Data Separation Models for Jira Data Center

[JosephQ5]

We are are an IT Team for a government contractor and considering a hosted Jira Data Center for our Software Dev Teams we support. Each software dev team requires strict data separation,  preferably at the database level. We've seen federated models, assuming these were using multiple instances of Jira Software with Crowd that appear to have met this need. However, given the discontinuance of Jira Software we would host Data Center and wondering what options are available to us?

[Craig Castle-Mead]

Hey Joseph,

If you need physical separation of data (separate databases), then you'll need 1 DC instance per team that needs isolation.

If there's no need to segment the user data, you could run a single Crowd instance that all DC instances authed through, and if you had a consistent domain (team1.company.com/team2.company.com etc) or connected the apps to Crowd via SAML you can get Single Sign On between the instances.

There's no out of the box "federation" that I'm aware of that would allow a user with access to team1.company.com and team2.company.com to be able to search via a single UI across both instances. You'd definitely be able to build a single pane of glass UI leveraging the REST APIs that queried each instance, merged the responses and displayed them to the user. Unsure of the ROI on this for you though.

CCM

[Nic Brough -Adaptavist-]

Welcome to the Atlassian Community!

Server and Data Center are technically the same software now, and the only real difference is how you set them up and licence them.

So, yes, you can flip your servers to DC installations and keep them completely independent and separated. 

You don't even need to do much migration or conversion work - if you are on a high enough version, you can apply DC licences to your servers (bear in mind you'll need to check any apps for compatibility and also get DC licences for the ones that are compatible) and you end up with a DC cluster with a single node.