Word on the Street

When looking at the majority of management solutions for Virtualization presently available, the first thing that may strike you is that these solutions and products don't seem all that different from what was available 5 or ever 10 years ago. There are many reasons why this has occurred, the chief among them is a legacy base on which most of these products have been designed. Just like Microsoft Windows has traditionally offered backward compatibility for Windows based applications, most management tools have had a great deal of legacy support built in or worse, the new products simply follow the old management paradigm without taking into account more efficient and simpler use paradigms. Looking at the design of most tools, the goal has been to make these tools match as closely as possible to that of traditional tools that administrators have been using. The logic appears to be sound at first glance, if an administrator is used to using a tool that works a specific way, there will be little in the way of new training or effort involved for them to learn the new product or tool. The unfortunate downside is that this leaves far less opportunity to adjust the interface to increase efficiency of administration or properly expose the capabilities that virtualization can offer.

The Cloud

As the cloud grows in popularity as an infrastructure that leverages virtualization (at least in the case of IaaS), current tool ability to support things such as transitions from one cloud to another or multi-tenancy scenarios will expose their weaknesses. This is because unless a tool has an architecture to support this type of highly Agile paradigm, it will not be easily adapted to work well with these new usage models. The only way the cloud will lower costs is if it can be effectively utilized by both end users and administrators and administration tools on the backend.

Adapting Slowly

Because most of the major players are attempting to tie in virtualization to their management suites or plugging virtualization management into a framework as an add-on, the ability for management to progress and innovate has been deeply constrained. This makes the speed at which business will be able to leverage the advantages of virtualization and the cloud a great deal slower. Creating an agile business based on tools and infrastructure that is slow to evolve isn't a combination for success.

New Entrants

With new solutions to Virtualization Management problems emerging from smaller companies and startups, the practical solution may be in leveraging a combination of old and new tools. By leveraging the existing processes and management tools that an IT shop already has in place, they are commonly able to cope with virtualization, but not take full advantage of it. By looking at new tools that give new insight and control into the virtual infrastructure and older tools, IT may be able to adapt more quickly even if most of their traditional tools can't.

This is a question that is often asked by CIOs and LOB Managers, but the answer seems to be different from company to company. All companies want (or at least should want) IT to be aligned with the business, to support business groups and the goals of the business. Companies also want IT to be able to tie themselves to different business functions to more easily adapt to the needs of the business. Traditionally, IT has been more of an independent business unit providing services to other business units throughout the company (and in some cases external customers as well). Unfortunately this approach causes tension in the organization as resources have been stretched thinner and costs have gotten higher. Proper business alignment includes not only having IT support business units, but implementing best practices and methodologies that are proven to work (such as ITILv3).

Virtualization alone cannot make IT become more business aligned, but with an implementation based on a Cloud approach (think next generation SoA) and ITILv3, things at least begin to head in the right direction. Even with Cloud and ITILv3, these things don't directly align IT and business, they simply provide tools, services, and work flows to being successful if things are properly aligned. This is analogous to having the best and fastest engine, transmission, and overall race car, but if the driver isn't very good, they still won't win the race. To win the race, IT must interconnect and map business process and function to IT resources and procedures.   

Aligning IT with the Business will also require the correct tools. This is where an interesting new problem lies, how can IT align with the business when all current and past tools have been designed for it to operate as a silo? This is made even more difficult as simply retro-fitting existing tools that don't take into account Virtualization or The Cloud would be a path to failure. This new crop of management tools has to allow IT to dynamically adjust IT resources to align with business units and for it to re-adjust that alignment dynamically or on demand at any time. If we assume for a minute that this will take place, then the viability of IT aligning with Business becomes far more practical and closer to reality.

"Neither rain, nor sleet, nor snow..." might be the motto of the United States Postal Service, but the VMware users in Columbus, OH and our Hyper9 team members are just as brave and share a similar belief.  Even though we're from Austin, TX, we weren't going to let a little bit (or a lot) of snow stop us.  Some of that can be attributed to bravery, and some of that could be the fact that we in Austin have no idea what snow is!

In spite of Mother Nature trying to ruin our fun, we had a great VMUG meeting in Columbus, OH on February 16th.  I think the city received something like 8+ inches of snow that day, and a lot of businesses closed down because of it.  It was touch and go for a while there; but luckily, there was a small enough window for our flight to sneak through and arrive in Columbus without being diverted somewhere else.

