How do Jira Cloud admins handle performance issues as instances grow?

[Sharad Dadarao Chate]

Hello Community,

I am a Jira Cloud admin working with growing projects and users.
As the instance scales, I would like to understand the most common performance bottlenecks in large Jira Cloud environments and best practices admins use to avoid them.
Any real-world tips or lessons learned would be appreciated.
Thank you.

[John Funk]

Hi Sharad,

Use as many shared configurations as you can - meaning Schemes. This also means you would create Company-managed spaces and not Team-managed spaces. Team-managed spaces also increase the number of custom fields and configurations for the system. 

Create Company-managed spaces based on existing spaces as much as possible. This will use the shared configurations in the first item. 

Use generic custom field names as much as possible when you create them. And then share the fields across projects. Also add a Context for each Custom Field as much as possible - linking to either one or more spaces. Global Contexts should be kept to a minimum and only used when necessary.

Keep down the number of third party apps as much as possible, as well as integrations. 

If you do these things, your instance should run fine. 

[Hans Polder _Devoteam_]

Hi Sharad,

It might be only partly related, but I would like to point out that, in my experience, it's 'more users, more problems'. Of course, you can have perfect documentation for your users in Confluence, but with more users, you will obviously have a higher chance of hitting edge cases, and more users that won't bother to read  your documentation, and want to ask for help. So, make sure you have a solid documentation & support structure in place to cope with that. 

For example: When I worked with a company that merged a couple of times and through those mergers, the Jira/Confluence instance grew from ~1000 to ~8000, there were noticeably more users complaining about Jira/Confluence Cloud being slow. It might feel like "shoot, the entire instance is slow", or it could very well be that simply the rise in user count simply increases the absolute number of users that report issues. In most cases it always turned out to be an issue on user side (slow internet connection, working via Citrix or other VPN connections, other local issues). 

So: It might feel that you have more issues 'because Atlassian didn't scale correctly on their backend', but it could very well be that you simply have more issues because you have more users (therefore increasing the odds of them reporting issues) ;-)

[Marc -Devoteam-]

Hi @Sharad Dadarao Chate 

Resources are granted as per subscription.

Main causes fro bad performance are :

• number of custom fields
• number of permission schemes
• bad performing automation rules
• integrations via API, what are there permissions and scope

[Trudy Claspill]

I would add to this list that use of third party apps can contribute to poor performance as the volume of data grows, depending on the construction of the third party app.