Preconditions

• Link JIRA 7.3.x with Confluence 5.9.x or higher.
• Ensure that JIRA is talking with Confluence via REST.
  • This is the default behaviour in JIRA 7.3.x, but can be toggled by dark feature.
  • Go to /secure/admin/SiteDarkFeatures!default.jspa, and check that "confluence.issuelinks.use.rest.api.disabled" is not present.
• Ensure there are two user accounts can log in to both JIRA and Confluence.
  • I'll call them "bob" and "alice".
• Create a project in JIRA whose issues can be browsed by anonymous users.
• Create a space in Confluence whose pages can only be seen by authenticated users.

Steps to reproduce

1. As Bob, in Confluence,
   1. Create a page that is accessible by any logged-in user. Call it something like "Logged-in test".
   2. Create a page that only bob can access. Call it "Page by Bob".
2. As Bob, in JIRA,
   1. Create a new issue in the public project.
   2. Link to both the "Logged-in test" page and the "Page by Bob" page.
3. Log out of JIRA.
4. View the issue as an anonymous user.

Expected results

...

Actual results

• In the UI, the anonymous user sees a "Page failed to load" message for the Confluence page link with a spinner that spins forever.
• In the JIRA log files, the following message will appear:

  Error occurred while generating final HTML for remote issue link: java.lang.RuntimeException: java.io.IOException: Failed to load Confluence Page from remote server Caused by: java.io.IOException: Failed to load Confluence Page from remote server Caused by: com.atlassian.sal.api.net.ResponseException: Status Code: 404, Status Text: Not Found, Errors: {}, Error Messages: [The requested Confluence content could not be found.]

Workaround

Enable the "confluence.issuelinks.use.rest.api.disabled" dark feature for the JIRA instance. This will cause JIRA to use XMLRPC to communicate with Confluence. Confluence's XMLRPC responses do not return 404 responses for unauthorized requests, so the errors aren't raised on the JIRA side.