Budgeting: Atlassian Data Center vs Server
August 2, 2018
Choosing between the Atlassian Data Center solution and Server solution is one of the first choices you need to make when deciding on an appropriate on premise setup.
Atlassian’s software development and project management tools are among the most versatile on the market. But before you use Confluence for team collaboration, Bitbucket for managing source code, or Jira Service Desk for handling customer requests, it’s important to establish which kind of Atlassian deployment is most suited to your organisation.
For each one of its major products, Atlassian provides two different deployment options for on-premise installation: a Server edition and a Data Center edition. Data Center is optimised for large user bases and where the application is mission critical; Server is for smaller userbases and less important uses. Data Center is more expensive but permits multiple application servers (clustering), whereas Server permits only one application server.
Why might you need to use Atlassian Data Center over Atlassian Server?
An earlier blog covered some of the general benefits of Atlassian Data Center and Server – but while various factors will influence your decision, downtime (and your organisation’s ability to handle it) may be the single most important.
Capacity and reliability should therefore be front of mind. A single node (i.e. application server) with internalised stores might suit a smaller organisation: the single node can handle the load and occasional downtime can be tolerated.
In contrast, a large business has more at stake: an hour of Jira downtime, for example, could have significant financial and operational consequences. The larger your user base, the more capacity you will need to ensure adequate performance, and the costlier an outage becomes.
Atlassian Data Center costs more than Atlassian Server, but it is important to weigh this against the cost of downtime and performance degradation. As a team becomes bigger, the cost of an outage grows, until at a certain size it exceeds the additional cost of Data Center.
The chart below illustrates this. Making certain assumptions about the cost of idle users, the red line shows the cost of one hour of outage against the size of the user base. The assumptions exclude any costs to reputation, opportunity costs of lost output, and dislocation costs of users taking time to switch to other activities and back. The green line shows the cost of Data Center against the size of the user base. The blue and orange lines show the cost of Server.
Each organisation’s decision about whether to use Data Center will be different. The decision will be based not only on a careful calculation of the cost of an outage, including reputational risks, lost sales etc., but also on factors such as performance and the time taken for upgrades (Data Center permits zero-downtime upgrades).
We can help you decide which option will be most effective for your business’ particular circumstances. To discuss Bitbucket Data Center, Confluence Data Center, and Jira Service Desk Data Center deployment options – or their server equivalents – talk to our specialists today.
Alternatively, for more detail on the cost of outages and to find out how many users we recommend use the Server products, download our eBook now: The Enterprise Guide to Atlassian Data Center Applications