Update existing issue via import

[Tao Yuan]

Hi, I am trying to update existing Jira issues via CSV file. I did have 'Issue Id' field for each issue. However every time I tried to update I got a new issue created for me. The existing one doesn't get updated at all. Can any expert help? Thanks!

[Andy Heinzer · Accepted solution]

Hi Tao,

I understand that you're trying to update issues in Jira Cloud via a CSV file, but that this is not working as expected and new issues are created instead. This is possible to do in Jira with a CSV file, but there are a couple of caveats to doing this.

There is a scenario where new issues are created instead of updating existing issues.  Jira has two different features that can both import issues from CSV files.  However one of them is only available to Jira administrators, and as such that particular External System Importer has some abilities that the other importer does not have. 

You can confirm which one you are using by taking a closer look at the site URL.  If your URL on that page where you created the screenshot looks something like this:

https://example.atlassian.net/secure/admin/BulkCreateFieldMapping!default.jspa?externalSystem=com.atlassian.jira.plugins.jim-plugin%3AbulkCreateCsv 

Then you're not actually using the external system import utility.  If you are a Jira admin, you can go to Cog Icon -> Settings -> System -> External System import.  If you're on the correct page, the URL will looks like this:

https://example.atlassian.net/secure/admin/ExternalImport1.jspa 

From there you can select your CSV file and import this. 

However from looking at your steps, I don't think you will be able to update existing issues based solely off the issue id field.  That field is important for establishing parent/child relationships of subtasks, but in order for you to update existing issues, you actually need to map to the IssueKey field instead.  This is explained in Importing data from CSV: Updating existing issues.

To update existing issues, your CSV file needs to contain a column that maps to Issue Key. If an issue exists for a given key, it will be updated.

Your CSV file will also require the summary field.  So it should still be possible for you to update these issues in Jira via CSV, but to recap, your file needs to have IssueKey, Summary, and any other fields you want to update on existing issues there.  And again, I'd recommend making sure you're using the external system import as an admin instead of the bulk create function.

Try this and let me know if this helps.

Andy

[Guy LAROCHELLE]

Jumping in here, why is the Summary field required ?!?  I do not want to update the summary, so why would I need to include it!

[Andy Heinzer]

The CSV importer is not smart enough to understand when you go to use it that you might only want to update existing issues.  It was initially designed as a way to get issues into Jira from an external source quickly.  It does that. 

Summary is one of the only required fields in the field configurations that you cannot make optional.  There is an existing feature request where users are seeking the ability to make Summary optional over in JRACLOUD-69507.

As such, the importer is trying to honor the field configurations that exist.  It doesn't know that it might be updating existing records, it has been built with in the specs that it will have to create new issues.  And when creating new issues in Jira from any source, you would you be obligated to have some value to provide to the required fields such as summary.

[Guy LAROCHELLE]

Frankly, it is probably the easiest thing to change in the whole of Jira, seriously. If the ticket key is there, then it is an update, that's all. The only required field should then be ticket key.

Also, the Jira ticket you gave is not the same thing at all. Summary should not be an optional field for a ticket. It's two different things.

[Algorithmism]

this is so painful.

Each Epic, User Story, Task, etc. should be treated as an object with a unique ID. 

Any time there are updates to that UniqueID, consider it an override....

Did they add a bulk update feature yet? 

[Ben Cundy]

Does bulk updates work in Jira Server configuration?

[Marc Isikoff]

Again, Atlassian wants us all to be on Jira Cloud - so forget that our companies went to a different cloud provider and most of us have Jira Server "in our company's cloud provider" and Jira Cloud just is not going to happen.

They need to understand that "solving" a problem with a solution only available to Jira Cloud customers disenfranchises a whole lot of users.

Also, External system CSV import - if I were to wait the 2 weeks for permissions to have someone in my company import CSVs I would have 2 weeks of data that wouldn't be included.

The cleaner solution is to allow CSV import that is available ubiquitously to have an easy path to update existing issues.

[Guillermo]

Jumping in here to raise voice and say it does not work when the field to be updated is resolution date: the csv is processed and there is no error message anywhere, but there is no update on the `resolutiondate` field.