Inspired by Mobility

I’m here at VMworld and will take the stage later this morning to present the End-User Computing portion of the keynote along with the EVP and GM of EUC, Sanjay Poonen.  One of the themes of our talk will be the idea of “inspired by mobility.”  I wanted to take a moment to explain in a bit more detail what I mean by this and our thinking behind it.

As Sanjay and I will talk about, we’re in a new era of computing: the mobile-cloud era.  This era is characterized by changes in apps, data, devices, and how end-users consume them.  Long gone are the days of one Windows laptop per user.  Now each user moves between three or four devices in a single day.  Users access many different types of apps – not just Windows apps (though they are still very prevalent within the enterprise), but also SaaS/web apps and mobile apps.  Indeed even the way users consume data is changing.  The rise of “magic folder” data sync solutions has radically simplified the way users manage data and collaborate.

As these changes have rippled through enterprises, IT has been struggling to cope with the degree of change.  Consumerization of IT is bringing in unmanaged devices to the datacenter, with employees demanding full support for IT services on these unmanaged devices.  The number of devices per user is growing faster than the size or budget of the IT department.  These are big problems faced by IT departments.  In the end, we need an architecture that can meet the requirements of this new era of computing.

A new architecture

The architecture we propose is what we call the mobile-cloud architecture: 

In a mobile-cloud architecture, all state lives in the cloud: applications, data, and policy and configuration.  In this model, devices are essentially stateless.  When connected to the cloud, they “light up” with the user’s apps, data, and policy.  In other words, the device is automatically personalized for its user.  (BTW, you’ll notice that this is an elaboration on the architecture I originally laid out in a previous blog called “The Cloud is the Device”)

iOS devices are a great example of the mobile-cloud architecture in action.  All important information lives in iCloud: what apps I’ve purchased, my files and photos, and my configuration settings.  If I buy a new iPhone and connect it to iCloud, and all my apps and data immediately start streaming down.  In a matter of minutes, I can have a new iPhone that has an identical set of apps and data to my old iPhone.

Importantly, in this model the individual device doesn’t really matter.  If I lose or destroy my device, no data is lost because all information is always synced with the cloud.  And that’s the key – data is dematerialized and effectively lives in the cloud, but can be cached on the local device for fast access.  This is exactly what I’m talking about with the notion of a mobile-cloud architecture.

The mobile-cloud architecture is powerful for three reasons:

  • First, it creates a truly seamless experience for end-users since all devices share the same apps and data.  This means that users will have a consistent experience as they move from device to device.  This architecture empowers users, allowing them to use the device of their choice and move between devices at will.
  • Second, it enables IT to centrally and consistently manage apps, data, and devices.  Management is also drastically simplified into what I call “point and click” style management.  (I’ll explain exactly what I mean in just a moment.)
  • Third, since automation is at its core, it’s easy to layer on governance and self-service from such products as vCloud Automation Center or VMware Workspace Portal.  This empowers end-users by getting their computing needs satisfied immediately while allowing IT to enforce its compliance and security requirements.

How do we realize this mobile-cloud architecture?  The good news is that mobile technologies are generally already there, as they have been redesigned from the ground up to fit into this new architecture.  Desktop, however, has a tremendous amount of legacy baggage that has kept it from realizing this architecture.  So our task is to help it get there.  We see two required pieces for enabling this: application delivery and desktop delivery.  Let’s discuss each.

Real-time application delivery

As I mentioned above, we take our inspiration from mobile when thinking about how desktop application delivery evolve for into the mobile-cloud era.  Specifically, AirWatch has made mobile application delivery very simple.

With just a few clicks of a mouse, an admin can entitle a user or groups of users to an app or a set of apps.  That app or set of apps is then automatically delivered to the user’s device, where it just pops up.  This is exactly the sort of “point and click” style management I referenced above.  The challenge is to realize this for desktop.

The very exciting news is that we now have a technology that can do this, in the form of CloudVolumes, which we acquired last week.  As it turns out, in CloudVolumes the workflow for an admin entitling a user or group of users to an app or set of apps is almost identical to that of AirWatch.  Here’s a screenshot of their admin entitlement UI, just to give you a sense of the similarity:

 As I blogged last week, CloudVolumes can deliver applications to Windows desktops (or RDSH) in real-time.  The effect is awesome: once an admin entitles a user to an app, that app immediately appears in the user’s Windows desktop.  The best part is that the app can also be immediately executed.  There’s no streaming or copying or anything in the background.  Instead, CloudVolumes leverages hypervisor technologies to optimize the delivery of the application such that it’s zero overhead.  CloudVolumes achieves exactly what we wanted: defining application state in the cloud and delivering applications instantly with minimal transfers to the device (a virtual desktop in this case).

Just-in-time desktop delivery

While we’ve solved the real-time application delivery problem, we still need to address the issue of desktop delivery.  The reality is that today the delivery of virtual desktops is time consuming.  From cloning to powering on and OS boot to customization, the end-to-end process can take many minutes before the virtual desktop is ready to accept user logins.  This problem compounds itself when trying to deliver hundreds or thousands of desktops at one time.  We need a better solution.

