Jira Data Center vs. Cloud: What testing teams are missing out on

Alisha Siddhartha

November 13, 2026

Picture this: A QA leader sits in on another team’s sprint review. Someone asks Rovo to generate test cases from the acceptance criteria, and a few clicks later there’s draft coverage on screen and a clear view of where tests are still missing.

It looks useful. Then comes the question: Could our team do that today?

If you’re running Jira Data Center, the answer is no. Not because your team is missing a feature or hasn’t configured something correctly. Some of the newer testing capabilities depend on Forge, Atlassian’s cloud app platform, or on Rovo, which runs on it. Neither exists in a Data Center deployment.

That gap is getting harder to ignore.

What this means for your testing team

Atlassian Data Center reaches end of life on March 28, 2029. That gives teams time to plan, but also leaves room for procrastination.

For testing teams using SmartBear Zephyr, the more important question is what you could be doing in Jira Cloud today that isn’t possible in Data Center. Atlassian is building more AI and automation capabilities into Cloud through technologies like Rovo and Forge, its platform for building and running apps. That means testing apps built for Cloud can take advantage of capabilities that aren’t available in Data Center.

For Zephyr customers, that includes AI-powered testing with the Zephyr Agent for Rovo, connected testing workflows through MCP, and no-code test automation.

As Atlassian continues to build on Cloud, the gap will continue to widen. Let’s cover what you need to know if you’re deciding whether it’s time or not to start planning your move.

Key takeaways

The deadline is coming, but the capability gap is already here: Atlassian Data Center reaches end of life on March 28, 2029, but Cloud already offers testing capabilities that aren’t available in Data Center.

Cloud opens the door to AI and modern test automation: Capabilities such as the Zephyr Agent for Rovo, MCP access to test data, and no-code automation depend on Atlassian’s Forge platform and on Rovo.

You don’t have to wait until 2029 to start planning: Zephyr supports the move from Data Center to Cloud, with migration guidance and tools to help teams assess their environment, test the migration, and move when they’re ready.

The Data Center deadline is coming, don’t wait for it

Atlassian has set March 28, 2029, as the end of life date for impacted Data Center products. From that point, Data Center subscriptions and associated Marketplace apps expire, and affected Jira instances become read-only. We’re working to keep your Zephyr test data accessible in read-only form beyond that date, and we’ll share more as we confirm details.

Date	What changes
March 30, 2026	New customers can no longer purchase impacted Data Center subscriptions or Marketplace apps.
March 30, 2028	Existing customers can no longer purchase new Data Center subscriptions, Marketplace apps, or subscription expansions.
March 28, 2029	Data Center reaches end of life. Subscriptions and associated Marketplace apps expire, and affected Jira instances become read-only.

That’s a long runway.

But the deadline isn’t the most important date for testing teams.

The more important question is what your team could be using today.

While Data Center continues to run, Atlassian is investing in Cloud, including the platform and AI capabilities that enable new ways of working. For testing teams using apps like Zephyr, that means capabilities such as AI-powered test generation, connected testing workflows, and no-code automation are already available in Cloud.

So, you don’t need to wait until 2029 to decide whether Cloud is right for your team. And a migration of this magnitude isn’t a weekend’s work, so the teams who get a head start set themselves up for success. You can begin by identifying what you’re missing today, what the move would involve, and what your team would gain from making it.

Your test record needs to keep up with how software is built

Software teams are moving faster, and AI is doing more of the work. Test teams have to keep up without losing the evidence they need to understand what was tested, what was missed, and whether something is ready to ship.

That puts more pressure on the test record. It needs to capture more than manually written test cases and execution results. As testing becomes more automated and AI-assisted, teams also need a way to bring those activities into the same place where they already manage their work.

For teams using Jira for their system of work, that means the capabilities of your test management app matter. If your testing platform can take advantage of the AI, automation, and app capabilities available in Jira Cloud, your test record can evolve with the way your team works.

That’s where the difference between Data Center and Cloud starts to matter.

What Cloud makes possible for testing teams

Some of the newer testing capabilities available in Jira Cloud are possible because Atlassian is building them on Forge and on Rovo, which provides the AI capabilities apps can work with. Neither is available to apps running directly on Data Center.

For teams using Zephyr, that opens up new ways to work.

Start with AI-assisted test creation: The Zephyr Agent for Rovo works inside Jira to help teams generate test cases from requirements, explore test coverage, and identify gaps. Instead of starting every test case from a blank page, testers get a first draft and apply their expertise to it.

Connect testing to the AI tools your team already uses: The SmartBear MCP Server for Zephyr makes test data available to AI tools and agents, so teams can bring testing information into workflows beyond Jira. This keeps test cases, cycles, and results connected rather than creating separate testing records across different tools.

Automate more of your regression testing: Zephyr’s no-code automation lets teams create and run browser tests without writing scripts, including against private environments, while capturing information such as video and network logs when tests fail. Results can feed into CI/CD workflows.

