The Future of Atlassian Data Center

[#1]

Hello Data Center community,

My name is Gosia Kowalska, and I am Head of Product for Data Center. I am posting here to discuss the future of Data Center and what its future looks like within Atlassian’s cloud-first strategy.

We know that we have not been shy about sharing our cloud-focused investment, and following TEAM 22 some of you have shared concerns about what that means for the future of Data Center products. We are keenly aware that Data Center is still mission-critical for many of you. As such, we want you to know that we’re committed to continuing to support you on your journey with Atlassian whether you are currently in the process of a cloud migration or choose to remain self-managed for the foreseeable future.

To be clear, we have no plans to end of life or end of support for our Data Center offering.

What is the future of Data Center?

Data Center remains an integral component of Atlassian’s offerings to ensure our software continues to meet the unique needs and requirements of all of your organizations. We are shaping Data Center’s future investment strategy to meet what you have told us are your most critical needs:

• Performance and scale

• Security and compliance

• Infrastructure and operation

In sharpening our focus we believe that we can best meet your business and mission-critical needs today while being able to evolve and change with you as your needs possibly shift in the future. Data Center will continue delivering a full enterprise-grade platform with all the features and capabilities you have come to rely on and love.

Any feature or capability that is currently on our public roadmap will be delivered and supported even if it is outside the bounds of our aforementioned future investment.

Focusing on your most critical needs

We will stay committed to delivering platform and product-level improvements to help your administrative teams meet their most critical needs, while some of these features and capabilities will drive additional and improved end-user value. Read on to learn more about how we will increase our focus and investment in the three aforementioned areas.

Due to Data Center’s focused initiative to support on-prem and self-managed customers' most critical needs, there will not be feature parity with our cloud offering.

Performance and scale

To continually enable you to meet your long-term growth goals and maintain the confidence that Data Center can support your user growth, increased loads, and more complex data models, we are doubling down on performance and scale. We will be proactively addressing performance and scale limitations in our Data Center products as they are identified, either by removing the bottleneck or by providing strategic guidance, documentation, and tools to admins to remove the limitations.

Future investments will focus on core platform features, customizable controls, and improved strategic guidance across capabilities such as:

• Cleanup and archiving

• Monitoring and insights

• Apps and integrations

Security and compliance

There’s no denying security demands and compliance regulations are continually evolving, and it is because of this that many of you are currently unable to migrate to our cloud offering. We understand a critical part of managing a behind the firewall instance is meeting and mitigating these risks and requirements. We will both continue, and evolve, our security uplift program to optimize our development and security practices and take a stronger and more proactive approach to all things security and compliance. This will enable us to not only respond quickly to the changing landscape but also alleviate as many complexities and obstacles as possible. Thanks to our security program we have delivered over 2.5 thousand security fixes since last July.

You can expect future work in areas such as:

• Building towards evolving compliance demands i.e. accessibility

• Continuing proactive security programs and practices

To learn more about Atlassian’s commitment and position on security and compliance as an organization head over to our Trust Center.

Infrastructure and operations

We recognize that just because you must remain self-managed does not mean you don’t desire common benefits provided by the cloud, such as reduced organizational costs and maintenance burdens. To help address this, we are focusing Data Center’s future development around containerization, both simplifying the journey to containers and Kubernetes, and optimizing it. We will be positioning Kubernetes as our recommended deployment option to increase flexibility while opening up multiple avenues to help address and solve future demands.

Expected opportunity areas include:

• Increased automation

• Easier administration

• Improved scalability and resilience

To discover more detail about specific features that we have shipped and are planning to ship surrounding these areas go check out our Data Center Roadmap

Maintaining best practices and navigating changes

Data Center customers can expect to continue receiving standard software best practices including consistent and secure releases, Long Term Support releases, platform releases, documentation, support (both paid and unpaid), CVE notifications, bug bounty programs, and the availability of all current feature functionality.

