Scaling Jira Software and Confluence Cloud to 20K users per instance

[#1]

Hi everyone! As summer kicks off in most parts of the world, we have some hot news to share about Atlassian's cloud scale journey. Today, we are excited to announce the general availability of 20,000 users per instance for Jira Software Cloud and Confluence Cloud.

To offer maximum licensing flexibility, we are also releasing new price tiers between 10,000 - 20,000 users across our Standard, Premium and Enterprise (annual only) plans. You can check these out on our pricing pages for Jira Software and Confluence. To support the new user tiers, Marketplace partners will also have the option to set pricing for their cloud apps between 10,000 - 20,000 users per instance. While this is a significant milestone our work is certainly not done by any means. We continue to allocate significant engineering resources to support your scale needs and aim to raise the per instance user limit to 35,000 next year.

We recognize that you leverage Atlassian products for some of your mission-critical projects. So to back up our scale investments, we continue to invest in offering the highest standards of reliability. This includes financially backed 99.95 percent uptime SLA with our Enterprise plan and 99.9 percent uptime SLA with our Premium plan. Over the last year, we have delivered significant performance improvements to our Jira Software Cloud and Confluence Cloud products. In our latest cloud performance blog, you can learn more about how we improved load times for Jira Dashboards (1.9x faster), Backlogs (1.4x faster), Roadmaps (1.5x faster) and Confluence pages (2x faster). Also don’t forget to check out our public roadmap for upcoming performance investments.

As a reminder, with the launch of Cloud Enterprise - our highest offering for Jira Software Cloud and Confluence Cloud - you can now take advantage of multiple instances (up to 20,000 users each) to support autonomy, security or customization needs. Some of the common scenarios we’ve heard from customers for setting up multiple instances are:

• offer autonomy to acquired entities or independent business units

• support advanced security needs such as data residency for certain teams

• support collaboration with external partners, suppliers

• customize based on Marketplace apps, project configurations and more

To learn more about whether you could benefit from a multi-instance environment, please review our support article on - how many instances of a product does my organization need. Also a useful aspect about Cloud Enterprise is you can leverage centralized licensing - pay for a user once and grant access to unlimited instances! And you get to maintain global control and oversight with ability to centrally manage user access and permissions, get usage insights as well as apply security policies across instances. Check out our latest demo on enterprise administration to learn more!

As always, a huge thank you for helping shape our cloud products. Please don’t hesitate to share your thoughts on the latest scale and performance investments and how we can continue improving your experience in the cloud!

[Dave Liao]

@Aditya Ghule - 20,000 kudos points to the team! Being able to serve that many users reliably is quite a milestone for Jira Cloud and Confluence Cloud - I know a few companies that were waiting for this. 💜

To the moon - for users! 😅 35,000 will be quite a feat.

[Aditya Ghule]

@Dave Liao - Appreciate the positive endorsement! Look forward to sharing the 35K news  

[saurabh.sharma4]

Kudos to the Atlassian team for making significant improvements to Cloud offerings. Looking forward to more exciting features and data security controls for Enterprise Customers.

[Dane Kantner]

how do we scale API limits imposed by Atlassian in the cloud w/ jira, and do any of these plans have different API limits?

[Rutger Coolen (Atlassian)]

Great question @Dane Kantner!

REST API rate limits in Jira Cloud are used to ensure that services are reliable and responsive for customers. These limits are typically higher for larger customers than for smaller ones. But, rate limiting is not used to differentiate levels of service for different plans or apps.

Note also that the actual limits are not published because the logic is evolving continuously to maximize reliability and performance for customers.

We ask that app developers follow the guidelines for how to deal with rate limits, see: https://developer.atlassian.com/cloud/jira/platform/rate-limiting/ This way we can manage peaks in load, while maintaining a reliable experience for customers.

I hope that answers your question, thanks!

[Ollie Guan]

This is good news. Cloud can provide good support for increasingly large-scale enterprise users.

However, we still expect Atlassian to continue to provide better support for Data Center. After all, this large-scale company needs to maintain localized deployment based on various factors.

[Aditya Ghule]

@Ollie Guan - Thanks for the feedback. And yes, we continue to invest in our Data Center solution as we understand this may be the most optimal deployment for certain scenarios or could even be the stepping stone before moving to cloud. You can check out our public roadmap for Data Center for additional details. 

[Gaurav Agrawal]

@Aditya Ghule Great news.

One question - Do we have any upper limit of supported groups? I know a use case where organisation maintain fine-grained permission at project/space level using groups. In that case, groups count would be much higher (assume 20k+ groups) to support 20k users.

[Aditya Ghule]

@Gaurav Agrawal - We typically don't define limits for groups but assume these would be less than the number of users in most scenarios simply because groups are formed to manage a collection of users with similar needs around access, permissions and more. So it would be interesting to hear more about your use case around creating 20K+ groups to support 20K users and how they are managed from an administration perspective. Also tagging my PM colleague @Rutger Coolen (Atlassian) to add any perspective he might have here.

[Gaurav Agrawal]

@Aditya Ghule Thanks for the response. We have internal tool to manage user access via groups. 

Lets consider Jira, and we have around 4k projects in Jira. Each project will have 3 role groups (project admin, developers, and users) so in this case we have around 12k (4k*3) groups to manage access for users. Same goes for Confluence.. which will turn into more than 20k groups.

This is to have fine grained access at the project/ space level.

[Jack Brickey]

While I do not expect to approach these numbers in my current business (never say never) it is reassuring to see Atlassian dedicating efforts to have cloud truly meet the needs of large businesses.

[Rohit Jagdale]

Hi Team,

Thank you for this new update and efforts you took for the same.

[Vikrant Yadav]

Wow!That's great news. Now we can migrate our JIRA Server 15k users to JIRA Cloud :)

Thanks!