Files/directories not included in a csv import/export between two Jira instances ?

[Bouabana]

Hello everyone !

I am about to migrate two projects from a Jira instance hosted by an external provider (Source instance) to a jira instance that is hosted by my organization internally (Target instance). Both instances are Jira Server/Data center instances.

My organization instructed me to do this migration only via a CSV Import/Export. I found out after doing research that this method doesn't allow to migrate Attachments, history and comments. Obviously, that is not an option.

So my questions are :

Firstly, what are the files that will not be included in a csv import/export ? beside the attachments, history and comments ?

Secondly, is there a way to move these files manually ? Or will they be lost ? Knowing that my organization wants to terminate the source instance, that is hosted externally, after the migration.

And finally, even if i can move these files manually, i want to know how to archive these files ? and store them outside Jira.

Thank you !

[Radek Dostál · Accepted solution]

CSV is never good enough - you cannot get history from it. It does allow import of attachments and comments, but you would need a lot of steps to prepare the data and build a reliable format in the csv.

For all intents and purposes, if you are migrating entire projects, you more often than not would want to use .xml backups.

Out of box, you would use https://confluence.atlassian.com/adminjiraserver/restoring-a-project-from-backup-938847691.html

There are 3rd party apps but I do not know of one which would allow you to import projects as part of trial, so might have to purchase a license which doesn't sound cost-efficient for just two projects.

Basically, once you go through the configuration part (because out of box .xml backup only imports issue data, components and versions), then you would be able to import the issues. That said it is a lot of overhead configuring those projects manually but unless you have any import/export plugins that's pretty much what you'd normally do.

XML backups include comments, attachments, and history. The XML backup itself has references to attachments - but you need to copy/move/symlink the attachments directory from source instance right next alongside the .xml file so that during the import, Jira reads the xml, reads the attachment references, and is able to find the attachment in the provided folder, and then can copy/create the attachment on the source instance in proper directory.

Long story short the above link should cover the steps, but overall you really would want to test this on a non-production instance first. It is likely you will hit into a lot of hic ups and might have to deal with some bug fixes along the way, it's not perfect.

But, it is the only way how you can get history/attachments/comments over, unless you're keen on building complex scripts which have flaws themselves.

[Bouabana]

Thank you !