Sharpening our focus in the aforementioned areas ensures that Data Center will consistently meet your most critical enterprise needs and demands. Additionally, we will continue to work to create a viable path to the cloud when you decide to explore this path.

On Wednesday, May 25th (PST) we will be hosting an all-day AMA to answer your questions about the future of Atlassian Data Center. Join the AMA here to ask your questions in advance and remember to join the new Data Center community group to stay up to date on all things DC

Cheers,
Gosia Kowalska and the Data Center Team

[Evangeline Black]

Thank you for this article @Gosia Kowalska . I have a minor project that needs completion but in a very demanding and dynamic environment. It is reassuring to know the data centre will continue to support the slower customers.  

[Rodney Nissen - ReleaseTEAM]

"Some of you have shared concerns."  Who, me? http://thejiraguy.com/2022/04/13/atlassian-we-need-to-talk-about-data-center/

Honestly, I still think large companies can absolutely use tools like Jira Work Management, Compass, and Atlas.  Unfortunately for many of those companies, Data Center is their choice for a myriad of reasons.   And I think that is a MASSIVE missed opportunity to genuinely make the lives of people using some of the largest deployments of your tools easier and more efficient. 

Yes, I want improved performance and efficiency.  I get that on the DC Architecture, that's not easy. But I also want to wow my users with improvements to how they work too.

[Dave Liao]

What Rodney said.

Quality of life improvements are craved by the Data Center user base, improvements that either can't be met with third-party add-ons, or make little sense to address by any third-party vendors.

There's a ton of JAC tickets that continue to get votes and comments from end-users and admins. Addressing some of these can generate good-will from your user base and win you more advocates.

[Ed Gaile _Atlanta_ GA_]

@Rodney Nissen - ReleaseTEAM that was a very well written and thought out article and I am very glad to see Atlassian addressing some of the concerns. 

We are fully on DC since we are in the Health Care realm and leadership has HIPAA compliance concerns in the cloud.  I know Atlassian says they are HIPAA compliant in the cloud - but that is for enterprise level only AND you have to turn off email. 

I have to imagine many large Government, FinTech, and Health Care are going to want/need to stay on DC.  Hopefully Atlassian stays true to their word.

[Nic Brough -Adaptavist-]

@Ed Gaile _Atlanta_ GA_  - you missed "Defence" and security organisations off the list.  I know there are a number that just laugh when someone says "Cloud" to them.

I've also got a rather difficult meeting later this week - a large customer trying to decide between DC and Cloud, they added "the recent outage" to the agenda...

[Mayuresh Sakharape]

Thanks @Gosia Kowalska for this assuring post for DC users. But I still feel you also need to develop new functional features in DC products along with above mentioned non-functional features.  Else you would end up driving away many potential customers.

[Ravi Sagar _Sparxsys_]

Thanks for the update @Gosia Kowalska 

[Carmen Nadeau]

For me, it read like, "we know you want more but we have only XX amount of time/money to devote to the DC so it will go to the infrastructure, be happy, we heard you"

I, too, am happy that the infrastructure will be.."upgrade" but, the reality that most of the new functionnalities that kept being deployed in the Cloud, will never see the light of day in the DC environment, is disheartening for us admin who love your products.

The fact is that all cies related to the finance sector cannot move to the cloud until there is a data residency in our country (fo us it is Canada). And with what just happened makes it even harder to convince the business people to move to the Cloud.  So DC is it for us for a while and, again, I will miss looking forward to a new version with all the new functionnalities available in this new release.

Carmen

[Rodney Nissen - ReleaseTEAM]

From today's post on https://thejiraguy.com

My Thoughts

First, let me start by saying I sincerely appreciate Atlassian putting this statement together. Considering the tone of most of my writing on thejiraguy.com is very much Pro-Atlassian, I can imagine that last week’s article came rather unexpectedly, as did the response where more people seemed to agree with my general statement than disagree. 