The bigger shift isn’t any one feature, it’s that testing can become part of the AI-assisted and automated development workflow, while the test record stays connected to work happening in Jira. That’s the opportunity Cloud creates: room for your testing process to evolve.

Which of these your team can use depends on the Zephyr Cloud edition you land on, so it’s worth settling that now rather than post-migration.

Capability	What it enables your team to do	Zephyr Cloud edition
Zephyr Agent for Rovo	Use AI in Jira to generate test cases, explore coverage, and identify gaps	Standard or Advanced
MCP access to test data	Connect Zephyr test data with AI tools and agents outside Jira	Standard or Advanced
No-code test automation	Create and run automated browser tests without writing scripts	Standard or Advanced
Data residency	Keep Zephyr application data in a supported geographic region	Standard or Advanced
Advanced execution capabilities	Expand browser and channel coverage and run tests in parallel	Advanced

The point isn’t to pick an edition from a feature checklist. It’s to start with what your testing organization needs. For a QA leader that might mean AI assistance for testers and more automation coverage. For an engineering leader, connecting testing to AI-assisted development. And if you’re thinking about the return, understanding what your team stands to gain by moving sooner rather than at the deadline.

And you don’t have to work it out alone. Our team can help you understand the options and support you through the move.

What changes when you move to Cloud?

Data Center may still support the way your testing team works today, but it has a fixed end date: March 28, 2029. That makes it important to understand what will change when you move, especially if your team relies on specific reporting, administration, or configuration capabilities today.

Reporting is one of the biggest changes. Data Center offers up to 70 reports and 60 gadgets, compared with 31 reports and 19 gadgets in Cloud. Confluence reporting capabilities aren’t available in Cloud at all. If your team relies heavily on these today, factor that into your migration planning.

Some administration and configuration will change. Zephyr’s global administration pages don’t exist in Cloud. As a result, the Data Center controls for project enablement, analytics, thresholds, global access, and project restrictions don’t carry over.

Two other features don’t come across either. Automatic traceability links, which sit in project (space) level settings in Data Center, aren’t available in Cloud. Folder custom fields, which sit under the app’s configuration, aren’t available in Cloud, and the migration doesn’t include them. If your team uses any of these today, it’s worth mapping out how you’ll handle them in Cloud as part of your planning.

Your test execution history comes with you. Historical test executions and their step-level results migrate to Cloud, so your testing record and results stay intact. Test execution change history, the record of who updated a status and when, doesn’t migrate today.

The important thing is to understand these changes early. Data Center is approaching its end of life, so the question isn’t whether your team can continue using it indefinitely. It’s how to plan the move in a way that preserves what matters to your team while giving you access to the capabilities available in Cloud.

For the full field-by-field comparison, see the Zephyr Server/Data Center vs. Cloud comparison.

How secure is Jira Cloud compared to Data Center?

For teams who’ve run Jira on their own infrastructure for years, this is often the question that decides the timeline. The good news is that it’s usually easier to answer than teams expect.

Atlassian works on a shared responsibility model, so it’s worth being clear about the split. Atlassian handles the security of its applications, the systems they run on, and the environments hosting them, and publishes the proof in the Atlassian Trust Center for your security team to review. Your team keeps what it should: the information in your accounts, who has access, and which Marketplace apps you install and trust. Zephyr is SOC 2 Type 2 certified and GDPR compliant, with current reports available in the SmartBear Trust Center.

And if your organization has regional requirements, data residency keeps your Zephyr data in the U.S., E.U. (Ireland), Germany, or Australia.

With Data Center, your team manages the infrastructure as well as access and governance. With Cloud, the infrastructure half moves to Atlassian, so your team can focus on product quality and delivery.

Thinking about the move? Here’s where to start

You don’t have to figure out your move to Cloud all at once. Start by understanding your current environment, what your testing team needs, and what the move would look like.

Start with what your team needs:
Look at the testing capabilities that matter most to your team today and the ones you’ll need as your development process evolves. This gives you a better starting point than simply comparing feature lists.

Check your current setup and migration requirements:
Review your Jira and test management versions, dependencies, and any prerequisites for moving to Cloud. If you’re using Zephyr test management for Jira, you can find the specific requirements in the Zephyr test management migration guide.

Test the move before you commit:
Migrations run through Atlassian’s Jira Cloud Migration Assistant (JCMA), which is the recommended and only supported approach for moving Zephyr data from Data Center to Cloud. You can build your own scripts against the Zephyr API instead, but there are limitations and we don’t support or troubleshoot that route. A test migration gives you a chance to see how your projects and test data will move, identify anything that needs attention, and get a realistic view of the work involved, without disrupting your live instance.

Plan the move around your business needs:
Review what will migrate, what may need to be reconfigured, and the commercial options available to you. Atlassian also offers migration incentives for eligible customers, so it’s worth discussing those with your Atlassian account team early.

