Troubleshooting
Troubleshooting: my schema import is stuck or failing
This page covers problems while importing a schema. For how imports work, see What schema import does.
I can’t find the import after starting it¶
Likely cause: Import details live under Organization Settings, which only organization admins can open.
What to do:
- If you’re an admin, go to Organization Settings ›
Schema Importsand open the import. - If you’re not, watch the schema’s tabs: objects appear as each stage finishes. Ask an admin to check any failures.
It’s been “Analyzing file structure…” for a long time¶
Likely cause: A large or complex script. Splitting takes longer when AI is involved.
What to do:
- Wait a few minutes. The page refreshes itself.
- If nothing changes after about 10 minutes, click
Refresh. The import keeps running on the server even if the page stopped updating. - If it’s still stuck, cancel and import the script in smaller parts.
A table fails¶
Likely cause: An unrecognized data type, a vendor-specific clause, or a reserved word used as a column name.
What to do:
- Click
Fixon the item and readWhat went wrong. - Edit the DDL. For example, swap
BIGLONGforBIGINT, remove a partition clause, or quote the reserved word. - Click
Retry with Changes.
If that doesn’t work: Leave it failed, finish the rest, and create the table by hand with Create Table › Paste Code using corrected DDL.
A table shows “Skipped”¶
Likely cause: A table with that name already exists in the schema. Script imports never overwrite existing tables.
What to do: Nothing, if you meant to keep your version. To take the script’s version, delete or rename the existing table and retry the item.
Indexes fail¶
Likely cause: The index refers to a table or column that didn’t import.
What to do:
- Fix the table first.
- Click
Resume All Failed, or acceptRetry All Dependentswhen the fixed table succeeds. - If the index refers to something that will never exist, edit its DDL to drop the reference.
Relationships fail¶
Likely cause: The foreign key refers to a table or column that didn’t import, or the column types don’t match.
What to do:
- Check both tables exist and the join columns have matching types.
- Remove vendor-specific options (such as deferrable constraints) from the DDL.
- Retry or resume.
A view, procedure or function fails¶
Likely cause: SQL that Kenseme can’t parse, often a vendor-specific extension, or a reference to an object that failed.
What to do:
- Check whether a dependency failed.
Fixis disabled until it’s fixed. - Click
Fix, simplify the SQL, and retry. - If it still fails, create the object by hand and paste the SQL on its
SQLtab.
Most objects have no description¶
Likely cause: The script had no description statements (sp_addextendedproperty, COMMENT ON). A script import doesn’t write descriptions of its own.
What to do: Select the objects on their tab and click Generate Descriptions, or use Import Descriptions on the Info tab if you have descriptions in a file.
The import shows “Failed”¶
Likely cause: The import stopped before finishing. The banner shows why.
What to do:
- Click
Refresh. It may have completed. - Read the message, fix the cause, and click
Resume Importor start a new import. - If the message points to a server problem, contact your administrator with the schema name and the time you started.
“An import is already in progress for this schema”¶
Likely cause: Another script import is still running against the schema.
What to do: Wait for it to finish, or cancel it from Organization Settings › Schema Imports.
A connection import can’t read the source¶
Likely cause: The connection’s credentials, network access or permissions.
What to do:
- Read the error in the dialog and click
Retry. - Ask an admin to test the connection. See Database connections overview.
- If the review page shows extraction warnings, some catalog categories couldn’t be read. The change list may be incomplete; check the connection’s permissions.