However, after Team ’22, I was deeply bothered by the apparent lack of Data Center content and felt I could not ignore the topic and still do justice to the position the readers of the blog have given me. And while Atlassian’s response did little to change my opinion of the situation, it at least shows they are aware of the impression most people seemed to have following the conference. 

As far as the content of the post, it was not surprising. It seemed to confirm Cameron Deatsch’s comments I posted last October – which was that we should expect little – if any – user-facing features in Data Center moving forward. However, I still feel this is a mistake.

As I stated in my earlier comment, I believe large companies can absolutely use tools like Compass, Atlas, and Jira Work Management. It was these large companies I had in mind when I was talking with Atlassian about their new Jira Core revamp during the development of Jira Work Management. And many of these large companies – for a myriad of reasons – have chosen to make Jira Data Center their preferred home.  

Don’t get me wrong; as a DC customer, I want security and performance improvements. But I also want new features to wow my users as well as my fellow Jira Admins. Unfortunately, I think Atlassian is missing a huge opportunity to make the lives of the people using some of the largest deployments of their toolset easier and more efficient. 

The cynic in me wants to say that this is intentional. That Atlassian purposely tries to make Cloud more attractive by withholding features that would otherwise make sense to have On-Prem.  

However, I fully admit that the opposite could be true: The architecture of the Data Center products makes it such that adding these features would prove an undue and expensive burden, which otherwise would not make business sense. However, if the latter were the case, I would hope that Atlassian would have addressed this apparent Tech Debt in their upcoming Jira 9.0 release. Unfortunately, this does not appear to be the case as of this writing.

Lastly, I want to draw your attention to one line in particular that a reader pointed out to me.

Any feature or capability that is currently on our public roadmap will be delivered and supported even if it is outside the bounds of our aforementioned future investment.

Gosia Kowalska, “The Future of Atlassian Data Center“, 19 Apr 2022

Please go back and read that passage in context. While granted, it does leave the option open to add to that roadmap in the future; it sounds like a final statement. I read such a passage: “We will finish with the roadmap we have now, even if it costs more than expected, but don’t expect more beyond that.” Again, as I’ve stated, I could be wrong, and I sincerely hope I am. But it does have an ominous tone to it. 

Why the Cloud hate?

I don’t hate Cloud. I think it’s a fantastic option for start-ups and is excellent for companies large and small that can afford to go over there. But I also know that there are many reasons an organization would choose to stay on Premise with their critical instances. Some like the security of having a system with all their secrets behind a firewall. Some have regulatory obligations to control their data. I’ve even heard a rumor of a U.S. DoD instance on an ultra-secured network that is completely isolated from the Internet. 

And I feel Atlassian should fully support their customers wherever they need to be. And as a Data Center customer, it’s honestly felt like Atlassian has recently fallen short for us. Even my users – who are not tuned into the Atlassian Ecosystem like I need to be – notice that they don’t see new features despite recent Jira Upgrades.  

But, I’m just one man. I have a megaphone, but I’m not thinking anything I say will change Atlassian’s mind. And, to be fair, I’m shocked that my first post blew up like that or that Atlassian felt the need to respond to me. So in that sense, I’ve already far exceeded what I hoped would come of all this.  

Aren’t you moving the goal post?

I mean, if I squint, I could see how you could view it as such. But I don’t think so. Let me explain:

My post last week responded to how, from my vantage point, Atlassian seemed to be going out of their way to avoid talking about Data Center. And to be entirely fair, Atlassian has addressed that concern 100%.

Let me repeat that in bold.  Atlassian has addressed the concern I brought up last week. They are now talking publicly about their Data Center products roadmap, which is more than I could have hoped for.

However, in their response, I now have a new concern, which is what I’m expressing here. Will Atlassian also address this concern? Honestly, I doubt it. I think this is baked into their long-term plan. And while they have no plans to end of life or end of support for their Data Center offerings, I still can’t help feeling it’s still the ultimate goal, and the DC Feature Drought is just one part of that goal. 

