Will switching to JIRA enterprise, increase the number of issues that can be handled?

[Arun_Thundyill_Saseendran]

I am in the process of designing an architecture for JIRA Enterprise (Data Center set up). I understand that increasing the number of nodes, increases the number of simultaneous users and also the reduces the response time.

However, I see that there is no reference to whether the number of issues that JIRA can hold increases when we switch to enterprise. If yes, is it linear. Say if with two notes I can hold 2 Million issues, with 3 will be able to have 3 million issues. 

Is there any documentation regarding this? The only documentation I saw is related to 1 million issue performance test, but that was on a single JIRA instance and not enterprise.

Thanks

Arun

[Otto · Accepted solution]

Data Center is focused on improving essentially two aspects of JIRA;

• Ability to maintain performance with increasing loads: JIRA. Data Center has multiple nodes onto which you can spread your concurrent load to avoid maxing out any single node. This can help avoid poor response time during peak load periods.
• High availability: Should one of the nodes experience an outage, users can be failed over onto another node. With Data Center users won't be unproductive due to losing their ability to manage their issues.

To specifically address your question:

Data Center allows you to have several nodes behind a load balancer. You can think of each node as being a JIRA server. So if a JIRA Server has 1M issues then it will take x time for a certain action (eg. user view issue page request). There will be virtually no difference in the amount of time it takes for this one user to see her page whether she is routed to node 1, 2, 3, 4, etc in the cluster (assuming nodes have the same load and hardware resources/config). So Data Center will not improve your ability to scale JIRA variables such as number of workflows, custom fields, issues, projects, etc.

As issue (and other important variables in JIRA growth continues) you can expect to see changes in response times. If you haven't seen it already https://confluence.atlassian.com/display/ENTERPRISE/Scaling+JIRA can give you some insight into how these variable influence response times. As you can see we've worked hard to continue to push out our scaling capabilities even on the server deployment option of JIRA. In JIRA 6.4 we've made substantial improvements in performance at scale and I can't recommend enough that customers take advantage of this performance improvement .

So in summary we continue to prioritise the ability to scale JIRA regardless of whether a customer chooses to deploy on Server or Data Center. Data Center will continue to be our recommendation for customers for whom;

1. JIRA is business Crititcal
2. There are performance challenges due to peak times when there is heavy load on JIRA
3. Maintaining availability and minimising outages is a key requirement.

Hope that helps,

Otto Ruettinger, JIRA Principal Product Manager

[Arun_Thundyill_Saseendran]

Hi @Otto Ruettinger Thanks for taking time and giving a detailed explanation. It was very helpful. A request to the JIRA team is, please work on increasing the capacity to hold issues. This is one thing that all of us want. A solution in which, if I have a cluster of 3 JIRA instances, then my the capability to hold issues should increase linearly. It would be great if this happens. However, from our end we are also looking into ways for making this happen. Will post if we are able to find do something. Thanks @Arun Thundyill Saseendran

[Otto]

Performance at scale is still number one on our list of roadmap priorities. :)

[MattS]

It all depends what will limit your use of JIRA at that scale. If it's concurrent users and response time the data center will help. If it's the size of the Lucene index then maybe not. Id recommend testing to confirm 

[Arun_Thundyill_Saseendran]

Thanks @Matt Doar [ServiceRocket]. My requirement is to increase the size of the index. Thanks for you post.