Thanks to everyone who made the journey and came out to the meeting.  And thanks to VMUG leaders Randy and Eric for helping to coordinate everything.

We'd also like to give special thanks to Steve Gruetter, a director at Platform Lab.  Platform Lab provides a great service, offering short-term physical or remote access to IT hardware, software and massive amounts of bandwidth for all types of companies for the explicit purpose of IT testing and projects.  And they also just happen to be the host facility for Columbus, OH VMUG events.  On this day, while the facility itself was closed from doing normal business due to bad weather, Steve was able to keep the facility open so that the VMUG event could still take place. 

We had some 50+ brave individuals attend the meeting, and that included the good folks from Panera Bread who we hired to cater breakfast.  Nothing better than hot coffee, bagels and pastries on a snowy morning, eh?    

At the meeting, Hyper9 talked about the practical ins and outs of virtualization reporting and best practices that address the challenges related to storage (how to squeeze the most disk space out of your environment as possible), VM sprawl (how many VMs do you actually have and which VMs are you actually using?), and management reporting (who owns what and how much does it cost?). 

We hope everyone had a good time at the event.  And we certainly enjoyed speaking with and meeting everyone.  Hopefully, we'll be invited back to Columbus in the future - Spring or Summer preferred.  :)

This was a mantra that many of the large software companies were speaking about several years ago (Microsoft, IBM, SAP, ...), "what happened to Business Agility?"

There was definite interest from businesses, as being able to react more quickly to lower costs and more effectively leverage resources within the business would attract virtually any executive, member of management, or business owner.

The Answer:

It turns out that implementing a system that coordinates an entire businesses resources and processes, allowing them to become more agile is a bit more difficult than the ease at which the marketing departments led businesses to believe. To fully realize these benefits there has to be a complete (or nearly complete) alignment between IT, division/group/line of business management, Executive Management, Production (Products, Manufacturing, R&D, etc.), Accounting, HR, etc. This is very difficult even with automation systems fully (and properly) implemented, such as SAP or other ERP packages, Cognos or other Analytics, custom applications with proper integration into other critical business functions, and a fully virtualized infrastructure.

The Reason for the Answer:

There are many potential reasons for Business Agility being difficult to attain, however the biggest reason behind the Answer is the ability to map the business and the processes to the proper resources (people, software, information). It is important to note that the mapping isn't a simple one way static mapping, but a multi-directional and ever changing map. If it were a static map, then it would assume that the business, processes, people, software, information, data and the like never change. What type of changes?

• Business conditions change on a regular basis, whether due to internal (organizational changes) or external factors (the economy)
• Processes change due to internal (better efficiency) and external (regulatory requirements) factors.
• People change internally (employees leaving, changing roles and responsibilities) and externally (customers and contacts are added and removed)
• Software internally (new packages are installed and old packages retired) and externally (suppliers and service providers add/remove/change capabilities and integrations) changes
• Information fluctuates internally (new ideas, business directions, strategies) and externally (discoveries, competitors)
• Data grows internally (new and historical data) and externally (changes in the market, deeper market knowledge, time)

The bullets above are by no means a comprehensive list, but merely a sample set. With all of these changes and shifts taking place, a static infrastructure and methodology for managing and mapping that infrastructure will be difficult at best, but more likely impossible to properly and efficiently manage and maintain.

This makes true or real Business Agility an impossibility until these needs are met.

What is required to meet the real needs of Business Agility?

A higher level system that will dynamically link all of these changing components and will allow for coordinated efforts to occur live as they are needed. This would begin to fulfill the promise of what Business Agility really was meant to be. Virtualization and some of the ideas behind Cloud Computing begin to offer the capabilities to do dynamic linking and proper coordination, however there aren't enough solutions that either allow integration with business applications and business processes or with the information and data they are generating. When solutions begin to offer these types of capabilities, that is when Business Agility and the real value it can potentially bring to business will be realized.

As the Cloud Market continues to take shape, many of the Cloud Providers have chosen to provide combinations of PaaS, SaaS, and IaaS. The reasoning behind this is that the market continues to be uncertain and probably will remain uncertain into the foreseeable future.

Why is the market uncertain?

Because different businesses have different demands and requirements. Mix this with the fact that developers are using the Cloud for portions of both public and private applications and you end up with no reasonable single answer to point to. Enter different types of cloud services for different businesses and business requirements.

