Infrastructure Mashups
In a previous technology life I was fortunate enough to create really fun software targeted at the web. The fun part was creating logical units of work that could run independent of the container to which they were published. Other developers could then reuse the components within their own applications internal to the company. Advancements in web technology spawned many building blocks which were created at first for internal consumption and then much later for public consumption. The adoption and creation of the web mashup was born. Standard protocols are what made this possible. As a developer I now had a ton of flexibility and tools to create applications built for consumption, regardless of whether they were published internally or externally. The flexibility of the components alone was impressive; the only frustrating aspect was the lack of clear direction to what or how I should proceed. As with all technology adoption, timing is everything. Today we have a number of applications that are made completely of other services, but organized in a way that makes sense for a given business application.
Applying patterns and practices to other technology disciplines is very intriguing. We are in a very exciting time where this pattern of a mashup can be reapplied. Due in part to the cloud frenzy, it can now be applied to infrastructure. This was evident when I saw multiple full page ads by vendors advertising cloud computing in the Wall Street Journal. Each ad with its own spin or interpretation on what the cloud is and why the consumer should care. The problem in all this advertising is that “cloud-computing” means something different to everyone. If I were an organization looking to put together a cloud initiative, I would be frustrated trying to figure out a direction. The issue presented is a wide array of implementation options, which can be a blessing and a curse. I call it the Cloud Infrastructure Mashup. I believe it’s not going to be a one size fits all approach, but one where each organization gets to choose the best services to meet their needs, and mash them up.
The Cloud offers many of the building blocks needed to build these infrastructure mashups. Storage, applications, data platforms, network, communications, and hosted infrastructure can be delivered on-premises or in the cloud. Left out of the mix of offerings are desktops. Where are the desktops? More importantly, when are they going to be in the mix? Today the way to offer desktops of this nature is VDI (Virtual Desktop Infrastructure). Often times the mere mention of Virtual desktops is visited with adverse reactions. Many organizations have bought-off on the promises VDI can provide; but testing reality proves dismal. Are desktops one of the last great frontiers to tackle in this world of cloud computing?
Microsoft, Google and others are trying to offer their OS in the cloud. Honestly I thought we were still a few years off from completely ditching the laptop or desktop for the cloud based operating systems–until now. I have used virtual desktops in the past and was along the line of thinking “VDI is generally pretty lame.” However, over the last month I have been made a believer of VDI, and this is the genesis of the mashup idea.
I am a solution architect for Fusion-io and the work I have been doing with one of my customers, V3 systems, is the catalyst for my change of heart. Their solution and engineering staff helped me see the VDI light again.
V3 created an optimization layer for their purpose built VDI appliances which is coupled with Fusion-io’s PCIe based solid state as local storage to offer the fastest, highest populated, and simplistic approaches to VDI I know of. All this happens regardless of the hypervisor. I know how fast the Fusion-io storage is, and I was blown away with what their optimization layer does with Fusion’s technology to provide unbelievable benefits for VDI.
So it came as no surprise that after some initial testing of 50 virtual desktops on one V3 appliance the cogs started to turn. My first thought was “I want a virtual desktop–in fact I want more than one.” The virtual desktops served up by the V3 appliance were night and day faster than my local box. Which happens to be an Intel i7 with 16gb of ram. Hence the Infrastructure mashup concept was formed and a flood of old application mashup feelings came rush’n back…
Finally, a viable approach to VDI that actually works and fulfills the promises of what VDI should bring. This core building block gives organizations functional VDI options, whether it is on the local LAN, WAN, or in someone else’s public/private cloud.
Organizations stand to gain the greatest benefit from infrastructure mashups. Just like the application space, a mixture of clouds can now form a solution. V3′s VDI solutions will be key in allowing organizations to dictate what the cloud means to them and show how desktops can fill a giant void. The beauty here is: what the cloud means to one can mean something completely different to another. Delivering building block services in this case for VDI truly accelerates the viability of the cloud. I have full confidence that what V3 has accomplished will be game changing. The last frontier for the cloud is here and companies can finally tackle that age old cliché of doing more with less.
This V3 appliance is the first practical solution to having computing power where you need it, when you need it. I look forward to computing remotely from a Google TV in my living room which runs faster than the local machine on my desk. From my perspective they got VDI right. V3 systems has the ability to grow this space, set direction and provide a very feasible and adaptable approach to utilizing the cloud for desktops. Riding the wave of application mashups was really fun and I can only imagine that this wave will be exponentially greater in size and excitement.
Great work V3! It will be exciting to see the direction and thought leadership you will provide to the VDI world.
Chris Featherstone is a Solution Architect for Fusion-io. Fusion-io is a leading provider of data-centric computing solutions – a combination of hardware and software that places data closer to processing, resulting in dramatic improvements in both performance and efficiency.