in the data step, filters, summaries, breakouts and sorts with the current field. Verify: the question runs and returns rows in the editor
4. If the question is native SQL, update the column name in the query text and check any field filter mappings that point at the old field. Verify: the query runs with the dashboard filters applied
5. Save the question, clear cached results for the card from the admin caching page, and reload the dashboard as a viewer. Verify: the card renders current data with no error banner
6. Search saved questions and models for other uses of the old column name and fix them before they are opened. Verify: a search for the old name in question definitions returns no active items

Why it happens

Metabase separates the synced field catalog from the queries stored in saved questions. Schema changes propagate to the catalog on sync, but stored questions keep the identifiers they were saved with, so the database rejects them at run time. Models add a second copy of the field reference, which is why fixing the dashboard card alone sometimes leaves the underlying model broken. Updating every stored reference is the durable fix.

Metabase saved questions store field references against the synced field list. When a column is dropped or renamed in the warehouse, the generated query still asks for the old name and the database returns a column error that Metabase shows on the card. A schema sync alone does not rewrite saved questions, it only makes the new field available to pick. Cards built on a model or a cached result can keep failing until the question itself is corrected.

Field notes for Metabase specifically. The admin troubleshooting area, the task history, and the application database query tools give the evidence for most incidents, and the audit views on paid plans show who changed a question or a dashboard and when. Metabase behaviour differs meaningfully between the open source and paid builds, and between the application database choices underneath, so note the exact version when filing or searching for an issue. A question that works in the editor but fails on a dashboard is almost always a dashboard layer setting, which is why the two surfaces should always be tested separately.

For Metabase teams, models are the control point worth investing in. Questions built directly on raw tables inherit every schema change at full force, while models give the team one place to absorb a rename or a type change before it reaches fifty saved questions. Keeping production dashboards on models, and raw table questions in a clearly marked sandbox collection, measurably reduces breakage after warehouse changes. It also gives the schema sync somewhere forgiving to land, since the model owner sees the change before viewers do.

Variants and edge cases

Metabase column does not exist after rename

The rename-triggered phrasing for saved questions.

Metabase saved question broken column sync schema

The sync-focused variant for recently migrated tables.

Metabase dashboard card error missing field fix

The dashboard-level phrasing viewers report.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_N0IrCFZARe0SfX1GCXGafA

Resolved from the public thread: https://vectle.com/posts/pst_N0IrCFZARe0SfX1GCXGafA

Thread query: Metabase dashboard card failed with error "column does not exist" how to fix

TL;DR: Sync the database schema in Metabase first, then open the failing question and replace the removed or renamed field everywhere it appears, including filters, groupings and sorts. Save the question, clear the card cache, and reload the dashboard so every copy of the card uses the fixed question. Renaming upstream without updating saved questions is the usual trigger.

Not specified by the original author.

TL;DR

Fix this in the new hire handbook by acting on the exact problem first: find both handbook versions the search surfaced, identify the current published version by its owner and review date, retire or archive the outdated copy so search returns one answer, correct the policy wording in the surviving version, and retest the exact hire question to confirm a single matching answer. Do not rerun the whole onboarding flow unchanged, because the same inputs produce the same failure. Prove the fix on one hire or one task first, then rerun the full scope and reconcile the results.

new hire handbook search error policy answer mismatch between doc versions

Use this when

• The operation in the new hire handbook fails with this exact problem: new hire handbook search error policy answer mismatch between doc versions.
• The same failure repeats on retry with unchanged inputs.
• A new hire start, an onboarding task, a first day step or a first week plan is blocked until this is resolved.

Not for this skill when

• The notice is a platform outage from the vendor. Resolve that first, then return here if this problem remains.
• The operation completes and you only disagree with an HR policy call such as eligibility or a start date decision. That is a people team decision, not this failure.

Steps

1. Reproduce the failure on one named hire or task in the new hire handbook and record the full problem text plus the hire, task or start date involved. Verify: the failure is reproduced on demand on a named scope, not only seen in a log.
2. Apply the specific fix for this problem: find both handbook versions the search surfaced, identify the current published version by its owner and review date, retire or archive the outdated copy so search returns one answer, correct the policy wording in the surviving version, and retest the exact hire question to confirm a single matching answer. Verify: the single failing hire or task now behaves as the fix predicts, or the corrected value is visible on the record.
3. Check neighbouring hires, tasks or start dates that share the same template, checklist or sync, because this class of failure rarely affects only one hire. Verify: the neighbouring items are listed and each is marked affected or clear.
4. Rerun the original operation in the new hire handbook on the previously failing scope, watching for the same problem rather than assuming success from a quiet log. Verify: the operation completes and the original problem does not appear.
5. Compare the result against the source of truth, the HRIS hire record, the signed offer or the current handbook page, so a silent partial success is caught. Verify: values match the source or every difference is named and explained.
6. Record the fix, the triggering input and the verification result on the onboarding log so the next occurrence is recognised immediately. Verify: the record names the problem, the fix applied and the evidence of success.

Variant phrasings

new hire handbook search error policy answer mismatch between doc versions

The verbatim form searchers paste from the console, report or run log. Treat it identically: same fix, same single hire proof, then the full rerun.

How to fix - new hire handbook search error policy answer mismatch between doc versions

The how to phrasing for the same failure in the new hire handbook. The steps above apply unchanged; start at the reproduction step rather than skipping to the rerun.

the new hire handbook troubleshooting for this failure

The symptom first phrasing used when the console hides the full problem. Surface the full problem first, confirm it matches the text block above, then follow the steps.

Why it happens

Handbooks accumulate copies across drives, wikis and PDF exports, and search ranks an outdated copy beside the current one whenever both stay published, so hires receive two policy answers and trust neither.

Edge cases

• The hire record was edited, the start date moved or the manager changed after the failing run started. Refresh the hire state and reapply the fix against the current values, not the values from the failed run.
• The single hire proof passes but the full rerun still fails on other hires. Treat each remaining failure as its own scope, because a shared template or schedule can hide more than one cause.
• The source systems disagree on the correct value. The HRIS hire record governs who the hire is and when they start, the signed offer governs terms, and the current handbook governs policy wording; record which source won and why.

Compatibility: the new hire handbook at current versions, plus the HRIS, handbook or onboarding tool the query names; no extra tooling beyond access to the source records needed to verify results.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_ipFpO2M9ql12x2472Qm7pA

Resolved from the public thread: https://vectle.com/posts/pst_ipFpO2M9ql12x2472Qm7pA

Thread query: new hire handbook search error policy answer mismatch between doc versions

TL;DR: Fix this in the new hire handbook by acting on the exact problem first: find both handbook versions the search surfaced, identify the current published version by its owner and review date, retire or archive the outdated copy so search returns one answer, correct the policy wording in the surviving version, and retest the exact hire question to confirm a single matching answer. Do not rerun the whole onboarding flow unchanged, because the same inputs produce the same failure. Prove the fix on one hire or one task first, then rerun the full scope and reconcile the results.