[Gosia Kowalska]

Thank you all for your comments. It is great to hear that many of you find this post reassuring. We do understand that some of our customers require more time to adopt Atlassian Cloud solutions and we are committed to supporting them on Data Center.

As we have outlined in the blog, our future focus is on the key strategic pillars of performance and scale, security and compliance, and infrastructure and operations. However, we will continue to evaluate and are open to learn and hear about how we can further drive mission-critical value within these areas across all users of our products.

@Rodney Nissen - ReleaseTEAM, @Dave Liao @Ed Gaile _Atlanta_ GA_ , @Mayuresh Sakharape, and any others I would love to better understand which improvements your users are looking for in Data Center products.
Feel free to reach out to me so we can set up a time so that I can learn more about how we can continue to help your teams. 

Please ping me on mkowalska@atlassian.com

Look forward to chatting!
Thank you.

[Phill Fox]

As an attendee at Team 22 I can confirm that a significant proportion of the folks I spoke to were still using the on-premise solutions. Many on server are still not convinced that the business case for Cloud works for them and are considering what the lifespan is of Data Center as they do not want to commit to a solution that will in a short period also be end-of-life. 

At times like the present where the global economy is still struggling to come to terms with recent events making the case for change and more importantly a larger budget allocation (without clear business benefits) is a challenge for most organisations.  

When faced with the difficult decisions about where to go from server organisations are facing tough choices. Do they go to DC accepting the roadmap commitments given here by @Gosia Kowalska ? Or do they go Cloud with the disruption to their current ways of working assuming that this is an option in their circumstances?  Or probably the worst choice from Atlassian perspective do they accept that change costs (time/money) and look at competitor products/offerings?

I hope that Atlassian realises there are a significant number of people on-premise who for a variety of reasons are not able to move to Cloud and that at Team 23 there is a reflection of this in the content provided. Even a single session on the future of Data Center as a standing item for all future events would be a great start. 

This post is a great start to keeping the discussion open about the future of Atlassian on-premise solutions but it is completely drowned out in the tsunami of material about Atlassian Cloud offering. 

[Full disclosure - I work for Adaptavist but the views expressed above are my own.]

[Jamie Echlin _ScriptRunner - The Adaptavist Group_]

> We do understand that some of our customers require more time to adopt Atlassian Cloud solutions 

The suggestion here is it's just a matter of time before all customers "adopt Cloud". Is that the intended meaning?

What sort of time scales do you think customers require? Potentially "forever" or do you have some range in mind?

[Troy Chaplin]

While it's nice to see Atlassian responding to these concerns, this still has not given me any confidence that DC will get anything more than some security and performance improvements moving forward. Nothing in this response has changed my opinion. I still feel that at some point in the coming years Atlassian will announce the end of life of DC and cloud will be the only solution they provide.

A few comments above this one, @Gosia Kowalska states "We do understand that some of our customers require more time to adopt Atlassian Cloud solutions and we are committed to supporting them on Data Center." I feel like @Jamie Echlin _ScriptRunner - The Adaptavist Group_ hit the nail on the head here. This one line here leads me to wonder if there's still an underlying motive to try and push more and more people over to cloud. Not bringing new features to DC, and not providing anything on the Team 22 streaming side for DC customers is certainly one way to do that.

Another item that's bothered me lately is that regardless of whether or not I remove it, the Cloud Migration plugin always finds a way back into my instance. Some might think that's a small thing, but I look at that as being a little heavy handed in forcing something on me that I do not want. And this plugin has also been the cause of a few upgrade failures for us in the past.