Just like CloudVolumes leveraged hypervisor technology to optimize the delivery of applications, we can leverage hypervisor innovation to optimize the delivery of virtual desktops.  In particular, the vSphere team has been working on something we’re calling Project Fargo.  Project Fargo enables the instantaneous cloning of a running virtual machine (essentially a virtual machine “fork”).  The cloned VM is identical in every way to the original and initially shares all memory and disk with it as well.  Project Fargo is very cool for two reasons: first, it gets you a new running VM in under a second.  Second, it’s a very lightweight VM because it shares all memory and disk with the original.  (To be clear, both the memory and disk are “copy on write” so if new VMs modify bits of their memory or disk, a separate copy is made for that VM.  We thus preserve security and isolation between VMs.)

The idea is then that we can leverage Project Fargo to quickly spin up new virtual desktops for us.  Rather than the multi-minute process we had before, with Project Fargo and some optimizations in how we do customization, we can have a new virtual desktop up and running and ready for user login in just a few seconds.  Very exciting!

Desktop reinvented, Inspired by mobility

Now let’s put everything together to see how desktop works in the mobile-cloud architecture.  In this new model, when a user wants to log in, vSphere will fork off a clean, base Windows 7 or 8 desktop.  CloudVolumes will then deliver all the apps and data for the user.  In just a few seconds, the user has their Windows desktop fully personalized and logged in.  This is done completely transparently to them.  The desktop will stick around for as long as the user is logged in.  The moment the user logs out, the virtual desktop is destroyed. 

The key point here is that the apps and data live in the cloud, not the desktop.  The desktop is really just a temporary vehicle for delivering the apps and data to the user.  This is exactly the mobile-cloud architecture in action.  There are many benefits to this architecture:

  • Management is radically simplified.  Rather than managing and maintaining the bits on a disk, all management is done through simple “point and click” admin UIs in the cloud.  The bits on the disk are ‘execution scratch space’ and will be automatically reconstructed from the state in the cloud when the user logs in.
  • Costs are dramatically reduced.  As I mentioned above, forked VMs via Project Fargo share all their memory and disk, leading to less resource usage per desktop and thus a higher consolidation ratio without a performance penalty.
  • Security and compliance is increased.  These desktops are more secure because everything is up-to-date and at the latest patch level.  They are more uniform because users get a brand new desktop every time they log in.  No old cruft to worry about – it’s all cleaned away with each log in.

We think this is a fundamentally better way to architect desktops.  It’s something we’re calling Just-in-Time (JIT) Desktop.  The name comes from JIT compilers, such as those for Java, where code is compiled on-the-fly as the need arises; we’re doing the exact same for desktop.  JIT Desktops enables IT to take application and service delivery to a whole new level. When IT is providing users with quick login times, access to a large catalog of applications, and seamless upgrades to the OS and apps without the users’ involvement, users will prefer this desktop to their unmanaged personal desktops.  This is desktop reinvented and inspired by mobility – providing desktop and application functionality just-in-time based on user-state stored in the cloud.


The mobile-cloud era is demanding new enterprise architectures to solve the age-old problems of app, data, and device management.  The old ways aren’t sustainable in this new era of device diversity and proliferation.  Instead, a new way is needed.  We’re inspired by the ease of management for admins and great experience for end-users brought about by mobility.  Our goal is to bring this mobile inspiration to all parts of end-user computing, but especially to the desktop space.  Powered by innovative hypervisor technologies, Just-in-Time Desktop reinvents desktop for the mobile-cloud era, aligning IT management and user experience across all devices.

What’s the most exciting part about Just-in-Time Desktop for you?  Have you had a chance to check out CloudVolumes? If you’re at VMworld this week, I encourage you to attend my Super Session today (EUC3319-S) to learn more or stop by the VMware booth to chat with our EUC team.

Other posts by

One Platform, Any Application – and Getting There with Cloud-Native

Today’s an exciting day for VMware as we hold our biggest launch ever for a plethora of new products. While Pat and Ben are unveiling all the new features, I wanted to give the Cloud-Native Apps perspective.  As you guys know, our focus is on helping customers develop next-gen applications and operate them in production. […]

EUC, Cloud-Native Apps, and the New, New Thing

Here at VMware, we’ve always worked to deliver solutions that allow our customers to confidently and securely deliver software applications critical to their business. The End-User Computing business unit was created to ensure secure and seamless application delivery to end-users.  We’ve dramatically enhanced our end-user delivery capabilities this year via: the acquisition of AirWatch, the […]

Integration & Innovation: Announcing Horizon FLEX

Today I’m at VMworld Europe where we made some exciting new announcements in End-User Computing.  For this blog, I wanted to focus on a new product we announced called Horizon FLEX.  FLEX is interesting because it’s both an integration and an innovation. Integration update First, let’s start off with what we’ve been up to with […]