So who are the primary players?

The larger players are offering at least two of the three service types and include:

IaaS includes offerings such as EC2 and S3

PaaS covers SimpleDB, RDB, and MP

PaaS with the Google AppEngine

SaaS under a suite Google Apps (Business Edition)

IaaS covered by Azure VMs (Service is estimated to be available in March of 2010)

PaaS includes Azure SQL Server and Others

SaaS includes MS Office Online

IaaS through Terremark

PaaS through SpringSource CloudFoundry

SaaS through VMware GO

(VMware doesn't necessarily provide these services directly, but does profit from them all)

By spreading their bets across all of these platforms, the Cloud Vendors are allowing the market to drag them to where and how it wants services. This is a good plan from the perspective of easily meeting the right market demands, however it may be difficult to provide good market differentiation in the long term. Over time, most people will be more concerned with PaaS and SaaS offerings from the consumer side. On the business side, assuming that Hybrid Clouds are the preferred deployment of cloud technology, IaaS and PaaS will weigh in more heavily as businesses need to invest in these areas to provide solid SaaS to business consumers internally to their organizations.

Uncertainty in the above prediction is also leading to fragmentation in each of the companies strategies. None of them want to spend money on a service or technology that isn't going to lead to money, whether directly or indirectly. On the flip side, none of these companies want to be driven or dominated by a competitor's technology or product. So their only choice is to spread their bets across as many of these solutions as possible and wait to see which becomes predominant.  And then, they can focus more time and money on the winning solution.

After reading a news bit on virtualization.info covering a stealth company called Nicira and its connection to Diane Greene (former CEO and Co-Founder of VMware), I thought it would be interesting to dig deeper into what was being done by Nicira and why it was different/significant.

What is Nicira working on?

Network Virtualization, but not the sense of what most virtualization industry veterans are used to, however. When people in virtualization think of Network Virtualization they think of two things: VLANs or Virtual Switches. But this is NOT what Nicira is doing. Instead, Nicira is building a framework built on top of the OpenFlow networking specification and codebase (which is Open Source). The Nicira framework is called NOX, which can be downloaded from noxrepo.org. The goal of NOX is to make the OpenFlow specification more accessible by allowing developers to write code to an API (currently C++ and Python are supported). Nicira is also developing a suite of applications that leverage their NOX framework. One of the applications already in the suite is called Ethane and it's centered around security.

How does Nicira plan on making money?

It appears that there are two paths that Nicira may take. The first is by providing commercial support for the OpenFlow Framework/Specification, their NOX Framework, and their Application Suite. The second is by selling/licensing their application suite commercially.

How is Diane Greene Connected to all of this?

Diane is listed as an investor on the Nicari Website. This doesn't appear to be the only connection, as the core of Nicari's technology and management come from Stanford University (some from Berkley as well). Stanford is where Greene's husband and Co-Founder of VMware Mendel Rosenblum is a Professor and received his PhD from Berkley. Mr. Rosenblum created VMware's core technology at Stanford under first the simOS Project and then the Disco Project. If Mrs. Greene invested money in this venture, there is likely some very innovative technology under the hood.

What is a Spring Container?

If you are a virtualization administrator type and heard about VMware's acquisition of Spring Source and therefore control of the Spring Framework (including the Spring Container) you might be wondering why you should really care or what this has to do with virtualization to begin with. Hopefully at the end of this blog entry you will have some of your own thoughts and opinions on where this all may be going.

A Spring Container isn't something that you put into a mattress or in your car's suspension, it is a way of looking at dependencies of an application (specifically libraries, etc. in Java). Wikipedia lists the Spring Framework as a collection of modules including an IoC (Inversion of Control) based container that allows Dependency Injection. What does that mean in english (to a non-developer type)? It means that Spring manages the integration and requirements that one component of a developer's code has with other components that it depends on and that depend on it. This process used to be something that was manually done and when applications were small wasn't very difficult. As Java applications grew larger and more complex, so did their dependencies. As the applications grew, they also began to change more rapidly, users made more requests for more functionality, requiring more changes, more capabilities, and even more dependencies. From this complex mess, a solution came about in a way to abstract away this complexity.

The complexity problem that was solved through abstraction of the dependencies mirrors the complexity problem that has been found in data centers and where virtualization has become the answer. IaaS based Cloud Computing is only viable because of virtualization and the abstraction of the complexities in provisioning and managing infrastructure.

