CSV instead of updating, is duplicating cards

[Luciano Sampaio (Thera)]

When using bulk update thru CSV, header file has Issue Key (mapped to Issue Id), Summary (obligatory field), and Description.
When importation is run, instead of updating cards of same Issue Id, it creates new cards, with new numbers, ignoring the Key value in the column, regardless if we set checkbox in Map Value field (already tried all options of checking the key or updatable column). Wht is the tricky for making csv file update the value of a list of cards listed in the csv file?

[Jack Brickey]

Sure seems odd on the surface. Certainly CSP import has its warts but as long as I had the issue key and summary than any other fields would be considered updates. I expect you may need to reach out to Atlassian support and share of the import file and let them try to inspect what's going on. One final comments/question I assume that the summary field is exactly the same and not being edited.

[Luciano Sampaio (Thera)]

We had tried again.

Initially Summary field was target to be updated as well, but based on your statement, we updated manually those fields before updating thru CSV. Our file contained 11 records, 5 for updating values of Description, and 6 for creating new records. Everytime it had imported as new cards. We tried by checking the checkbox Map Field Value, in all different manners (none checked | ID + summary | only ID | all 3 | only summary).

Every attempt resulted creation of 11 cards. 

Last attempt was to delete lines of cards to be created to have only updating data in the CSV, and attempted to run all possibilities of checkbox. None of them gave us updated data, only new cards. We are opening an issue!

Thanks for the feedback

[Jack Brickey]

If I understand what you're attempting to do here is what I would recommend. Of course you may already be doing this in which case definitely reach out to Atlassian support.

1. Leave the summary fields untouched and export the issues you wish to update using CSV.
2. Using the exported file, open in excel and make any modifications you wish however leave the issue Key and summary columns completely alone. Save the file as CSV.
3. Import this new CSV file which incorporates your edits
4. Change any summaries desired within the Jira application.

[Luciano Sampaio (Thera)]

Unfortunately, we exported the content, changed the column and re-imported, resulting same behavior (duplicating cards).

So we'd opened a bug to support. oddest information is that we'd performed it once and it worked. I suspect some change in the instance level configuration is retrieving this undesired behavior, but I have no clue idea!
Will update this post by support feedback whenever it happens!