@Gosia Kowalska you asked a few others what they were looking for as far as features go, here's my 2 cents worth:

• Proforma: what's the plan for DC customers here? Can we ever expect better integration between Proforma and insight? I would LOVE to be able to use Insight custom fields in a Proforma form.
• Clarity on the roadmap for companies and plugins that Atlassian acquires. Proforma again servers as a great example. I can find info about the roadmap for proforma for cloud, I even got emailed info about it. As a DC customer I haven't heard anything since it was acquired. I have been stuck on 8.21 for a while as this plugin wasn't compatible. That got changed yesterday, but not with a version release, so anyone else waiting for a new release had to manually go look.
• Work management. I believe it was @Rodney Nissen who also mention that this would be a fantastic addition to the DC world. Why does this have to be a cloud only option?
• Request types: I would love to have the same drop down list of forms in a portal group available in DC. It's a slick way for user to move to a different form when they realize they've clicked through to the wrong one.
• Look and feel. I love navigating around DC, it's easy and straight forward, well organized. I don't feel the same about cloud. What I do love about cloud is that it looks and feels like a modern web app. The web designer in me would love to see a big leap forward in the UI in DC, it doesn't really feel like a modern app.
• Everything else. I don't see why a DC customer can't expect and plan for all the new features that are being introduced to Cloud. I, like many other DC users want to get excited about new features and do more with out instances.

Paying a lot of money for a tool that doesn't improve outside of the standard security and performance won't fly in todays marketplace. Atlassian needs to recognize this, and understand that a lot of admins are probably already looking around at other solutions in the marketplace as a potential fallback plan. And you can certainly bet that the competition is watching this very closely as well. Lack of clarity on the future of DC combined with a major outage and concerns on proper disaster recovery on cloud should be a cause for concerns for all customers and should be top priority for Atlassian.

[Troy Chaplin]

For some reason a previous comment I made has been removed. Can I ask why? I was not rude, I was not disrespectful, I echoed similar concerns from others in the other comments here, I raised a couple of concerns of my own, and I provide a short list of things I would like more clarity about.

Is this not acceptable? Other comments here are not much different than mine, I just wrote more information. It's rather frustrating to try and get involved in the community, take the time to think out a comment, review it and make sure it's well written and informative, only to have it removed.

Update: post was flagged as spam and has been restored

[John Dunkelberg]

@Troy Chaplin I saw the notification for your post and wondered what had happened.  Hopefully some automated moderation error, and the post can be restored?  To my reading, it was very much what @Gosia Kowalska had asked for in the earlier post.

Update: and now the post is back!

[Troy Chaplin]

Thanks @John Dunkelberg, it appears as though my original comment has now been restored. This makes me happy as I want to be a part of the community here, I am still relatively new to the Jira Admin world (about 2 years now), and having my first attempt at having my voice heard get removed was just frustrating.

[Nic Brough -Adaptavist-]

@Troy Chaplin  the automatic spam fitter hit it.  One of the leaders corrected it just before I got there.

[Troy Chaplin]

Thanks @Nic Brough -Adaptavist-, for future reference, was there anything specific that was flagged in what I wrote so I can avoid that on future comments?

[John Dunkelberg]

For our organization (we're in the 5,000 - 10,000 user band), we will continue to evaluate Cloud vs DC.  My colleague just completed an assessment recently, and while he hasn't made formal presentation of his finding, the short version is that Cloud is still not ready for our usage.  I'll be glad to share more once I can accurately represent his findings, but two things are at the top of my list:

• Customization.  We use ScriptRunner and other tools, including internally developed tools, so that Jira can better support our internal processes.  We have to be confident that this functionality can be done in cloud.  For our internal tools, my understanding is that they will have to be re-written from the ground up for Cloud, so we would need to see a increased benefit that matches that cost.  I would speculate this is also why we see that many of our valued third party plug-ins either do not yet exist in Cloud, or are not yet at parity for functionality offered in their Data Center versions.
• User experience continuity.  With a previous SaaS tool we had considerable pains when a new version was pushed to our environment that caused changes in our user experience, ranging from broken workflows to a simple need to update internal documentation and communicate to the user base.  After a lot of feedback to that vendor, they established sandbox environments so that we (and other customers) would see changes weeks before they were pushed to us, so we had time to provide feedback on workflow breaking bugs as well as communicate to our users.  We left that vendor to come to Atlassian Jira, and have enjoyed that we control the change flowing to our users.  It is not clear to me that Atlassian's Cloud offerings can even match that former vendor's approach.

