Austin
April 25-29, 2016

Event Details


Moving a Running OpenStack Cloud to a New Data Center

The data center is all out of space and you’ve been told that you need to move to the new building. How do you move a running OpenStack cloud to a new building? How will you handle moving to a new IP space on new switches? What happens to running instances and existing floating IP addresses? How do you minimize outages? What order do you move the servers in? What impact did this have on customers? And also can you come up with a new network architecture to move onto?

After months of planning and coordination with outside teams and vendors, we physically moved all our servers to a new data center in under two weeks with no API outage.

Our talk will answer all the questions and explain the pitfalls and successes we found when moving our production OpenStack cloud into a new data center. We will cover API services, compute services, networking, and storage.


What can I expect to learn?

Attendees should come away with a checklist of things to consider when faced with similar challenges, whether a full move or simply an expansion.

Tuesday, April 26, 2:00pm-2:40pm
Difficulty Level: Beginner
Principle Software Engineer
Matt works at Accenture helping clients transform IT infrastructure operations with cloud and Devops. Previously, Matt was a Principal Software Engineer at Charter Communications and used devops to deliver self-service IT infrastructure and services using OpenStack. Matt's focus was primarily on configuring and deploying OpenStack using tools like puppet and ansible. A fan of open source... FULL PROFILE
Time Warner Cable
Craig is a Lead Software Engineer at Time Warner Cable and uses devops to deliver self-service IT infrastructure and services using OpenStack.  Craig's is primarily focused on storage services and delivering new storage based services to our customer base.  When not working on delivering new services his primary goal is to enhance and add more stability to our current... FULL PROFILE