Volkswagen’s “Cloud First” Approach to Infrastructure Decisions
April 25, 2016 Nicole Hemsoth
As automotive companies like Ford begin to consider themselves technology companies, others, including Volkswagen Group are taking a similar route. The company’s new CEO, who took over in September, 2015 has a background in computer science and began his career managing IT department for the Audi division. Under his more technical-tuned guard, IT teams within the company are taking major strides to shore up their infrastructure to support the 1.5 billion Euro investment in R&D for forthcoming electric and connected cars in the near future.
Part of the roadmap for Volkswagen Group includes a shift to OpenStack to manage IT complexity. The technical tangles stem from a global business that does far more than manage design, production, and financing of its twelve automotive brands. Heavy equipment manufacturing, fleet management, insurance, and of course, the daily business operations of well over a half million employees add to the burden as well, especially as the company seeks to centralize and rethink its global infrastructure.
Speaking from the OpenStack Summit this morning in Austin, head of infrastructure for Volkswagen Group, Mario Muller, said there is still a great deal of work to be done to improve workflows, optimize at scale, and standardize infrastructure and applications, all with the goal of speeding IT operations. These have all been moving targets for Muller, who has been managing IT infrastructure in the automotive industry for close to two decades, with a 16 year stint at BW group.
“We are still too slow, we need to get faster, and at the same time, we need to reduce costs. This means we have had to reinvent our IT infrastructure in Volkswagen Group.” As one might imagine given the location of the talk, one path to lead this greater efficiency and speed is to adopt Open Stack.
The company started its OpenStack journey in mid-2015, but even still they have a very heterogeneous environment because they still need the connection of the new cloud into their old ecosystem because that is where the data still resides. The second version of their OpenStack environment, in conjunction with Mirantis, will be ready in mid-2016 and will support customer-facing applications, but also core smart production analytics, 3D rendering, and other more compute-intensive workloads central to the company’s R&D thrust.
On top of the forthcoming second OpenStack IaaS layer, Muller says they will be working with CloudFoundry on the PaaS environment. “We are going to need this development environment for our users to get great integration and toolchains for new deployments,” he said, pointing to the success of the first steps in adoption.
At the core of the new OpenStack efforts on the hardware side is a new 2,000 square meter datacenter, part of which will be reserved for their OpenStack cloud. “Just for this implementation we have 900 cores, which will span around the world to APAC, the Americas, and this is not counting what we have as a hybrid version to work with the cloud providers.
Again, this is not anything major—it’s a rack of servers, but everyone has to start somewhere. And for a company the size and scope of VOlks
“We have made a commitment to going cloud-first with all new deployments,” Muller said. And while the new datacenter is just one new piece of the puzzle, he sounded enthusiastic about the prospects, backed by enterprise OpenStack partner, Mirantis.