On September 15th, 2008 at the Venetian Hotel in Las Vegas, Paul Maritz walked on stage for his first VMworld. It had only been just over 8 weeks since the press release came out (July 8) announcing the immediate replacement of the founding CEO Diane Green by Martiz. The crowd was eager to see what the new leadership was made of and where the company was heading. What was this ex Microsoft executive going to do with "our beloved VMware", for many people view the company as a family.
The keynote was the first announcement of the vCloud Initiative. Some might say a new direction for VMware, others a natural evolution.
It is very interesting to look back to the very start of vCloud and see the companies thinking given the knowledge we have today. A few things have happened since that date, the purchase of SpringSource and Zimbra, the launch (and some might say fizzle) of vCloud Express, the vCloud API being published and submitted to the DMTF.
Now that it is 2010, its a good time to go back to the beginning, and see the vision, one of the first pushes from the new CEO, that started the vCloud journey.
What follows is a transcript of vCloud elements (introduction and body) of the VMworld Keynote, Sept 15 2008, Paul Maritz, President and CEO VMWare.
We are moving fundamentally away from a device centric world, to a world that is application, information and people centric. We are all about how do we take infrastructure and treat that infrastructure as a common substrate on which we can build experiences that allow services to be provisioned ultimately to users in a much more flexible way without having to redo the underlying infrastructure. And allow us to gang together all the underling elements of that infrastructure to increasingly think of that infrastructure as a single giant computer of which applications and services can be fundamentally provisioned in a more manageable, scalable and extensible way. When we divide this challenge up we split it into three themes, the Virtual Datacenter Operating System, the vCloud Initiative and the vClient Initiative, the three Vs.
We are responding to is basically the following needs. On the one hand we see our current corporate IT customers under a mandate, particularly in this economic climate, to use the underlying compute resources they have in a fundamentally more efficient and flexible way. To be able to start behaving internally like a hosting provider or like an Internal Cloud, or reaching for the self service datacenter or reaching for utility computing or reaching for autonomic computing, they are all phrases that mean the same thing. How do I take my collection of infrastructure resources and increasingly see them as a single giant computer on which I can flexibly both todays application workloads and tomorrows application loads.
The next thing is how do you look outside and realise that there are increasing numbers of people who are building compute resources in their datacenters, in their publicly accessible datacenters, how do you take advantage of that? How do you think about that as part of your strategy. And our vCloud Initiative speaks to this.
[...]
We want the flexibility to allow you to start having much more freedom about where you pull your compute resources from. There is tremendous discussion in the industry today about the Cloud and we believe that the key point here is not about an either an, do you do it all internally or do you drop everything and try and do everything externally, its all really about flexibility and federation. So the needs to look at the cloud are real.
All of you are being asked to do things more quickly, more cost effectively, you need to look very hard if you want to cary capital expense on your books or have the capital expense carried on someone else’s books. So there is a tremendous interest and a genuine need to look at what can be done with off premise computing resources but it comes with real problems.
Generally these external clouds are not compatible with what we find inside the enterprise environment today. A lot of their infrastructures, the way that they have approached things, requires you to essentially write new applications. And on top of that they are all different. So how do you make the decision basically which guy you are going to use, are you going to get locked into them, do you have to rewrite your application etc. We think we can use the techniques we are using today to basically isolate application loads from underlying infrastructure inside the datacenter and apply those basic approaches to the cloud as well.
So we are working on an initiative to be called the vCloud Initiative. Which starts with that extended definition of an application load that I talked about earlier, the virtual appliance going to the virtual application or vApp standard that allows you as a IT organisation to understand when you have got your application load thats in a form that is well described, isolated enough, well formed enough, to be able to hand that to someone else and have them run it in their infrastructure. Conversely, it is a set of standards and a vocabulary that will allow cloud vendors to speak to you in terms of what application loads they can accept and run in their environment. And the hope is here and the intent is here that that will allow IT organisations and cloud operators to be able to talk in meaningful terms to each other and it does not become an either or situation, it becomes a situation where you can make flexible decisions about which parts of your application loads you want to run internally or externally and change your mind as the time goes by.
And clearly we have received a tremendous amount of interest in this area. We have over a 100 cloud operators who have signed on to be part of this vCloud Initiative including some big names like Verizon and British Telecom and Sungard and Savvis and T Systems etc. So they will be working with us to be able to articulate to you what kind of services they can provide to you and conversely you can understand how to get yourself and your applications into a position to be able to take advantage of that.
So we will be extending our Virtual Datacenter Operating System to not only have infrastructure services, application services, but have a set of services that have to do with federation, what we call our Cloud vServices, that will allow you to now federate across and distribute your application load across both, resources inside the firewall and resources outside the firewall.
Rodos
No comments:
Post a Comment