Why do developers use it?

There are a select number of IoC based frameworks and VMware is focused on Spring, which is a Java centric framework. There are several Java centric IoCs as well as several .Net based IoCs (Spring.NET is Spring's .NET based framework (although not as popular as their Java based IoC). This entry's focus is on Spring and Java based IoCs. Developers often want to integrate different components together quickly and in a uniform way and want to have reusable code. Spring is the largest and most popular Java IoC and has the largest Java enterprise developer base. This developer base translates to a large number of reusable pre-written integrations for nearly all of the popular libraries and applications that an enterprise Java developer might need or want.

Possible motivations behind VMware acquiring Spring Source

As the market progresses and matures, several things are becoming apparent. The first being that the hypervisor is becoming more commoditized. The second is that the cloud is going to become more important, however VMware has been a virtualization company and their competitors are more than virtualization companies. Look at Microsoft and their broad portfolio of Operating Systems, Applications, Technologies, and now a growing number of Cloud offerings under Azure. Microsoft doesn't make money off of Hyper-V itself and makes little money off of the basic management tools for Hyper-V, they do however make a great deal of money off of their applications and hope to make money off of Azure. In essence, Microsoft has an entire ecosystem behind them. On a different cloud approach, look at Google who arguably has a giant set of offerings in the cloud from Google Apps to their App Engine. There are millions of web developers that have created and continue to create applications that leverage Google's web services and APIs. So what has VMware been missing? The answer: An ecosystem beyond virtualization and management.

Where VMware may be heading

So where does this leave VMware? VMware needs to put together a strategy that creates a full ecosystem and that is what they are attempting to do. In acquiring Spring Source they have gotten a large active community of Enterprise Java Developers. They also gained a significant amount of mindshare around leading Java projects including Spring's CloudFoundry (which was a Spring Acquisition), Spring's Groovy and Grails (which Spring Acquired from G2One), and Hyperic's instrumentation and monitoring (yet another acquisition). With this broad user base and now VMware's acquisition of Zimbra, it is becoming obvious that VMware wants to be a premier player in the Java Developer and Cloud ecosystem (by cloud ecosystem this means Cloud Computing from the perspective of all three platforms SaaS, PaaS, and IaaS). This would allow them to compete head to head with the likes of Microsoft and Google on both a platform and technology level as well as a cloud and application level. There are many questions that remain unanswered as to if all this is going to successfully come together and if so, how long will it take. The key will be to keep the development community happy and growing as this will likely prove to be the greatest long term potential for VMware to continue its growth.

For Reference:

Additional IoC Containers (For those interested in additional IoCs):

• Guice (by Google)
• Butterfly
• PicoContainer

There aren't many independent sources covering "Cloud APIs" - whether referring to "Independent" groups, standards bodies, or specific providers. This lack of sources without an agenda of some type makes it incredibly difficult to bring together good information that isn't somehow tainted. By creating this new Wikipedia entry, the hope is that a solid, independent, and over-arching resource can be created for the community to share. The cloud computing concept covers a broad range of topics, and therefore this Cloud API entry covers a broad range of available APIs.

The Cloud APIs are broken up into two Primary types, each having three sub-types. Think of a 2 row by 3 column grid (I will be updating this post and likely the Wikipedia article with a table of this sort, attempting to begin classifying of each Cloud API as time permits). The rows consist of Cloud Provider Cloud APIs and Cross Platform Cloud APIs, while the columns consist of Infrastructure (think IaaS), Services (think PaaS), and Applications (think SaaS). These categories (or combinations of them in specific cases) should allow for the classification of all current Cloud APIs and provide for a common understanding in the industry.

Please view or review the Wikipedia entry for Cloud APIs (which can be found at http://en.wikipedia.org/wiki/Cloud_APIs). Everyone is encouraged to go and take a look, edit carefully, and contribute thoughtfully to this effort as the more information that is put in, the more the community will get back out.

Some applications are more easily adapted or are just plain better running in the Cloud than they are hosted or running locally. Some people don't necessarily subscribe to this notion, however after looking as some of the reasoning behind this, it may make a bit more sense. Realizing that not only are some applications more easily put in the Cloud, others may be more difficult or nearly impossible to put entirely in the cloud (without substantial changes to code). Games, CAD, and other video intense applications are good examples of hard to move applications.  Non-web enabled video intense enterprise applications are becoming fewer today than compared to even 5 years ago, however they are still out there and the Cloud can only host the backend for applications of this sort, the rendering side has to be done locally. It is important to note that companies like Citrix have developed technologies that allow applications like this to render intense video remotely, but in the overall public Cloud there just isn't much traction for this (although there is a great deal of traction within the walls of the enterprise).  There are many applications that don't require this level of intense video, which is what is being focused on here. Business applications commonly have low rendering requirements, but have all sorts of other complex architectural requirements. Let's look at a few of these requirements and how the Cloud can help solve them.

The Scaling Story

Many successful business application implementations end up growing in the number of users and uses rapidly over time. This usually causes scalability issues and a great number of headaches for architects and administrators. In a Cloud based environment, it is trivial to throw a few additional resources at an application either on a temporary or permanent basis. This isn't so easy in a traditional environment giving the Cloud a distinct advantage in its flexibility and associated costs. By being able to leverage resources outside of the walls of the datacenter for specific needs on demand, a business application's response time SLA can be kept far easier than before.

Data Storage

Many SANs today are on a constant churn of storage space, changing partitions, hot and cold LUNs (for those unfamiliar with LUNs, think of specific chunks of space on the SAN), and a complex infrastructure to manage and secure. This is the reality of storage in data centers today. In a cloud storage scenario, these issues are left up to the cloud provider and the track record of cloud based storage is pretty good and steadily improving. Because of the ability to create storage grids under the covers, storage cloud providers can lower their costs, increase their storage space utilization, and easily add additional storage ahead of space and performance demands.

Databases

Commonly databases require complex and redundant systems with many systems either over utilized or underutilized. Most medium to large organizations use DBAs (DataBase Administrators) to managed and maintain their databases. Properly implementing and managing a database is not an easy task and with constant updates, patches, and performance tuning required, it is expensive. The cloud can provide a great alternative to running a database in house, the only drawback is your data doesn't reside on premise. For some companies, having their data locally is a must or in some cases required by law. Most companies have several databases that could be moved into the cloud without risking catastrophic loss or legal action.

Network / Connectivity

Some companies pay out high dollars for fast and reliable connectivity either from one location to another, whether that be an office to another office or from one datacenter to another. This scenario may or may not include internet connectivity. All of these network connections are pricey to implement and maintain, mostly due to the cost of the networking equipment and the network administration required to manage them. Performance and reliability for internet based applications can be very important to companies that rely on the web as a primary channel for revenue. Moving to a cloud provider can in many cases offer greater response times and availability than attempting to run a server either locally or co-located in a local hosting facility, all while not having to give up administrative control over the server (whether it be physical or virtual).

Centralization

By centralizing applications in the cloud, management applications can potentially become more uniform (this depends on what management application(s) are being used. The ability to spin up additional instances to provide better performance, greater concurrency, or redundancy for brief or long periods of time is a prime benefit to the centralization of the cloud. Other benefits of centralizing include the ability in some cases to replicate tiers in other geographic locations and to be able to control tiers through one centralized security policy.

Conclusion

Applications that are running in the cloud have many distinct advantages especially as resources become scarce - (whether they are compute or personnel). Small to medium sized businesses can benefit from offerings in the cloud for mission critical applications more rapidly than enterprises, primarily due to the great expense that most enterprises have already invested toward building out data centers of their own and potentially privacy and regulatory requirements. There are however pockets within enterprises which have specific projects or needs which can easily (and in a non-disruptive way) leverage the cloud and its benefits without necessarily violating IT standards. For small and medium sized businesses, the classic example of an easily accepted need is that of cloud application provider Salesforce.com which has been highly successful and offers a feature rich solution completely in the cloud, giving sales groups easy collaboration, lead generation management, contacts management, etc.  Salesforce has been so successful with their sales application, that they branched into other applications such as support.

Depending on what the needs are and what current applications are out there to meet those needs, the Cloud can be a powerful and cost effective tool to leverage no matter the size of the organization.

With the recent announcement of the vCloud API SDKs for both Java and Python, it seemed as though it might be relevant to investigate the available Cloud APIs, what they do, where they come from, and how they work. There are four Cloud APIs that are abstraction APIs (at least four of which that seemed to have a relevant amount of information around). Each of these APIs has its own take on how Clouds should be provisioned and managed through their API. All of the APIs do appear to have a common goal in creating an abstraction from specific cloud implementations so that a developer can create an application that will work on many different clouds while making a single API call.

What relevance does this have to a Virtualization Administrator? Today, probably not very much, but in the future, a great deal. Cloud APIs make it possible for developers and even applications themselves to generically interact with the infrastructure without having to interact with the infrastructure owner or administrator directly. This automation has many advantages in the speed at which applications can be developed and deployed because of the elimination of many manual processes and ties to specific operating systems and infrastructure specifics. This means that Virtualization Administrators will eventually be more of Infrastructure Security and Maintenance and less of Application Implementation and Support.

Each of the four Cloud APIs were briefly examined along with source code check-ins and whatever documentation was available. The tables below are not guaranteed to be perfect or error free, however they do provide a good idea as to which API provides what capability in addition to providing guidance as to where each is headed and how quickly it may get there.

(Click Image to Enlarge)

Is Cloud Computing Really New? Where did it come from? These are questions that are often unanswered or not covered in all of the buzz that companies like Google, VMware, and Microsoft are putting out there around the Cloud. Let's step back and look at where the Cloud actually came from.

In the beginning, there were great big hulking monstrosities that took up rooms of space, these were called mainframe computers. Mainframes brought us many of the terms used in computing today, along with a model that we are in essence moving back toward because of the efficiencies that it gave us. Out of the Mainframe Era, came the Mini Era. Mini Computers were smaller and less expensive than Mainframes and gave us a more practical approach to compute ownership for businesses and Universities. The PC Revolution hit following this and people became interested in doing more with their workstation than simply using it as a terminal to run applications on the Mainframe or Mini.

Enter the Client/Server Era, as more people had PCs placed on their desktop and more applications were written for Windows, people became interested in networking these machines together. Enter the Client/Server model from Large Enterprises to Small Businesses, companies were wiring their resources together to leverage PCs and have them jointly use File Servers, Printers, and the like.

The Internet had already been around a long time at this point and many University/Government Computers had been linked to the Internet. Now some companies were providing businesses and consumers with the ability to dial-up and use the Internet. As broadband became more popular and more companies began to leverage the Internet as a resource as well, the web became relevant and companies began to need a web presence. Around this time, a group of providers started to offer desktop applications as a service.

The Application Service Providers (ASPs) as they were called began to offer applications using remote access technologies such as Citrix Metaframe (now XenApp) and Microsoft Terminal Services. These technologies worked reasonably well, but weren't adopted for a number of reasons. These reasons included ownership of data, concerns over SLAs, and Licensing/Costs. As the ASPs were in decline, there was a fledgling company introducing a technology that they had been working on for PCs to allow them to run other operating systems in software. That company was VMware and the product was VMware workstation. Interestingly, VMware wasn't the first company to provide virtualization capability, that honor goes to a company once known as Connectix (they created VPC, the Connectix QuickCam which was sold to Logitech). Connectix created VPC a software package for Macintosh computers to allow them to run Microsoft Windows in a Virtual Machine, which ultimately allowed Windows and OS/2 Computers to run Microsoft Windows in a Virtual Machine as well.

Another thread of the story is around Service Oriented Architectures (SoA). The idea of exposing resources as web services making them more easily and uniformly accessible has been around for a while, but building the components, tools, and an infrastructure to support it has been tougher. VMware's push of Virtualization into the Enterprise allowed Applications to be coupled with Operating Systems and become highly portable. This was the push that was needed to make the idea of the Cloud actually function the way it was envisioned. SOAs were now something that could be built and torn down dynamically providing the last component technically to create an efficient cloud model.

Enter Amazon EC2, the first commercial grade public cloud. Why does Amazon have this distinction? There were other companies doing Cloud Computing before Amazon (Salesforce.com on the SaaS side and ProTier/Surgient on the Dynamic Infrastructure Provisioning Side), however none of them brought the web community (including Developers and Admins that needed resources) into the picture the way Amazon did. Amazon promoted an architecture and services, while creating a realistic model for charging for its Cloud Resources (a granular consumption model).

Today, there are many Public Cloud providers and companies offering solutions to create and manage Private Clouds. Public Cloud providers include:

• Amazon AWS and EC2
• Google Web Apps and App Engine
• Microsoft Azure
• RackSpace

Private Cloud Enabling software providers include:

• VMware
• Citrix
• Microsoft
• Surgient

The Cloud is the old reinvented as new based on all of the technological advancements from the proliferation of Virtualization and broadband Internet to the ubiquity of Web Services in a Web 2.0 based Applications. What makes the Cloud likely to be a permanent reality? Companies continue to shift from seeing Servers and IT as a strategic resource to viewing them as another type of facilities based service (this harkens back to Utility Computing). Another important note is that Cloud moves towards the vision that IBM mentioned several years ago (but at the time wasn't practical to implement and still has some significant hurdles, but moves closer and close), that being Autonomic Computing. The idea behind Autonomic Computing was (or for that matter, still is) that systems will become self-healing and will be able to fix, tune, and adjust themselves. This is where things will head, it is just a matter of how long till we get there. What isn't being said, but many are thinking is that this sounds a lot like HAL9000 or Skynet from (The Terminator Movies). Let me put those fears to rest, there is little doubt that there will still be a human somewhere in the loop checking policies or providing approvals and checks.

These are exciting times for the Cloud Computing space and what is happening in that space. Cloud Computing will be the hot place for Application and Infrastructure for the next 5 years at least.

With Yesterday's announcement of the launch of VMware GO, VMware has finally entered the Hosted (Cloud/Web) Application market. Providing services to small and medium businesses to more easily manage their ESXi implementations. On the management side, VMware GO provides a guided installation, ESXi Media Download and CD Burn (from ISO), VM Provisioning (P2V , Manual Creation, or Virtual Appliance - via the VMware Marketplace), Console Application Linking is Provided (the VMware Remote Console can be Launched from the VMware GO interface), Patch Scanning and patching to one or more Guest OSs and ESXi Itself, and Reporting rounds out the capabilities (very rudimentary patch and configuration reporting, but it is free). Many questions remain as to what finally tipped the scales for VMware to pull the trigger at this time (Amazingly just days after the news of the Zimbra acquisition hit).

This move may be an attempt by VMware to begin experimenting with ways of creating Hybrid Cloud environments in a future iteration of VMware GO (Interesting to note that GO was apparently created in partnership with VMware Partner Shavlik). How would this work and how does GO work today? The video embedded below is an Overview of How VMware GO works.

Where do you think VMware will take GO from here? Leave your comments...

The year 2012 is surrounded in a shroud of mystery, with many believing that some sort of cataclysmic or transformative event will take place December 21^st or 23^rd of that year.  Predictions of impending doom stem from interpretations made about the Mayan and other ancient civilizations and the Long Count calendar which is said to "end" sometime around that same date in 2012.

IT folks aren't strangers from thinking about facing doom and gloom in the future.  While we may not be following the Mayan calendar at work, our own calendar has certainly put misplaced fear in many of our hearts.  Remember the "Year 2000" problem or Y2K?  I for one was certainly glad the Millennium Bug didn't end my career like it was supposed to.

But Gartner recently announced a bit of prognostication that got me thinking.  Yesterday, the analyst firm launched their 2010 and beyond predictions which spanned some 56 markets, topics and industry areas, with around 250 predictions in total.

The one that quickly caught my eye stated: "By 2012, 20 percent of businesses will own no IT assets."

What?  No IT assets in one out of every five companies in as little as two years from now?  Wow!  That's a pretty powerful statement and quite a bold prediction.      

Gartner says this will be the result of virtualization, cloud-enabled services, and employees bringing in their own desktops and notebooks to work.

Gartner writes, "The need for computing hardware, either in a data center or on an employee's desk, will not go away. However, if the ownership of hardware shifts to third parties, then there will be major shifts throughout every facet of the IT hardware industry. For example, enterprise IT budgets will either be shrunk or reallocated to more-strategic projects; enterprise IT staff will either be reduced or reskilled to meet new requirements, and/or hardware distribution will have to change radically to meet the requirements of the new IT hardware buying points."

This prediction is not entirely a surprise, however; as we have witnessed this change happen over the last seven years to hosted Websites and e-mail.  Most companies are using outside providers for these types of services today rather than burdening their internal IT staff.

Virtualization has definitely done its part thus far in shrinking down the data center footprint of server equipment.  But this latest prediction has one in five companies with a zero footprint - NO IT assets!!!  That seems like an awfully heavy burden to place on the cloud community.  It means that cloud technology will need to mature a lot faster than has virtualization in order to reach those goals.  

