Presentation: Tweet"Continuous deployment for a billion dollar order system"
Continuous deployment is causing organizations to rethink how they build and release software. Atlassian is adopting this model throughout the company, but not all teams have the same challenges to overcome in doing so. Steve Smith lead the team that converted the company's critical order-processing system from a monolithic, single-server application to a continuously-deployed, high-availablity platform. Along the way there were a lot of practical and organizational issues that needed to be addressed in adopting this development model; in this presentation he shares some of the experiences and lessons of doing so. This talk covers continuous deployment from a number of different angles; high-availability requirements, development processes (in particular git-based branching workflow), practical delivery technologies (including analysis of the trade-offs), and organisational considerations and bottlenecks (e.g. SOX/PCI compliance issues).
Download slides