GSA Connector and Jira superuser account

Question

Hi,

We're in the process of setting up a Google Search Appliance connector that will crawl Jira to index related documents and pages for searching. We've been informed by the third party, that a service account with super-user access is needed in order to effectively restrict results based on security.

We'd like to know if there is a better solution other than providing a superuser account to the third party for the GSA connector to properly display all available documents and pages, including those that have specific restrictions?

Thanks in advance.

Nic Brough -Adaptavist- · Answer

JIRA does that automatically.

That's why the GSA needs a privileged account - so it can read and duplicate the JIRA restrictions.  (I don't know how GSA might "effectively block restricted pages").

The people I've worked with who have implemented crawlers to index their JIRA installations have kept it simple though, in order to avoid the risks of leaking data via the crawler.  Given the scenario:

* Alice uses JIRA and has restricted an issue to just her team.
* Bob uses JIRA, but is not in Alice's team, so he can't see the issue in JIRA.
* Bob uses the GSA, but he's in a group that has full access to everything in GSA.  So he can read the data the GSA has pulled from Alice's restricted issue.

In other words, if you need any level of security, your GSA access has to match your JIRA access rules exactly.  My clients kept it simple by saying "we are only going to index things a normal, or anonymous, user can see".

SCARM Team · Answer

Thanks for your input. Also, is there a mechanism to effectively block restricted pages/documents from being rendered?

Thanks

Nic Brough -Adaptavist- · Answer

It needs to be a JIRA admin, not a System admin account, because it needs to be able to read the restrictions set up inside JIRA to be able to work out external visibility.

If for example, issue ABC-123 is restricted to the group "penguins", then you need a JIRA admin account to look at the issue security scheme to extract that information and replicate it in the crawler index.  A non-admin user will not be able to see the issue at all, or look at the scheme that says "only penguins"