CSV import, Next-gen, assign tasks to epics. Possible?

[Heather Kendall]

We are trying to import tasks into our next-gen product.

Notably, to gain access to any epic fields to map we cannot go through the project's csv import dialogs, but have to go to the system import dialogs.

When we look at "epic link" and similar fields in advanced view columns in our lists, the values are blank for tasks that are assigned.

So, we are wondering if there is something else we should be mapping (if so, what are the steps to make that happen?) or if the next gen projects just don't support assigning tasks to epics as part of the import yet.

Thanks!

[Andy Heinzer]

Hi Heather,

I see that you are looking to import issues into a next-gen project and are having some problems with maintaining the epic/child relationship.  Next-gen projects have a different means of determining an issue's epic than the classic projects do. 

In classic projects, issues will tend to have an epic link field that references the issuekey of the epic itself, this epic link field is a global field, meaning it is used throughout all projects in Jira.  However next-gen projects don't actually have this global epic link field to use.  Instead they are showing this relationship by referring to the epic's issueid under the parent field.

If you are exporting this from another next-gen project, then when you do so you can see something like this:

Every issue will have a unique issue id.  That issue id of the epic type issues is referenced in the parent field.  That same parent field is also used to link subtasks to their parent task issues. So it is possible to maintain this epic link in a next-gen, but you have to do things a bit differently here.  First, we need to make sure that the issue id you are importing here are unique to this Jira site.  If they already exist in Jira, then using the external system importer could overwrite existing data.  In this example above, GOG-9 is one of the epics (which an issue id of 10233), and its child issue is GOG-1 which references that id in the Parent field.

I think the best way to manage this is to actually break it up into two imports: first import of just the epics themselves and the second import will then bring in all the other issues. 

When you do this first import, don't map the issue id or issuekey fields.  That way these imported issues will be considered new issues will have their own new unique issue id values.  Mapping the issuekey or issue id could overwrite/update existing issues that would match on those values.

After you import the epics, you can then do a new export of just the epics to see what their issue id values are, and then use that information to update the csv for the second import so that the parent field will match that issue id value of the epic.  Provided that the second import maps that parent field and the new csv file as this updated parent id value, this should work.  Admittedly this is more complex to achieve, but I believe it is possible.  If this project has subtasks, you might have to break this up into three imports to accommodate the subtasks relationships in the same manner.

Let me know if you have any questions or concerns about this.

Andy

[Katelyn Botts]

This is the most helpful response I've found on this topic! Unfortunately, I'm still struggling. :( 

Does the order of the column headers matter for the import? I read somewhere that it does. Everything I import is becoming a story or task but I'm having troubles with getting epics to import as epics, and I'm having troubles with getting imports to link Epics --> Stories --> Tasks/Subtasks.

[Katelyn Botts]

Or Epics --> tasks --> subtasks would also be fine

@Andy Heinzer - is this something you can help me with?

[Katelyn Botts]

I keep getting an error like this

[Andy Heinzer]

@Katelyn Botts There are actually two different means to create issues in Jira via CSV, and these two ways are not created equally.

Method #1: External system import.  > System > External System Import. This method is only available to Jira Administrators, but this import can allow you to map some fields that the other importer cannot.  More details in Importing data from CSV.

Method #2: Bulk Create.  From the Issue Navigator (JiraURL/issues/) go to the ... in the top right corner and choose Import issues from CSV.  This method is available to any user with create permission in one or more projects.  This method though is limited in regard to what issue fields it can set on import.   Off-hand I'm not certain this 2nd method would be able to manage this parent/child relationship.

If you're a site-admin, I'd almost always recommend the first method as it lets you set just about all the issue fields possible.

I would be interested to learn which method you are using here.  It looks like issues 49, 50, 51 are being imported as Stories and not sub-tasks.  Which would make me believe that the issue type is not getting mapped for some reason here.  The column order does not matter to the best of my knowledge.  If that doesn't help here then in your case, I would actually recommend creating a support case for this over in https://support.atlassian.com/contact

I suggest this in your case, because I can see that you are a site-admin entitled to support, and honestly to help you here I think we'd need to see some details of your CSV file here which you probably don't want to share on our public forums.

For other users that might find this post:  If you're using a Free plan of Jira Cloud or you are not the site-admin, then you won't be able to create the support ticket by following that link above, instead you would be redirected here to community again for support per our support offerings.  That said I'm still willing to try to help, please just understand that I might need to seem more of some sample data to be able to better understand the import problems here and try to recreate them.

Cheers,

Andy

[Katelyn Botts]

Thanks @Andy Heinzer

I'm using the first method; the external system import. Should I be able to import stories and tasks within the same csv import? This is a test import - no real data - but I'm having troubles getting the tasks to import over. Also, how about Epics? Is it possible to do a single import that maps Epics, Stories, and Tasks correctly?

[Andy Heinzer]

Hi Katelyn,

I see you created a support case for this over in https://getsupport.atlassian.com/servicedesk/customer/portal/23/JST-588442

Please include your sample CSV file to that ticket, and indicate which project (or project key) you are trying to import these into.  Someone from the Cloud support team will need to take a closer look at your data and your project settings to better understand the specifics of this kind of data import.