If you’re using Zephyr test management, the Zephyr test management Cloud migration resources can help you understand the supported paths and prepare for the move.

And you don’t have to do it alone. Whether you’re moving from another Jira test management tool or already using Zephyr test management, our team is here to help you understand the path to Cloud and find the right approach for your testing organization.

Evaluating test management for Jira?

If you’re choosing a test management solution for Jira rather than migrating an existing one, Cloud should be part of the decision from the start. New Data Center subscriptions and Data Center Marketplace apps are no longer available to new customers, so it’s worth looking at where each solution is investing for the future.

When comparing options, look beyond today’s feature list. Consider where the app runs, how it’s evolving in Cloud, and what capabilities your testing team can access as your needs grow. With Zephyr test management for Jira, that includes AI-assisted testing with the Zephyr Agent for Rovo, connected testing workflows through MCP, and no-code test automation.

If you’re evaluating Zephyr test management for Jira, the Zephyr documentation covers how it works in Cloud, and a demo is the fastest way to see whether it fits how your team tests.

Ready to see what Cloud could mean for your team?

You don’t have to wait until 2029 to start exploring your options. See what your testing team could unlock in Cloud, understand what the move would look like, and decide when the time is right for you.

Talk to the Zephyr team to better understand your migration path, or start a free trial to see what Cloud could look like with your projects.

Frequently asked questions

When does Atlassian Data Center reach end of life?

March 28, 2029. Data Center subscriptions and their associated Marketplace apps expire on that date and instances become read-only. New customers lost the ability to buy Data Center on March 30, 2026, and existing customers lose it on March 30, 2028.

What happens to my Marketplace apps when Data Center ends?

They expire with the subscription. Atlassian’s end of life covers third-party Marketplace apps for affected Data Center products as well as Atlassian’s own, so those apps stop working when the instance goes read-only on March 28, 2029.

What happens to Zephyr when Data Center ends?

Zephyr for Data Center expires with the rest of the instance on March 28, 2029. We’re working to keep your Zephyr test data accessible in read-only form beyond that date, and we’ll share more as we confirm details. The supported route to keep working is a migration to Zephyr on Jira Cloud using Atlassian’s Jira Cloud Migration Assistant.

Can I still buy Atlassian Data Center?

Only as an existing customer, and only until March 30, 2028. New customers could no longer purchase Data Center subscriptions or Data Center Marketplace apps after March 30, 2026.

What do I gain by moving test management from Data Center to Jira Cloud?

AI-assisted test creation, agent access to your test data, and no-code test automation. In Zephyr that means the Zephyr Agent for Rovo generating test cases from Jira work items, an MCP server exposing test data to agents outside Jira, and browser automation with no scripting. Several of these require Zephyr Standard or Advanced.

Is Jira Cloud secure enough for regulated teams?

Yes, and many regulated organizations already run on it. Atlassian publishes independently verified certifications in the Atlassian Trust Center, and Zephyr is SOC 2 Type 2 certified and GDPR compliant, with reports available in the SmartBear Trust Center. If you have regional requirements, data residency keeps your Zephyr data in the U.S., E.U. (Ireland), Germany, or Australia.

What features are available on Jira Cloud but not Data Center?

Everything built on Atlassian Forge or Rovo. Forge is Atlassian’s cloud app development platform, and Rovo is Atlassian’s AI product that runs on it. Neither is available to apps running on Data Center. For Zephyr that covers the Zephyr Agent for Rovo, MCP access, no-code automation, test step suggestions, and cross-browser and parallel execution.

Can we migrate Zephyr data ourselves using the API?

JCMA is the only supported route, and it’s what we recommend. You can build your own scripts against the Zephyr API, but there are limitations and SmartBear doesn’t support or troubleshoot that approach.

Is Rovo available on Atlassian Data Center?

Not in the way it works in Cloud. Atlassian offers connectors that let Rovo reach some Data Center content, but Rovo agents built by Marketplace vendors run on Forge, which is Cloud only. The Zephyr Agent for Rovo therefore requires Jira Cloud.

Does my test execution history migrate to Cloud?

Yes. Historical test executions and their step-level results migrate. Test execution change history doesn’t.

Which Zephyr capabilities depend on edition rather than deployment?

The Zephyr Agent for Rovo, no-code automation, the MCP server, and data residency are available on Zephyr Standard and Zephyr Advanced. Moving to Cloud doesn’t grant them on its own.

Is Zephyr Data Center still supported?

Yes. We continue to support Zephyr for Data Center with critical fixes and bug fixes through to Atlassian’s end-of-life date. There’s no active development for Data Center, so we build new capabilities for Cloud.

How long does a Data Center to Cloud migration take?

It depends almost entirely on the size of your estate. A test migration run against your own data is the only reliable estimate, and it has no effect on your live instance.