Virtualization And Cloud Q&a With Bernard Golden
I had the pleasure of meeting "cloud guru" Bernard Golden at Interop Las Vegas during a panel last spring
. As we look beyond VMworld and towards Interop New York, I was able to get some insight from Bernard as to the future of cloud computing. Bernard is the CEO of Silicon Valley cloud computing consultancy Hyperstratus and author of Virtualization for Dummies, the most popular book on the subject ever published. He is also the co-author of Creating the Infrastructure for Cloud Computing.
ScienceLogic: You recently attended VMworld and found it to be heavily focused on the cloud. What do you think is the next step for virtualization and cloud vendors? You mentioned storage as a big factor, what else?
BG: The next step is that virtualization will be baked into the computing infrastructure rather than being grafted on to achieve consolidation and higher utilization. With that native presence, a whole range of products - think security, monitoring, management, and so on - are targeted toward virtualized environments rather than being retrofitted to imperfectly support it. In a sense, we are midway through a shift to a technology infrastructure in which 'physical' doesn't exist - or at least is abstracted away from operational applications and tools. In addition to storage, networking will be a huge area to address, because increased workloads can lead to saturated networks.
ScienceLogic: As a known expert on both topics - how do you explain the relationship between virtualization and cloud?
BG: Virtualization is the foundation of cloud computing. Virtualization abstracts computing away from specific resources and enables easy transfer from one physical resource to another and easy addition and subtraction of resources from an application. Cloud computing overlays virtualization with a set of software services (aka, a software stack) that supports automation, automated application of policies, billing and chargeback, and so on. One may think of cloud computing as the logical extension of the implications of virtualization.
ScienceLogic: There are multiple definitions out there for cloud computing, with NIST's definition of cloud being the most cited. If you were to create the ultimate definition of cloud - including what components make up the cloud and what the impact is on business; what would it be?
BG: Really, we feel we can't add anything to the NIST definition, so we rely on it for our conversations about cloud computing. Where we add value to the definition is in helping organizations understand its implications along with why cloud computing is different, important, and critically, how to apply cloud computing to better solve technology and business problems.
From a technology perspective, cloud computing supports speedy provisioning and elastic applications, enabling characteristics that heretofore were extremely difficult to provide: highly variable loads, 'just-in-time' resource provisioning, reduced upfront investment, highly granular costs tied to actual resource use.
From a business perspective, cloud computing allows much more rapid response to changing business conditions or new business opportunities, short-lived offerings, and better correlation between business objectives achieved and IT costs imposed.
Overall, we are moving to the 'more' IT world: more apps, more rapid change, more data, more business initiatives that are IT, rather than initiatives supported by IT. An example of this is clickstream analytics, which allows companies to better understand customer desires and behavior, with a goal of making it easier to do business and increasing revenues. The drive to cloud computing is often described as increased agility - the ability of both business and technology to more rapidly respond to an increasingly unpredictable world. We provide an agility checklist to organizations so that they may assess their agility issues and evaluate how cloud computing can help them improve their agility.
ScienceLogic: You offer training and education for getting up to speed on cloud computing. What are the biggest challenges and barriers to adoption that you hear about? How can companies be more prepared for the cloud?
BG: We see three barriers:
First, understanding how to architect applications to leverage the characteristics of cloud computing that the NIST definition outlines. Even if the cloud infrastructure supports elasticity, self-service, and seamless response to load variability, that doesn't mean an application automatically has those capabilities if it is placed in a cloud environment. The application has to be created (aka 'architected') to enable and take advantage of the cloud computing environment's support of those characteristics.
Second, organizations often fail to take into account the necessary changes in processes and organization required to leverage cloud computing. IT operations has to change its working processes so that self-service is possible. Groups that formerly operated as silos (e.g., network and storage) have to cooperate and collaborate in a cloud computing environment.