[Jimmy Seddon]

Hi @John Dunkelberg,

My company is smaller than your user base (about 2000 users), and we have only been on Cloud since we started so we are in that state of "you don't know what you never had".  I completely understand where you are coming from, and I agree with your first point completely.  The ability to customize the Cloud products is not possible to the same extent as the Server/DC products at this time.  I believe Atlassian is working on that, but they have quite a ways to go.

I did want to share that they do have something to help with point #2 as it relates to sandboxes and "more" control over changes being released to your Cloud instance.  Please note that I believe you need to be on the premium subscription plan to have access to these capabilities but thought it was worth sharing:

https://support.atlassian.com/organization-administration/docs/manage-product-release-tracks/

Atlassian offers a Sandboxes and "Release Tracks" which allows you to optionally delay the roll out of changes to your Cloud environment to give you time to prepare your users for the upcoming changes.

I think you have the correct approach of evaluating the cost (not just financial) of moving to Cloud vs what you have on Data Center and sharing that feedback with Atlassian.

[John Dunkelberg]

Thanks @Jimmy Seddon - I just read over that doc.  The Production bundled / Sandbox Preview would allow us to see ahead of time "Visible changes that require people to adjust the way they are working, but do not cause significant impact to the organization."  That seems to leave open the question of seeing changes that would "cause significant impact to the organization" - the doc implies that such changes as "the new issue experience" and "the fabric editor" would use the continuous model?  That seems unexpected.  So we're shielded from small changes like moving a button, but not shielded from major disruptive change?  I can only guess that I'm not understanding this correctly?

Edit: I left feedback on that page, as I realized I shouldn't count on a comment in the forums as a means to communicate to the relevant Atlassian team. :)

[Jimmy Seddon]

I'll be honest @John Dunkelberg we are not making use of the Release Tracks feature.  I presented it as an option and our management team wasn't interested.

I believe what they are talking about when they mention things like the "new issue experience".  This was something that started rolling out over the course of a year and users had the option to "opt-out" in their profile settings.  So while you wouldn't be able to control the feature being available to users, you would need to instruct them in how to turn it off if they wanted to keep using the older view.

Then Atlassian make very intentional communication when the completion of the feature was going to be well in advance of it being rolled out so you had time to prepare those that chose to continue with the old view before they were forced to the new one.

I hope that helps!

[Aaron Matthys]

I echo @Rodney Nissen - ReleaseTEAM 's thoughts posted on https://thejiraguy.com. 
Suspect Atlassian will only release new products and features to Cloud and leave Data Center behind. All in an effort to push more customer to their Cloud offerings. 
Currently, the Cloud's price and migration path is not appealing. 

[Lukasz]

Any feature or capability that is currently on our public roadmap will be delivered and supported even if it is outside the bounds of our aforementioned future investment.

I think it`s hard to understand it in any other way than how @Rodney Nissen - ReleaseTEAM rephrased this statement:

“We will finish with the roadmap we have now, even if it costs more than expected, but don’t expect more beyond that.”

I would love to read the Atlassian comment on that. 

[Nic Brough -Adaptavist-]

@Troy Chaplin - I have little knowledge of how the spam filters work, there's no obvious triggers in your postings - links to external URLs (especially ones of dubious provenance), you've not posted before, junk messages filled with semi-english text and special characters, etc.  Sometimes swathes of log trigger them.  There's also a handful of words that trigger a different behaviour - the posts are put straight into moderation, rather than the spam bucket.