Improve S3 destination-prefix handling and validation in BRIE / Data Export

XMLWordPrintable

      Current experience

      A customer reports that Jira and Confluence exports targeting the same S3 bucket produce paths such as:
       
       
       s3://example-export-bucket//example-site_JIRA/s3://example-export-bucket//example-site_CONFLUENCE/
      The additional slash appears as a / folder. Downstream tools that normalize consecutive slashes cannot locate the objects using the resulting normalized paths.

      A previous successful test used a single policy with a named destination prefix. Multiple policies sharing a bucket is an observed correlation, not a confirmed cause. Empty-prefix handling and shared-storage behaviour require validation.

      Requested improvements

      • Construct object keys without adding unintended leading slashes or duplicate separators.
      • Validate destination prefixes when users create or update storage configurations.
      • If an empty prefix is unsupported, require a valid prefix and display an actionable validation message.
      • Detect conflicting destinations between policies and explain how to configure separate prefixes.
      • Preserve valid use of a shared bucket rather than blocking all multi-policy configurations.
      • Provide a destination-path preview so users can verify where exports will be written.

      Expected experience

      Users can configure predictable export destinations, for example:
      s3://example-export-bucket/Jira/example-site/s3://example-export-bucket/Confluence/example-site/
      Any unsupported configuration is identified before the export runs.

              Assignee:
              Unassigned
              Reporter:
              Abhinav Srivastava
              Votes:
              3 Vote for this issue
              Watchers:
              2 Start watching this issue

                Created:
                Updated: