Come for the products,
stay for the community

The Atlassian Community can help you and your team get more value out of Atlassian products and practices.

Atlassian Community about banner
4,359,801
Community Members
 
Community Events
168
Community Groups

The Future of Atlassian Data Center

44 comments

Long time admin of Atlassian products here (since 2008). I love Atlassian and she used to love me too. But lately, it feels like she's just not that into me. Our company has many instances of Atlassian products; the largest data center installations are Jira (15k users), Confluence (40k users), and BitBucket (8k users). We're in the financial services sector and have no plans to adopt Atlassian's cloud offering.

First, I want to say that I'm glad to see Atlassian put out a statement re-affirming their commitment to the data center offerings. It's not enough though. The imbalance between Atlassian's promotion of the cloud vs. the on-premise offerings is just too great. People naturally come to the conclusion that the data center offerings are part of what's being discontinued and I regularly get questions from executive management and end users about this.

Similarly, the pace of innovation for Atlassian's cloud-based offerings vs. on-premise speaks louder than any blog post. The data center improvements in performance and scale are nice (I still remember when 200k issues in Jira was the max recommended per instance), but this won't be enough to keep us in the long run. Likewise, the improvements in audit trails, monitoring, and system administration are all good (it's taken years and they still aren't as complete as they should be, in my opinion). But again, it's not enough. In my opinion, Atlassian needs to deliver features and integration on the data center side comparable with their cloud offering because that's what their competitors are doing. I'm not going to name names here, but others are keeping rough parity between cloud and on-premise versions despite disparate code bases and are re-architecting discreet applications so they become services feeding a common integrated UI. It seems to me that delivery of a comprehensive integrated solution from Atlassian is possible... it's just a matter of commitment (which is currently lacking).

For what it's worth, our company abandoned Bamboo years ago because it fell too far behind and in the past year, we've moved 10k users out of Jira as well. We remain committed to Atlassian products for the time being, but the relationship isn't really headed in the right direction at this point.

Like # people like this

Good to hear from you on this subject @Dave Thomas 

Like Dave Thomas likes this

Once again, thank you everyone for your comments and those of you that have directly reached out. Thank you @Rodney Nissen , @Troy Chaplin, and @John Dunkelberg for sharing some specific examples of investments that you would like to see delivered with our Data Center products. This is extremely important for my team and I know they are watching all the comments to include your feedback in their future considerations.

I’ll try to address as many of your questions and concerns as I can but I would really like to continue this conversation via a phone/zoom discussion as I think it could be more beneficial and we could have a better flowing conversation.

Let me first respond to Rodney’s interpretation of this statement:

“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.”

What this statement means is that while our future feature investments are focused on performance and scale, security and compliance, and infrastructure, anything that is currently on our roadmap will still be delivered even if it is outside the bounds of these three topics. We will continue to deliver and focus new features and capabilities in the aforementioned areas.

It does not mean as you mentioned “We will finish with the roadmap we have now, even if it costs more than expected, but don’t expect more beyond that”.

Let me give you an example. While it is outside the three focus areas, we will soon be shipping Automation for Jira as part of Jira Software and Jira Service Management. This will come with a number of benefits for your teams, including admins and end-users alike. Although we did pay very close attention to performance and security, this initiative has been primarily focused on helping teams gain efficiency.

Once again just to reiterate, although the bulk of our investment is focused on the three topics that are top of mind for our Enterprise customers, we will deliver on our public roadmap commitments even if they are from outside of this scope.

Our public roadmap is reviewed and updated on a quarterly basis. You can see things shipped and added to it.

I hope this helps clarify any misunderstanding or confusion.

There were also a few comments referring to this statement:

“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."

Atlassian has never been shy about sharing our cloud-focused investment. Much of this investment has been spurred by desire and demand of customers wanting a SaaS offeringWe do believe that Atlassian Cloud will help us fulfill our mission to unleash the potential of every team. We are happy to see large enterprises successfully adopting our Cloud products and we continue to address problems that we know are stopping others from moving to Cloud.

Having said that we understand that Data Center products remain mission-critical for many customers and we do not have any plans to end of life or end of support our Data Center offering.

We continue to invest in Data Center. Please keep an eye on our upcoming roadmap updates to see new expected investments. I would once again like to reiterate that we have identified our three focus areas through consistent feedback and open conversations with our customers. As we have identified these three mission-critical areas we will continue to develop new capabilities to support them.

Thank you everyone for this conversation. I really appreciate all your feedback and comments. I will be more than happy to jump on a call if anyone would like to discuss this further. 

Like # people like this

Thanks for the reply @Gosia Kowalska, if you ever look to setup some continuing conversations via zoom or another method, I would definitely love to be a part of those. I think that's a great idea!

Like Dave Liao likes this

Okay, so maybe it`s a good idea to organize a Q&A meeting for this community?

Like # people like this
Dave Liao Community Leader Apr 22, 2022

@Lukasz Przybylowicz agreed!

An AMA (Ask Me Anything) format would be good. The community can contribute questions and vote up questions they're (we're!) interested in.

Like # people like this

Gosia,

Thank you for your article and taking the time to address everyone's concerns. As you said, Atlassian's customers consider Data Center as Mission Critical. And that may be true. Only time will tell.  But for those of us who have been around since the beginning, that statement was also once said about the onPrem Server Version when Cloud hit the streets.

This is especially important for those organizations needing less than 500 licenses. Surely, Atlassian isn't going to intentionally alienate all of its <500 employee customers that for one reason or another cannot move to the Cloud...and there are plenty of good business reasons not to.

As someone who subscribes to the rule of "Don't complain unless you have a viable solution", I think it best to throw out some ideas my team, and even some of our customers, have had regarding this topic:

  1. Create licensing tiers allowing for a single node Data Center "lite" version that starts at 50-users. It's scalable, the first step is easier, oh, and it's on their servers.  But maybe this has been the plan all along? Make an announcement at the 11th hour, January 1st-ish, 2024, that server customers can make a lateral move that's not as costly?  Many marketplace vendors show DC add-ons starting at 50 users.  I bet this solution is easier to implement than most.
    Seriously, don't wait until the 11th hour to rescue the already disenfranchised customers, who will already have left by then. Remember the very first Jira Service Desk licensing scheme? Atlassian waited too long to fix the problem and even ignored its solution partners, beforehand, who said they couldn't sell a product that required a license for the entire company headcount when their Service Desk only had a fraction of the total users.
  2. Speaking of add-ons, going to DC is going to make purchasing add-ons a lot more costly. If the company has less than 500 people but they have to pay for add-ons at that same headcount regardless?  Every purchase will be seen as another reason to not stay with Atlassian. Cloud or otherwise. Just on principle.  Many of our Server and DC customers have asked for years for a way to manage add-on licenses at a lower level than the application's headcount. For example, 100 users prefer SmartDraw while a different group of 100 prefer Gliffy. Why not use groups to manage each add-on's access?  Look, if you can manage add-on licensing using Groups like you did for Jira Service Desk Agents, in the beginning, you can manage other add-ons the same way.
  3. Has Atlassian considered managing licensing using the concurrently in-use license model? For larger organizations, 5k, 10k, 20k, etc.  Meaning, only charge for the average number of licenses in use at any one time all over the world. Surely someone has thought of this one before.
    A company has 1000 employees in North America and 1000 in India, but they have to pay for 2000 licenses even though they only use half that at any one time.  Less than that actually.  Customers pay for 24hr/day, 7days per week, 365 days per year, but typically only use them 8hrs/day 5 days per week, don't they?
    As an example, those people who fly rented airplanes pay by the engine hour using a thing called a Hobbs meter. Ubers and Taxis operate by time of use. And so does AWS for some of its products. 

We're just saying, let's get creative and make it look like we care about the customer's bottom line instead of thinking of ways to drive the customer to your Cloud products.  Because that's what it's looked like for the past 5 years.

Thanks for listening

Like # people like this

Thanks for the update.

Hi @Dave Liao , @Troy Chaplin , @Lukasz Przybylowicz we plan to hold an Ask Me Anything (AMA) session on Data Center in the near future. The exact date and format are still being worked on; we will share the details with the DC Community as soon as we have them. We are really looking forward to all your questions and hope for a lively discussion!

Like # people like this

If you want people to move to the cloud you need to make it a desired end point, not something that seems forced - and migrations need to be easy AND COMPLETE. 

User migration is a nightmare, as is moving over all of the pieces that the tool doesn't do.  Dashboards?  Filters?  Are we supposed to just find all of this and rebuild it from scratch, then change all the owners on the cloud site, one by one?  That alone is a full time job, and enough of a reason to consider Data Center, or another product.

Like # people like this

@Gosia Kowalska 

Your statements:

'Let me give you an example. While it is outside the three focus areas, we will soon be shipping Automation for Jira as part of Jira Software and Jira Service Management.'

This is not am example of Atlassian adding functionality to DC. This Atlassian buying out an add-on and incorporating into their offering.

 

'We are happy to see large enterprises successfully adopting our Cloud products and we continue to address problems that we know are stopping others from moving to Cloud.'

Once again you've missed the point. We are not "waiting" for Atlassian to address problem area's -

WE ARE NOT MOVING TO CLOUD.

We are working in industries (financial, DoD, FDA, etc.) whose regulation and security requirements means moving to cloud will never be an option.

 

By only supporting 'these three mission-critical areas' and providing no new functionality or user capabilities, you've just told all of us we will NOT be getting the support we need from Atlassian and to start looking elsewhere for a new product.

Like # people like this
Will C Rising Star May 05, 2022

Thanks for writing the article, I really do hope that DC gets the attention and development that it deserves.

As a lot of the previous posters have said Cloud is a no go for a lot of companies for various reasons.

I have noticed some developments already rolling out to cloud but not DC, one of the recent ones was https://www.atlassian.com/software/jira/guides/expand-jira/jira-slack-integration#get-started

being able to create a Jira ticket from Slack, which means that we have to build our own integration or go to a 3rd party, which when DC costs so much already is disappointing.

Like # people like this

This post is extremely reassuring that Data Center is sticking around. This has been a concern of mine since the discontinuation of server products. Thank you for confirming this.

However, as others have already stated here, no statement made here commits to Data Center in the long term. It simply commits to completing the existing roadmap.

Additionally, every statement seems to indicate that Atlassian will find a route to get everyone to the cloud. That simply will not happen. I will not re-hash all of the reasons here, as I have stated them in the past, and others in these comments have said them better than myself.

I can only hope that Atlassian sticks to its word here.

Like # people like this

Prior to Atlassian's acquisition of Code Barrel, Automation for Jira for DC was released every two weeks, with features pretty much on par with Cloud.

I've found it puzzling that Automation has gotten so far behind on DC. It surely is the easiest way to ensure that customers on DC build with the least future barriers to move to Cloud.

Yet, where is user smart values on DC? How about processing 4XX web request replies? Both of these have existed on Cloud for more than 6 months.

Absent DC-Cloud parity on Automation, more customers need more app-based Scripting and face further cost and resistance to take the leap.

Likewise, it would be really good to get Confluence Automation on DC, again easing progression for those who choose to go to Cloud.

Lastly, no-code platforms have limits. Even platforms like Zapier drop to code in your choice of language in a Functions-as-a-service system such as AWS Lambda. I would think Atlassian's value proposition would be stronger supporting an official code container accepting webhook REST calls with optional callbacks for long-running asynchronous processing. Given the commitment to Kubernetes this could be reasonable.

Like # people like this
Dave Liao Community Leader May 25, 2022

To those still watching this thread, ask your questions in the AMA (Ask Me Anything) thread that Atlassian started, and is conducting today:

https://community.atlassian.com/t5/Data-Center-articles/Future-of-Data-Center-Ask-Me-Anything-AMA/ba-p/2023869

@Martin Cleaver especially - I have the same question as you! 

Like # people like this
Like Dave Liao likes this
Dave Liao Community Leader May 25, 2022

@Martin Cleaver - you’re learning! 😂 Glad you saw my comment, I was about to repost your original comment (with appropriate attribution). 💙

Like Martin Cleaver likes this

Hello,

I'm techlead and I've been working with Atlassian tools for 3 years.
Like you, I wonder what will happen to data center Atlassian tools in the next few years.

In my opinion, there is a double movement on the one hand to highlight the cloud solution while improving it (some integrators have completely followed suit)
On the other hand, neglecting the DC solution (no dev, no visibility, no long-term certainty)
This will end up hurting the DC solution.

Explanations:

1/ at the time of this writing, there are a lot of new features for the cloud in the roadmap and very few for DC.
Cloud roadmap 
DC roadmap 

2/ Atlassian buys apps, it's good but adds features in the cloud and not in DC.
Example: "Access Insight fields through Forms" is planned by the end of the year on Cloud but is not even considered on DC.
See in cloud roadmap 
See in DC roadmap 

I particularly draw your attention to this point.
How is it possible to buy Insight and Proforma and not even get those apps to work together on DC?


3/ When certain developments seem essential for automation, again purchased apps (Build and Deploy' triggers), they are added to the cloud but not to DC.
Better, Atlassian explains "It is not available natively in Jira Server / Data Center but there is now an app in the marketplace from an Atlassian partner" (here)
This is exactly what we want to avoid: paying for new apps because Atlassian does not develop updates.
Furthermore :
- it's risky, there won't always be a partner publisher to develop instead of Atlassian
Example: "4 New Form Automation Components for Jira Service Management": developed by Atlassian on Cloud and nothing on DC (here)
- this at a cost

4/ Atlassian gives no certainty about DC's long-term future
Gosia says "To be clear, we have no plans to end of life or end of support for our Data Center offering." but what will it be in 6 months in a year?
Not providing visibility, even for the next 5 years, discourages new customers to say the least and above all challenges many customers using DC.

Conclusion
As Gosia puts it, Atlassian "have not been shy about sharing [its] cloud-focused investment"
In itself, I don't see the problem but if the effort is focused on the Cloud, in particular to catch up with DC, new functionalities are added in the Cloud and it is not clear why they are not added in DC.
I can very well understand Atlassian's efforts to bring the Cloud solution to the level of the DC solution, but the DC solution will only last if Atlassian remains objective on what can not be developed on DC and what must be. On the one hand to satisfy customers and on the other hand to keep DC competitive...without relying too much on partner publishers.

Best regards,
Admin Dewet

Like # people like this

Atlassian may want to become a cloud-first company, but the people paying for the software aren't all going to feel the same way.

It shouldn't be my way or the highway.

What happened to 'don't (aggressively love) the customer'?

Like # people like this

Comment

Log in or Sign up to comment
TAGS

Atlassian Community Events