We started our production from OpenStack Havana while it's time for Kilo. There will be a new OpenStack release with lots of bugfix and new features every 6 months. It's really a challenge for Enterprise IT to define its upgrade strategy and perform a successful upgrade. In this session, we will introduce our upgrade strategy from an Enterprise IT point of view by comparing different upgrading strategies. Meanwhile, we'll also talk about the reason why we need to perform the upgrade based on business and technique perspective.
Even though you make the decision to upgrade, the road is hard and steep. There will be many unexpected issues if you don't have a mature test plan and enough rehearsals. The roll back plan is also very significant to avoid tragedies. In addition, automation or leveraging existing CI system is the key to reduce upgrade duration and avoid some manual mistakes. And lastly, how do you verify the cluster after the whole upgrade process? We'll share the best practices of performing a rolling upgrade and the expectation of future improvement of Openstack community.