clustering = ON for a single-node Data Center; what difference does it make?

[Darren]

I've recently inherited a Jira Data Center 8.22.2 instance.

I'm unfamiliar with Data Center / clustering : my last Jira Admin 'hands on' was Jira Server v6 (i.e. just as Data Center was coming out) so this is me being upfront that I'm a novice with Data Center, clustering, blah.

Going through  > System > System info just now I noticed

Clustering

ON

even though this is a single Application Server install (so a bit like old Jira Server)

A quick question to the community:  does this matter?  What impact is this likely to have for a single Application Node cluster, if anything?  What does this actually mean for a single-node setup?

[Andy Rusnak · Accepted solution]

Hi Arnold,

Thanks for posting to the Atlassian Community!

The installation you are running now, as I understand based on your post, is basically just a single cluster Data Center install. There shouldn't be any major impact to having that status set to on, but there may be some minor overhead accrued and some files may be stored in a different location from the standard server install you might be used to. The clustered environment uses a shared home directory, for example, for files that need to be shared across nodes.

In running a single cluster Data Center install however you won't be able to take advantage of some functionalities available to a multi-node cluster install such as zero downtime upgrades, server redundancy in the event of a server failure and scalability.

The Adding or Remove Data Center Nodes article discusses turning off clustering entirely should you want to go that route. Towards the bottom of this page you will see the instructions for "Moving to non-clustered Data Center". If you never had multiple nodes in your solution, you may not need to make any changes to load balancers as the document suggests, but each environment is different so it should be investigated.

It would be strongly recommended if you are going to make any changes like this to your solution you do so in a non-production instance or ensure there is a backup strategy in the event the changes would need to be rolled back.

Please let me know if you have any additional questions related to this and I would be happy discuss further.

Best,

Andy

[Darren]

Thanks Andy!

Actually I have a vision to move us to a clustered Data Center environment, for the reasons you mentioned in your reply.  This is a project we're just kicking off this week.

Thank you for your response though : it's given me a little more insight.

[Andy Rusnak]

My pleasure Arnold!  Happy to help, and if you have any additional questions as you start looking into clustering please feel free to let us know.  We are always here to help. 

Best, 

Andy