Expected Behavior

Issues will be imported based on the ID specified in CSV file, which will be discarded after the import process is complete. This scenario is suggested when importing issues and subtasks on Importing data from CSV - Encapsulating JIRA data structure in your CSV file

Actual Behavior

The values to the Issue ID field during a CSV import are saved to an "External Issue ID" field. When performing new imports, they will usually not work properly because it's rather difficult to find unique ID values. We should have two different options upon import: "Issue ID" - a value that will only be used within that import process to link issues; and "External Issue ID" for when users want to map the ID that came from the source system.

Steps to Reproduce

Try to import a file containing the following data twice, mapping "id" to "Issue ID" and "parent id" to "Parent ID".

id,parent id,summary
1,,"Test"
,1,"Test sub-task"

The value set on "Issue ID" will be replicated to the "External Issue ID" value of the issue with summary "Test". Upon a new import, it will skip that issue and register the following on the import log:

External issue 1 already exists as PROJ-1, not importing.

Notes

This may cause problems with sub-tasks links if you use the parent-id field as it will be mapped to the already existing external issue id and not the current CSV issue id.

Workaround