IBM’s Arvind Krishna: Linux Will Fuel Next-Gen Cloud Applications

29

The next generation of applications will be smarter — with built-in data, mobile and social capabilities — and it will be built on private clouds that run on Linux servers, says Arvind Krishna, general manager of development and manufacturing in the Systems & Technology Group at IBM. What’s still unclear, however, is the business model for delivering these future applications, he says.

Arvind Krishna, general manager of development and manufacturing in the Systems & Technology Group at IBM. Krishna will present his vision of how such cloud applications will be delivered in his keynote talk at LinuxCon and Cloud Open in New Orleans, Sept. 16-18, 2013. Here, he discusses the role of Linux and open source software in the cloud, what a next generation workload will look like, the biggest challenges facing the cloud today and potential solutions.

Can you give us a preview of your keynote?

My presentation is entitled, “Linux, Cloud and Next Generation Workloads.” It is about the changes in our systems that are driving a new generation of applications and delivery methods.  This next generation is built upon open source and open standards.  IBM has been working with Linux for many years to help accelerate collaboration and innovation. And now, with our recent OpenPOWER announcement, the members of this new consortium can use POWER microprocessor technology to innovate even further.

What is a “next generation workload,” by your definition?

When I say next generation workloads, I’m referring to a new class of application solution that combines rapid access to massive amounts of data with advanced compute capabilities, to deliver simple, timely insights about complex systems.  This involves any kind of big data, analytics, cognitive, mobile or social computing.  The demands and opportunities presented in each of these computing areas are growing exponentially.  These workloads are well-defined inside IBM under the SOE (system operating environment) or SDE (software defined environments) projects.  They are things like V8, Node.js, MongoDB, etc. that are new, emerging applications in the Linux space.  

What is the role of Linux and open source in managing such workloads in the cloud?

It is very influential already and its significance will only grow more dramatically in the near future.  IDC Research Director Matthew Oostveen was recently quoted as saying, “Private cloud adoption will drive converged systems adoption, which will in turn drive uptake of Linux-based servers.”  I believe that Linux will both manage workloads and be the engine for the next generation of applications.  That is where all the new applications are being developed (r.e. the SDE/SOE stacks).  

What is the biggest challenge facing the open cloud today?

The open technologies supporting cloud have been quickly maturing in a highly collaborative fashion.  This is evident with technologies such as OpenStack which have quickly grown in maturity and are being used in real world environments.  The more significant challenge facing the open cloud today, besides the technology, is the implications associated with the delivery of infrastructure, platform, software and pricing models.  The changes to the business model with considerations for public vs. private, pricing, billing, etc. often present a more significant challenge when introducing cloud delivery models.

How do we overcome this problem and what’s the role of Linux and open source in meeting that challenge?

Our partnership with Linux developers is a very high priority for us and you can expect us the “walk the talk” here.  I’ll explain this further during my keynote.  One cannot dispute that Linux is the innovation incubator where all the great ideas are being explored whether it’s at universities, with entrepreneurs, or in Internet data centers.      

How is IBM advancing the open source cloud going forward?

IBM has been actively participating in the OpenStackcommunity across our product development labs assisting with everything from blueprints, test, code reviews, code contributions and bug fixes.  In addition, we’ve been extending OpenStackto cover IBM’s broad server products to include Power Systems, System z and storage products.  IBM is not alone in this effort, as there are 60+ companies contributing code to the current development release of OpenStack.