If we plan on getting to a zero footprint of IT assets, cloud technology definitely seems to have the best chances to get us there; but that would also assume we are talking public cloud technology, and not private.  In the interim, many may choose to attempt a hybrid cloud in order to help bridge the gap.  But the clock is ticking, so people need to get moving if we are going to meet a 2012 time frame.    

As we enter 2010, 2012 no longer seems to be that far off into the future.  So if you live in the world of IT, make up your mind.  Which will it be... the end of IT as we know it?  Or will it be the end of the world itself?  I saw the movie... so I'm voting that we change IT if we have to!

David Davis does it again with yet another extremely helpful video, this time, an 18 minute demo that shows how to run VMware vSphere 4 inside a VMware Workstation virtual machine.  This is a great way to test and learn about VMware vSphere if you don't have any spare hardware. 

And stay tuned, Train Signal is going to be releasing his VMware vSphere training course soon enough. 

Looking to learn and master VMware's latest virtualization technology?  A good place to start is learning from the virtualization experts.  Check out these books that will be hitting the virtual bookshelf soon enough. 

Mastering VMware vSphere 4 (Computer/Tech) - By Scott Lowe

As part of the highly acclaimed Mastering series from Sybex, this book offers a comprehensive look at VMware vSphere 4, how to implement it, and how to make the most of what it offers. 

Coverage Includes:  

Shows administrators how to use VMware to realize significant savings in hardware costs while still providing adequate "servers" for their users

• Demonstrates how to partition a physical server into several virtual machines, reducing the overall server footprint within the operations center
• Explains how VMware subsumes a network to centralize and simplify its management, thus alleviating the effects of "virtual server sprawl"

Now that virtualization is a key cost-saving strategy, Mastering VMware vSphere 4 is the strategic guide you need to maximize the opportunities.

Link to Amazon. 

Scott Lowe's virtualization blog.

VMware vSphere and Virtual Infrastructure Security: Securing the Virtual Environment  - By Edward L. Haletky

Complete Hands-On Help for Securing VMware vSphere and Virtual Infrastructure by Edward Haletky, Author of the Best Selling Book on VMware, VMware ESX Server in the Enterprise

As VMware has become increasingly ubiquitous in the enterprise, IT professionals have become increasingly concerned about securing it. Now, for the first time, leading VMware expert Edward Haletky brings together comprehensive guidance for identifying and mitigating virtualization-related security threats on all VMware platforms, including the new cloud computing platform, vSphere.

This book reflects the same hands-on approach that made Haletky’s VMware ESX Server in the Enterprise so popular with working professionals. Haletky doesn’t just reveal where you might be vulnerable; he tells you exactly what to do and how to reconfigure your infrastructure to address the problem. 

VMware vSphere and Virtual Infrastructure Security begins by reviewing basic server vulnerabilities and explaining how security differs on VMware virtual servers and related products. Next, Haletky drills deep into the key components of a VMware installation, identifying both real and theoretical exploits, and introducing effective countermeasures.

Coverage includes:

• Viewing virtualization from the attacker’s perspective, and understanding the new security problems it can introduce
• Discovering which security threats the vmkernel does (and doesn’t) address
• Learning how VMsafe enables third-party security tools to access the vmkernel API
• Understanding the security implications of VMI, paravirtualization, and VMware Tools
• Securing virtualized storage: authentication, disk encryption, virtual storage networks, isolation, and more
• Protecting clustered virtual environments that use VMware High Availability, Dynamic Resource Scheduling, Fault Tolerance, vMotion, and Storage vMotion
• Securing the deployment and management of virtual machines across the network
• Mitigating risks associated with backup, performance management, and other day-to-day operations
• Using multiple security zones and other advanced virtual network techniques
• Securing Virtual Desktop Infrastructure (VDI)
• Auditing virtual infrastructure, and conducting forensic investigations after a possible breach

Link to Amazon.

Edward L. Haletky's virtualization blog.

VMware vSphere 4 Administration Instant Reference - By Matthew K. Johnson and Jase McCarty

VMware vSphere 4 Administration Instant Reference is a quick-reference guide for day-to-day administration of VMware's newest virtual infrastructure software. The book includes design features such as thumb tabs, secondary and tertiary tables of contents, and special heading treatments to provide quick and easy lookup, as well as quick-reference tables, lists, and step-by-step instruction to provide VMware administrators answers on the spot. This book is the perfect companion to any book on VMware, including Sybex's Mastering VMware titles.

Link to Amazon. 

Jase McCarty's virtualization blog.