All import set rows are being skipped

[bookman1502]

I'm doing these exercises to gain proficiency in ServiceNow. In section 6 for import sets, I'm asked to export data from the 'Windows Servers' list (the cmdb_ci_win_server table), make some changes to existing entries, add a few more entries, and then "Using this modified data, create a data source and transform map to import the CSV data back into the instance. Make sure that the correct CIs were updated and the correct CIs were inserted. Did you set up the correct coalesce value?"   I have created a transform map with a source table of Computer[imp_computer] and a target table of Windows Server[cmdb_ci_win_server]. which coalesces on the field model_id (for which a database index already existed, leading me to think there's a pretty good chance that was a good choice). I have auto mapped the matching fields for this transform map, and have left the fields that didn't match up alone. When I click the Transform related link it always says "No import set available. Create and load an import set first." I click on the link to create a new import set, use the existing import set table Computer[imp_computer] and then upload the csv file I created and modified earlier from the Windows server list. I leave both the sheet number and header row set to 1 and click Submit. I then click Run Transform and select the transform map I've created. Then when I go to the import set, all the rows have a state of Skipped and the import set run time is 0 seconds. Of course, this means when I go back to the Windows servers list nothing has changed. Could someone please point me in the right direction here?

Thanks.

[coryseering]

Hi Robert,

Did you modify any data in the CSV file in fields that are also mapped in your transform? For instance, did you update the Operating System or Disk Space fields?

The transform/import skips records where the values being imported are already equal to the existing values.

You probably want to coalesce on something unique to each record, likely the Asset Tag. The Model ID *may* be different for each record, but since this is a reference field to another table nothing guarantees that uniqueness. In the out-of-box demo data for Windows servers, they all have the same Model ID.

[bookman1502]

Yes, I did change the OS field as well as the OS version number. Originally, all of the OS field values were Windows 2003 Standard but I changed some of them to Windows 8, and the OS version number column was completely blank and I added some value for each of the records. I can't coalesce on Asset Tag because while that field exists in the target (windows server) table, it doesn't exist in the source (computer) table.

[coryseering]

Hi Robert,

The Source should be the Import Set created by you when you uploaded the CSV file. The CSV file should have that column because it was created as an export *from the same table that is your target table*. It should have all the same fields.

You can also coalesce on sys_id (which should also be in your export), but since the purpose of that exercise is to see how you would import data from another source and you wouldn't necessarily *have* sys_ids in that case, Asset Tag seems to be the best coalesce value for the target table you chose.

Can you show us the Import Set that got created when you loaded the CSV file?

[coryseering · Accepted solution]

Oh, wait. You did an export from the list view. You only have the columns in your export that were present ont he List when you exported it.

You should be able to add the Asset Tag to the List, export it as CSV, make your modifications, and import that CSV to a new Set.

Then you can coalesce on Asset Tag.