As an Atlassian Marketplace Vendor, my company attempts to pass the Atlassian Data Center compatibility requirements documented here, and we’ve stumbled across a requirement that we fruitlessly attempt to interpret.
Detail any user testing steps you take to ensure features work in standard Data Center multi node deployment (such as our AWS deployments).
Does this mean that we must deploy on AWS?
Where can I find documentation on “standard” Data Center deployments?
How is AWS deployment different from a self-hosted data center in a controlled environment?
As a partner that supports an extremely large data center deployment (130k+ Crowd users and 3k* Concurrent users, 2mil+ Issues) we interpret it as you have tested that your app/add-on can be distributed across an n4+1 node data center deployment and that it scales with concurrent users.
We have identified that there are apps (add-ons) that work great as server and as a data center node = 1 instance but once you start scaling out the nodes to n4+1 the app starts to affect the replication index between Jira Nodes and or Confluence/Bitbucket nodes.
A case in point is the Cleaner for Jira. Their filters cleaner had an impact on performance with large data center instances, thus they rolled back the feature so that they could scale it. I think that Atlassian is trying to get App developers like your company to consider, large scale deployments as that is where the industry is going.