Third, organizations typically fail to understand that most of the existing infrastructure support applications like network management, application monitoring, and capacity planning have to change to support cloud computing. Every part of the IT organization value chain is affected by cloud computing, but not everyone recognizes that fact.
ScienceLogic: In your opinion, how does monitoring and management fit in with the move to the cloud? Do you think that people are factoring the need to monitor into developing their cloud strategies?
BG: As I mentioned, monitoring - specifically, the changed monitoring requirements for a cloud computing environment - is often an afterthought when companies begin using a cloud environment. A virtualized environment, with transitory resources, along with dynamically assigned identifiers (e.g., temporary MAC addresses rather than a fixed MAC address) is challenging for traditional monitoring and management products that assume a stable, physical environment.
So it's a paradox - the malleable, elastic, ever-changing nature of cloud computing makes monitoring and management more important and more difficult, all the while making existing products less appropriate.
ScienceLogic: Who has it right when it comes to cloud computing? We think that service providers are leading the charge - but what about government and enterprise companies? Where do they fit in with cloud success?
BG: Surprisingly, government entities, particularly in the Federal government, are leading the change to cloud computing. Federal agencies are under extreme pressure to reduce costs via consolidation and are turning to cloud computing to extend virtualization and gain the same kind of benefits outlined earlier: greater agility, better elasticity, support for short duration applications. In addition to creating its own cloud computing environments, the Federal government is also turning to external cloud providers offering both SaaS and IaaS services; this use is represented by companies like Terremark, Google, and Amazon.
Turning to enterprises, they are also pursuing both internal and external cloud initiatives. Many business unit-led efforts are placed into external cloud service providers, while internal IT groups are evaluating or implementing internal clouds.
ScienceLogic: Some people worry that when the cloud comes, their job will go. But, you recently said that system administrators and other data center administrators will just have to work smarter. What are (3-5) top tips for sysadmins and others to consider when moving to the cloud? What new or improved skills will the sysadmin of the future need to have?
BG: I feel the right way to think about this is to consider what skills are required to support the shift from piecework to mass production. In piecework, the skills required are those of doing the work itself, while in mass production the skills required are those of standardizing and mechanizing the skills.
This means that the sysadmin of the future will be able to capture the steps required to, say, provision a virtual machine instance, and create a set of automated scripts that do that provisioning - including the scripts that install the monitoring and management frameworks, that support dynamic addition and subtraction of resources, and that integrate with billing and chargeback systems.
So, key skills or areas to get knowledgeable about include: (1) new system management and monitoring tools that support virtualized, dynamic environments; (2) automation frameworks like Chef or Puppet; and (3), last but not least, application lifecycle management concepts for applications that are constantly changing and blurring the distinctions between separate stages of an applications lifecycle (e.g., development, test, staging, production).
Bernard Golden is the CEO of HyperStratus, a Silicon Valley cloud computing consultancy which works with clients in the US and throughout the world. The firm's clients include Sensato Capital, Sun Microsystems, and the Silicon Valley Education Foundation. Bernard is the Cloud Computing Advisor for CIO Magazine. Bernard's writings on cloud computing have also been published by the New York Times and the Harvard Business Review. Bernard is the author of Virtualization for Dummies, the most popular book on the subject ever published. He is also the co-author of Creating the Infrastructure for Cloud Computing (Intel Press, 2010).
by: Larissa Fair
The advantages of gold plated jewelry Bangkok Condo For Sale - A Golden Opportunity Gold As An Investment Development Of Stainless Steel Jewelry Silver Jewelry Gifts For Your Better Half Jewelry Gift Ideas Are The Best Choice In All gifts For Her Ideas Top 5 Items Your Jewellery Box Is Incomplete Without Five Star Defeated: Golden Week Does Not Cut Prices On Withdraw Get A Wow Gold Guide To Make More Gold Jewelry Store Houston: Fine Gifts Of All Kinds Drill A Satisfactory Hole With Diamond Core Drill Bits Need Help To Select Sterling Silver Charm Bracelets? Jewelry Display