Quick! When was the last time you thought “network” when you heard “cloud computing“? How often have you found yourself working out exactly how you can best utilize network resources in your cloud applications? Probably never, as to date the network hasn’t registered on most peoples’ cloud radars.
This is understandable, of course, as the early cloud efforts try to push the entire concept of the network into a simple “bandwidth” bucket. However, is it right? Should the network just play dumb and let all of the intelligence originate at the endpoints?
First, an introduction. My name is James Urquhart, and I joined Cisco in the middle of November as their new Market Manager for the cloud computing and virtualization markets. I’m humbled to be working with such a deep team at the center of one of the most successful technology companies of all time, and I hope to use my extensive background in enterprise IT, distributed systems and cloud/utility computing to push both Cisco and its markets to create and sell innovative cloud and virtualization technologies.
For a detailed look at my thinking around cloud computing for the last two years, check out my blog, The Wisdom of Clouds. With almost 1000 subscribers, I’ve been lucky enough to engage in a conversation with many of the best and brightest in the cloud field for quite some time now. I hope you’ll join the conversation both there and here at the Data Center Networks blog, where I’ll be a regular contributor moving forward.
Now, back to the subject at hand.
Greg Ness, one of my favorite bloggers about the role of the network in the next generation of enterprise computing, wrote a post this weekend that captures the good, bad, and ugly of dynamic networks and network automation in today’s enterprise IT departments. At its core, the concept of automation in general in IT has been a patchwork process, with individual contributors crafting tiny pieces of technology to handle repetitive tasks, and some basic event responses and human processes automated when the cost of not doing so got too high. If an IT organization has gone further than that, and adopted a policy of automating as much infrastructure management as possible, the focus of those efforts have almost always been on server and storage provisioning. As Greg points out, the network aspects are almost always simply left to manual processes--which results in great resistance to automation by the network administration professionals in these organizations. Besides, change in network configuration is bad, right?
But Greg makes a great case for why CIOs need to make an all-out commitment to drive their organizations to embrace automation end-to-end, and why network automation becomes a necessary and desirable component to that. I have to say that I agree with him when he says:
“It is easier to create a server and move it with virtualization, yet the network required the same manual configuration tasks to keep up. The lesson: system automation needs to be accompanied by network automation; or dynamic systems require dynamic networks. That is essentially the case for Infrastructure 2.0.
“Without dynamic infrastructure dynamic systems will produce less than stellar results and will increase pressure on the network and increase anxiety within network teams “doing more with less.” Because of this conundrum, network vendors and CIOs have a massive incentive to deliver on the promise of dynamic infrastructure, and the signs are encouraging.”
I can tell you that this very fact is a driving force in Cisco, and that much of our thinking is going into questions about how the network can best address cloud computing, which in turn leads to how we best simplify data center virtualization, which itself is driven by questions about how we take infrastructure from static, fixed architectures to dynamic, virtualized ones.
There are two paths that I see need to be addressed in this space. The first sees a lot of action today: enterprise data center virtualization and automation technologies. I like what Infoblox is doing in this space, but it is a piece of the overall pie. Cisco is looking to address end-to-end IT automation and virtualization with a combination of partner technologies from the likes of VMWare, and our own successes in the Catalyst and Nexus lines (e.g. the Nexus 1000v). Stay tuned on that front for some eye raising announcements. The second path is what we like to call the “Intercloud”, but I’m going to save a detailed explanation of that effort for a later post.
The writing is on the wall. The next frontier to get explored in depth in the cloud world will be the network, and what the network can do to make cloud computing and virtualization easier for you and your organization. I look forward to sharing the excitement with you as the story unfolds.