How to import attachments using CSV into Jira Cloud?

[Radu Rendec]

Hello Everyone,

We want to import the data from our legacy issue tracker into Jira Software (Cloud) and we are currently experimenting with the CSV import tool. While the tool seems to work well and satisfy most of our requirements, I am having trouble importing attachments.

I know that the attachment field in the CSV can use an (external) http:// or https:// URL to retrieve the file from, but that would require (1) setting up a web server and uploading the files and (2) making the content publicly accessible, so that the Jira Cloud instance can access it. The former means extra work and the latter is downright unacceptable, since the contents of the attachments is confidential and that would violate the company policies.

It appears that the CSV import tool also supports file:// URLs, but that's not well documented. I noticed (by trial and error) that all paths following file:// are considered relative to /data/jirastudio/jira/home/import/attachments. Is there a way to upload content to that location in the Jira Cloud instance? For example through the Restore/Import tool?

I already tried to use "Restore system"/"Import media" and upload a zip file with an "import/attachments/" path inside, but that didn't work. I'm assuming the zip file is extracted somewhere other than /data/jirastudio/jira/home.

I also looked at the REST API as an option to upload attachments, but it doesn't seem to support specifying a different user and time, so the attachments are always created as the authenticated user and with the current time (the time when the API is called). Obviously, this is not suitable for importing historical data from a different system.

Thanks,

Radu

[Bastian Stehmann · Accepted solution]

Hi,

If you use http/https you could add username and password in the url, so the data can be protected and is not public available. But you will need a Webserver anyway.

I don't think there is a way to upload the data somewhere to the filesystem in cloud.

[Matt Doar]

I've used that approach before, but it was a while ago. These days with Jira Server I download the attachments and then run a separate script to call the REST resource to add an attachment. That way I get to preserve the attachment name. However the original author name and date are not preserved, and the updated date is changed.

So no great solution for me with on-prem. There might be something better with the Cloud v3 REST API?

[Radu Rendec]

Thank you for the suggestions! I am going to accept the answer because it seems to be my best option (or actually my only option). It's not exactly what I was hoping for, but gets the job done.

This looks to me like a serious limitation of the Jira CSV import tool, or rather an unfinished/incomplete feature. I don't know if/how it can get the attention of the Jira design team. I am sure they get many requests from many users and they don't really pay attention unless there is significant "noise" around a specific request.

[Saifulmazli]

Did you managed to get his ? I am trying to create 1200 issues with 1200 attachments on jira cloud, and it's not making life easy

[Radu Rendec]

I haven't had any breakthrough since my original post. It looks like your only two options are external URL (with username/password) or uploading the attachments through the REST API. The former requires setting up a publicly visible web server, while the latter doesn't support overriding the attachment author and creation date/time.

We decided to go with the REST API and found another limitation or rather side effect. When you create an attachment through the REST API, the assignee gets an email notification. One more reason to think that the REST API was not designed for importing historical data.

There are bug reports against all these problems. They were created years ago and eventually closed as "won't fix".