Resolve an import problem
First distinguish a file that cannot be read, a blocked preview and an uncertain execution result. Only execution writes changes; a failed or cancelled preview does not undo an earlier import.
Keep the original file and note the organization, data context, matching key and options you used. These make it easier to understand the result.
Correct the file
| Symptom | Check and correction |
|---|---|
| File cannot be read | Read the reason shown. Save an old or mislabeled Excel file as XLSX; renaming its extension is insufficient. |
| Excel file is protected | Save an unprotected copy when permitted, or export the data as CSV. |
| No useful columns | Put headers on the first row. For XLSX, place the data in the first worksheet. |
| All values appear in one column | Check Delimiter for CSV. The column delimiter differs from delimiters inside list cells. |
| Accents are corrupted | Check Encoding; prefer UTF-8 when exporting the source file. |
| Warning about extra cells | Check quoting and the number of cells per row. Do not assume surplus values were imported. |
| Dates or numbers are rejected | Check the column’s Date format and supported formats. Map intentional empty markers to No value. |
| Choice or yes/no value is wrong | Open the options below the column’s destination and check its value mapping. |
| A property is missing | Confirm the chosen type. Hidden and calculated properties are not destinations for imported values. |
After correcting a file, select the corrected version and inspect the mapping again. Do not rely on an old preview.
Correct identity and mapping
A plausible-looking spreadsheet can still address the wrong records. Compare the preview's names and IDs with the records you intended to change.
| Symptom | Check and correction |
|---|---|
| A row is skipped | Check that its matching key is populated. Unknown IDs do not create new records. |
| Too many assets are created | Check name spelling, the chosen key and creation of missing related assets. |
| A rename becomes a creation | Match by the existing Boldo ID or stable unique key, then provide the new name. |
| Fewer records than file rows | Several rows may be merged into the same asset or relationship. Inspect duplicate identities. |
| A supposedly unused name affects matching | A recognized name column can provide a fallback in an asset import. Remove it if it must not participate. |
| Relationship instance is ambiguous | Supply its ID or appropriate distinguishing properties in relationship import. |
| Endpoint mismatch | The ID or key identifies a different connection. Correct the identity; existing endpoints cannot be moved by import. |
| A key or name conflicts | Find the existing holder and decide which record should retain the value before retrying. |
| A restoration conflicts on a unique key | The archived relationship kept a value that an active one now holds. Give it another value in the same row, or free the value first. |
| Relationships swap unique-key values | Each one takes the value another gives up, so no order can write them. Import in two passes: first give one of them a temporary value. |
An unknown unique key on a relationship can request a new instance. It is not equivalent to an unknown relationship ID. Use the relationship matching rules when investigating unexpected creations.
Explain unexpected changes
Empty property cells preserve existing values with Skip empty values. They can clear those values with Overwrite with empty. The owner column is an exception: empty or entirely unresolved owner lists preserve existing owners.
A value mapped to No value follows the same empty-value setting. Check that mapping when a marker unexpectedly preserves or clears a property.
Unknown owner emails produce warnings rather than blocking every other change. Use emails of members in the current organization. When some emails resolve, the recognized list replaces existing owners; inspect it before confirming.
Imported access domains apply to creations. They do not change the domains of existing assets. Use asset editing for that operation.
In asset imports, inspect Missing relationships when links disappear. Check the direction and target column before mapping relationship properties. For a precise update to a parallel connection, prefer the standalone relationship import with its ID.
In relationship imports, inspect Archived separately from the handling of existing relationships. Explicit values can request archiving or restoration even when property updates are skipped. A matched archive can also be restored when no archive value is supplied.
Resume after an interruption
- Check whether you were preparing a preview or executing a confirmed import.
- If execution started, inspect the relevant active and archived inventories before submitting anything again.
- Compare a few affected records and the available result counters with the file.
- Correct the file or options for the changes still needed.
- Run Verify again against the current data.
- Confirm only after reading that new preview, then check the final result.
An asset import can save asset changes before a later relationship error. More generally, an error after saving or a lost connection is not proof of zero changes. Retrying the original file blindly can repeat or extend changes.
There is no general “undo the completed import” action on this page. Restore prior property values from your retained data through a reviewed correction. Use archive restoration for items that were archived rather than permanently removed.
If Imports is unavailable, verify your organization account type: Administrator or Owner is required. Read any plan or capacity message shown. Reloading the page does not grant missing access.
For a persistent problem, contact support with the relevant error and minimal example. Remove unrelated or confidential data from the example.