We recently upgraded Jira server from 6.4.14 to 7.2.9 and found the following 2 potential regressions in version 7.2.9.
problem 1: Has anyone experienced the the problem when using 'Project Import' to import a project to another jira instance that it creates the same label twice in the LABEL table in the same issue?
problem 2: When deleting an issue from the Jira navigator(on browser), the label(s) associated to the deleted issue still remain in the table LABEL.
I have not encountered the specific issue you mentioned in regards to having duplicate labels within JIRA, but I'm curious to learn more about your project import scenario. Are you importing a project taken from a 6.4.x backup into a 7.2.x version of JIRA? If so, there are a number of caveats noted in the project import guide in regards to making sure that the project you are importing is already on the same version of JIRA as the destination.
Restoring a project - Atlassian Documentation
I suspect that you might be seeing both issues due to this version mismatch. If the database layout was changed between these two versions of JIRA in regards to labels, then it could explain why you might be seeing this duplicate entry here.
I would recommend that you update the data in the backup by performing an upgrade to JIRA 7.0.x first, update all the plugins needed for that version, and then upgrade to the 7.2.9 version, again updating the needed plugins. From there you can create a fresh XML backup that you could then use to import this project.
Also in regards to the labels field, there is an open feature request to be able to delete these old or unused labels in https://jira.atlassian.com/browse/JRASERVER-33672
At the moment, JIRA does not have a means to delete these from the UI.
Thank you so much for your response.
The source data where I took the export actaully came from the same 7.2.9 version. Our current project is to move a Jira project from one instance of Jira to another instance of Jira, and both Jira instances have matching version.
Here are steps to reproduce the problem.
1. create a test project in a newly installed Jira 7.2.9 server instance.
2. create an issue with one label name in the test project.
3. take a backup from this Jira 7.2.9.
4. delete the label and delete the issue created in step 2. (note: merely deleting the issue, you would find the associated label would be left in the DB table(LABEL) without being purged).
5. perform project import from the exported zip file in step 3 into the same Jira 7.2.9 server instance. You should notice that 2 same label names being created in the DB table LABEL. You will not notice there are duplicates from the UI (browser) display.
Let me know if I need to open a support case in case you could reproduce the problem from above steps.
Thanks again, James
I was able to repo this in my 7.4.2 instance. You are correct that the project import does appear to be creating a duplicate entry for the labels table in the database, but there does not appear to be any duplicates in regards to the UI where we can see the issue in JIRA.
I'm still not sure why this is. I then went through and removed that label from this issue in the UI once more. Doing so removes both entries from the database. So it does appear that this duplication is a result of the import, but that JIRA UI does not appear to have any problem with having that same label exist twice in that table.
Could you let me know more about why this is a problem for you? Are you doing some external SQL reporting on the JIRA database on this table? If so, I could then see why this is a problem. However if not, and you are only using the JIRA UI, it's harder for me to understand why this is a problem. It's not a requirement to explain why this is a painpoint, but it would help us to then understand how to best address this problem.
It might be we still need to create a bug report or feature request to address this. However it might also be possible to create a SQL command we could run to remove these duplicates, but before we go down that avenue, I just want to know more about why this is a problem.
I am part of data migration team reponsible for migrating data from either Jira systme or non-Jira system into a centralized data-center Jira . The amount of data migrated overtime is large and would case a huge data integrity issue in the target data base if project import is creating duplicate labels. In addition, our DBA would not allow this to happen, as it might impact whether we are allowed continue to do migration until the problem is fixed.
This does not appear to be the problem when I was using previous Jira versions (6.4.x). It appears his might be a regression bug in Jira 7.x.
Thanks again, James
Thanks for explaining that. I agree that the project import should not be doing this. However I have not found any other reported cases for this behavior. So I created a new bug on this over at https://jira.atlassian.com/browse/JRASERVER-65870
I would recommend that you vote on and watch this issue for updates. I do not have a time table for if/when this might be fixed. But I would also invite you to comment on that bug there to try to help explain why this defect is a problem in your scenario, in your own words.
See how to use Jira on the go! Demo Den Episode 5 is here: meet Jira Mobile with Jira Product Manager Rayen Magpantay. Demo Den is in our monthly series where a Jira PM demonstrates...
Connect with like-minded Atlassian users at free events near you!Find an event
Connect with like-minded Atlassian users at free events near you!
Unfortunately there are no Community Events near you at the moment.Host an event
You're one step closer to meeting fellow Atlassian users at your local event. Learn more about Community Events