A way to connect AWS file server to Jira

Question

Hi,

We have a Jira instance (server version 7.3.0), running in Chengdu China. We're a virtual staging service where we take an image, convert it to a 3D space and then create it into a photorealistic rendering.

We do the staging outside of China and the rendering process in China. Currently, we have to store files in Jira and Google Drive.

Our China office has access to the Jira server and a few excepted web sites in the world.

Google drive is only to offer customers files they can use outside of China. Our customers are also limited to what websites they can visit.

We do have AWS China which we have found to be the best when it comes to file serving. However, AWS China is not user friendly for the rest of us non-technical people to host files.

So no Google Drive, DropBox, or other out of the box file hosting services. We do have an FTP, but the chances of something going wrong especially with our contractors is large.

We are currently using Jira's file hosting service to hold files in the 3-4Mb range. Recently we've started to have a need to link 300-400 Mb files to Jira tickets. This is causing additional headache, because Jira is not really designed as a file hosting service, but we need to have the ticket and the file so closely linked that they seem inseparable.

What we would like to do is have it that when we drag and drop a file into Jira, we're not actually storing the file in Jira's server but in AWS China's file hosting service.

I'm looking forward to hearing your thoughts and follow-up questions.

Lauri Gardner · Answer

HI Dave,

I really like your solution and taking your time to research and answer my questions.

EC2 would be a good way to go. Let me chat with my Chinese counterparts and see what they say.

Thanks again for the option giving!

Regards,

-Lauri

Dave Norton · Answer

That certainly makes things challenging :)  Thanks for helping to clarify the situation!

Would it be possible to run Jira in AWS, on an EC2 instance? It looks like you can mount an S3 bucket using Storage Gateway or even third party file systems.

If you mounted your S3 bucket to $jira-home/data/attachments that would in theory, let you store Jira attachments in the S3 bucket. That'd require loads of testing - and as with any remote file system for attachments, you're at the mercy of disruption between the server and the storage. It's probably not such a big thing when it's all in AWS, but it bears keeping in mind.

Would moving your instance from on-premise to AWS be a potential solution for you? That'd help avoid the use of third party apps in Jira, while maintaining Jira's native drag and drop functionality. There would be some effort to move it, and cost associated with running it - but it looks like all the services required are present in AWS China.

Let me know what you think!

Cheers,
Dave

Lauri Gardner · Answer

Hi Dave,

One of the challenges is that AWS China and AWS rest of the world are two different systems and their connectors have to be built by a DevOps team. It still wouldn't solve the issues of not using up our server's storage space.

Cyberduck for example cannot access AWS China. I've tried before.

The plugin still means we would need to link the file manually to the Jira ticket, which is fine on the surface, except that with Jira the ability to drag and drop files is the real power, as we are doing a lot of tickets with a lot of attachments.