Anybody using non-AWS S3 storage to store Confluence attachments?

[Andreas von Euw]

We are using a non AWS-S3 Storage Provider. To store attachments we would like to enable the confluence s3 feature. That basically works. We can upload new attachments and store them on s3. 

But what about the existing attachments. Any recommendation how to copy them to the s3 bucket? 

Another topic is the reliabliity of this solution? Because same setup does not work in Jira Datacenter. Answer from Atlassian is more or less, it's your problem. Your storage-provider does not support the latest security features of s3. So what happens when Atlassian chooses to implement the some on Confluence ..... ? 

But funny thing is: Internally Jira and Confluence use the same version of the AWS SDK libraries .... just not using them identical :(

Any recommendations about this topic from the Community?

Thanks, and best regards, Andy

[Anwesha Pan _TCS_ · Accepted solution]

Hi Andy @Andreas von Euw 👋🏼

To answer your 1st part of the question: To copy existing attachments from non-AWS storage to an S3 bucket, below are few methods :

• AWS CLI
  • aws s3 cp :For individual files or smaller transfers.
  • aws s3 sync: For synchronizing entire directories, especially useful for large numbers of files or keeping a local directory mirrored in S3. It only copies new or updated files.
• AWS Management Console
• AWS DataSync
• Programmatic Upload: If you have a custom application or need to integrate the upload process into a larger workflow, you can use AWS SDKs in various programming languages (Python, Java, .NET, etc.) to programmatically upload files to S3.

and many other options are available as well.

I hope this helps.

Thanks,
Anwesha