From the Blogosphere
The Three Biggest Tech Barriers to Cloud Computing
A Network Perspective
May. 11, 2009 02:05 PM
I recently posted about the Dizzying Economics of Cloud Computing when it occurred to me that the technological barriers must be equally mystifying for many. So I thought I would initiate a discussion about the barriers to the adoption of cloud computing by the enterprise.
At stake are the valuations of a gathering storm of public companies in technology, from Cisco, Juniper, F5 Networks, VMware, IBM, VMware, Microsoft and Citrix to advertising player Google and bookseller Amazon. The shape of adoption and growth will be impacted by how these barriers are addressed.
As I mentioned in "Dizzying Economics..." there is a business case war coming between the current cloud providers and the enterprise IT world. As these barriers are broken with innovation, cloud will move from consumer and SMB (small medium business) into larger and larger enterprise deployments. As a networking pundit, I see the three barriers in a network-centric perspective, from network security to the physical network and network management requirements.
Cloud Will Depend on New Approaches to Security
The biggest payoff of cloud computing comes from its potential to consolidate millions of servers (and the considerable amount of management and energy expenses) into dynamic meshes that can be created on demand; transforming IT from hardware-bound silos into just in time IT services delivered at any time from the most advantageous location.
Yet each one of those silos today is protected by a variety of mostly static security technologies. There are few viable solutions to the movement and change enabled by VMotion, or the ability of a server to move to its most cost-effective location at the time it is spun up. And most of those solutions were architected for specialized hardware, not commodity blade servers.
Today's world of network security solutions were simply not architected to keep up with the movement required for cloud to deliver on its promise of consolidation and cost efficiencies.
Yet without that movement, cloud just becomes another silo with tactical automation. One of my favorite blogs for cloud security is Chris Hoff's blog. Rather than drag you through the issues, check it out.
VMware purchased my alma mater Blue Lane Technologies last year. I think it was a smart move to address the emerging needs of its data center customers lured by the promise of cloud computing.
Cloud Can Break Static Networks
This year there has been a surge of content about Infrastructure 2.0 or dynamic infrastructure. Out of the discussion have been some interesting comments by conversation participants, from Cisco's Doug Gourlay and James Urquhart to F5's Erik Giesa and Lori MacVittie and VMware's Mark Thiele.
The network effects of cloud computing, which broke through the noise at a Cisco data center blog in December 2008, kicked off a discussion of multiple issues: from network switching to network management challenges. Cisco's Gourlay will be talking about the impact of cloud on the network at the exclusive, "thought leaders only" FIRE Conference later this month in San Diego. Don't expect him to pull any punches.
The physical hardware switching issues are being addressed but the leading players are by no means finished with what needs to be done in order to enable the pulsing fabric needed to ensure the cloud's integrity and business case promise. There have been an exciting and relevant string of announcements, including Cisco's UCS, VMware's (8 blade) cloud OS, and Citrix/Xen/KVM virtual networking stack, not to mention the Juniper and IBM global cloud push.
Network Automation is Critical
With increased movement and growing legions of connected devices (from netbooks and multifunctional cell phones and traffic-rich gadgets there is a burgeoning need for networks to move from manual configuration to automation of menial and high risk tasks. There are therefore a host of appliances that are likely to get grouped from the current Wild West into a single category because of their ability to automate the management of networks. They include IP address management and dns appliance offerings, in addition to network monitoring gear.
Infoblox (my employer) is approaching the network effects of the cloud by leveraging its expertise in core network services DNS to deliver an integrated Grid technology solution, while others are offering specialized subsets with (I suspect) the longer term intention of being bundled into broader offerings.
As the cloud tears down silos, one trick pony solutions (including freeware) will have an uphill battle for relevancy, especially as enterprises tear down silos. That is why I left Blue Lane during the VMware acquisition for Infoblox. CIOs are about to discover the significance of the network and core network services to unleashing the power of cloud. On demand It services will require unprecedented levels of automation and integrity.
The nature and extent of how the tech leaders address, security, physical infrastructure and network management barriers will impact their market valuations and growth potentials. You can expect a fresh round of partnerships, acquisitions and startups to reignite the tech sector as the economy stabilizes and enterprises shift to making strategic new investments.
You can follow my comments in real time at www.twitter.com/archimedius or catch the conversation as it happens at www.infra20.com.