December 04, 2008
Las Vegas isn’t exactly Silicon Valley East, but it does manage to keep IT journalists living here busy with countless conferences. This week, the Gartner Data Center Conference is in town.
Now, I respect Gartner as much as the next guy, but it has to stop scheduling keynotes at 8 a.m. I have no clue how these pent-up datacenter folks get up that early after spending their nights experiencing Sin City, especially considering I was unable to make it to the MGM Grand that early -- even to hear about cloud computing and disruptive datacenter technologies. (Shocking!) Luckily, I’ve got slides.
Let’s start with the cloud. In his presentation titled “The Future of Infrastructure and Operations: The Engine of Cloud Computing,” Gartner’s Thomas Bittman presents cloud computing an external implementation of a real-time infrastructure (RTI). RTIs are distinguishable by their capabilities around automation, service-orientations, dynamic provisioning, SLAs, etc. Cloud service providers themselves already have made the move to RTI, and organizations wishing to leverage cloud computing to its maximum potential should be on their way to RTI, too. His timeframe for widespread organization RTI adoption is 2010-2020. Of course, virtualization enables all of this, and advanced organizations already are in the midst of virtualizing their operations.
(For what it's worth, Bittman seems to have warmed to the cloud since last year.)
Speaking of virtualization, Bittman breaks it into three distinctive versions thus far. Virtualization 1.0 was about consolidation and cost savings, Virtualization 2.0 is about agility and speed, and Virtualization 3.0 is about alternate sourcing, or cloud computing. Initially, the latter might consist “simply” of moving VMs between physical servers to meet demand. In more advanced stages, this would include the much-hyped hybrid model -- moving workloads between the datacenter and off-site cloud resources.
In order to handle this “cloudsourcing,” Bittman believes smaller enterprise will turn to “service brokers” -- the evolution of today’s systems integrators and VARs -- who will orchestrate cloud resources and take responsibility for service levels. Larger enterprises, Bittman suggests, will form business- and IT-savvy “dynamic sourcing teams” that will make day-to-day decisions regarding sourcing.
According to Bittman’s presentation, RTI, or internal cloud, will further be enabled by meta operating systems (see VMware’s Virtual Datacenter Operating System) and advanced blade racks featuring integrated compute, storage and memory, aggregated CPU and memory, and virtualized I/O.
In a presentation I did attend, Gartner’s Donna Scott expanded on RTI. For starters, she said that, presently, there is no such thing as a datacenter-wide RTI; they only exist in pockets. However, she added, RTI is being spurred by external forces that include grid computing/HPC/cloud computing, shared services, SOA, IT pressures (cost and performance) and datacenter concerns, like modernization or constraints. The driving force? Scott’s June 2008 survey and a quick audience poll both found “agility” the leader, but “cost savings” is No. 2 and rising. (Let’s not forget that much has changed since June. Cost savings could be a chart-topper very soon.)
Another audience poll found “management process maturity” and “organization/culture” as the top two inhibitors to RTI -- a result that makes sense. These are the exact same issues that plagued grid computing in the mainstream enterprise, and the same I keep hearing cited around cloud computing -- two delivery models that fall under RTI’s penumbra. Other inhibitors include infrastructure maturity, unproven technology, lock-in concerns and ROI.
It’s not all about the future, though. Scott says that RTI is entrenched in high-performance computing environments, and progress is being made in the following areas: server virtualization, J2EE and .NET apps, database virtualization, disaster recovery, shared testing environments, loosely coupled server HA (acting as loosely coupled clusters), repurposing of production nodes (e.g., from batch to OLTP), and in dynamic capacity expansion for specific applications.
Scott also talked about the need for service governors, the brains of RTI. Service governors take action based on SLAs, policies, application demand, etc. However, she added, you’ll likely need several governors across different tiers, applications and environments. This definitely seems true as it relates to VMware, for example, whose VDC-OS initiative seems all-encompassing, but only supports VMware environments. That said, companies like IBM might be eliminating the need for multiple governors with cloud management solutions like Blue Cloud.
In two areas of particular concern to me, Scott named as application service governor leaders Cassatt (with Active Response), DataSynapse (with FabricServer) and IBM (with WebSphere Virtual Enterprise). For infrastructure service governor leaders, she noted Cassatt again, as well as Univa UD (with Reliance), CA (with Data Center Automation Manager), VMware (with DRS), Novell (with ZENworks) and IBM (with Tivoli Intelligent Orchestrator and Tivoli Provisioning Manager).
As for how to build an RTI, Scott suggests doing it from the bottom up. First, she says, standardize and consolidate. Next, virtualize. Step three, standardize processes. Finally, automate policy-based actions, move toward service-oriented IT, etc. -- all those cutting-edge steps that seem might seem like pipe dreams right now.
I’ll end with Carl Claunch’s list of the “Top 10 Disruptive Technologies Affecting the Data Center.” No surprise, cloud computing is on the list -- No. 2, to be exact. (I hope no one is upset that I’m not defining cloud here for the billionth time, and because I didn’t actually see the keynote, I can’t offer any burning insights from Claunch himself.) He also includes on the list, at No. 3, computing fabrics, which are very similar (if not the same) as the next-generation blade racks Bittman spoke of in his cloud keynote. This makes sense to me because, as I noted on several occasions last year, IBM said its first Blue Cloud solution would be in the form of an IBM BladeCenter. Whether this actually happens, we’ll have to wait and see.
In order from 1 to 10, Claunch’s list of disruptive technologies is:
Posted by Derrick Harris - December 04, 2008 @ 4:21 PM, Pacific Standard Time
Derrick Harris is the Editor of On-Demand Enterprise
No Recent Blog Comments
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?