So I started to work my mojo to counter their mojo; we got cross-mojulation, and their heads started exploding. - Austin Powers
This week VMware found something that it had been missing for at least six months - its mojo. At VMworld Europe in February, it was clear to me that VMware had lost its mojo. Gone was the swagger that built VMware's loyal following, in its place was a sense of doubt. The doubt brought with it the perception of a company not sure about its future direction. I'm sure the VMware product managers would argue that they have always been confident in their future direction, but to me, their public perception showed otherwise.
It's clear to me that with Paul Maritz's keynote on Tuesday, VMware is a company that is once again taking bold steps to aggressively push the envelope on how we look at IT, from the data center, to the application, to the desktop. Paul Maritz was brought in to find VMware's mojo and to ensure that it isn't lost again, and I think he's done it. To see why, let's look at the details:
- VMware Announces the Next Generation of Virtualization – the Virtual Datacenter Operating System (VDC-OS)
- VMware Announces vCloud Initiative for Enterprise-Class Cloud Computing with Support from Industry Leaders BT, Rackspace, SAVVIS, Sungard, T-Systems, Verizon Business and Others
- VMware Announces VMware Ready Program for Management Solutions
- VMware Optimizes Software Delivery and Distribution with Virtual Appliance Solutions
- VMware Announces vClient Initiative to Expand Virtual Desktops to Universal Clients – Desktops that Follow the User to Any Location
- Cisco and VMware Accelerate Innovation in Data Center Virtualization
Let's start with VMware's Virtual Datacenter Operating System (VDC-OS). VMware is going to catch some flack for calling itself the data center OS, but it is absolutely the right move. VMware got to where it is today by pushing the envelope with an incomparable vision of the future data center. VMware was so confident in their vision that the reaction from other vendors and IT pros had no impact on them. In 2001, several of my colleagues took pleasure in calling VMware "VM Scare" and found the idea of a day when we virtualize production x86 workloads completely ridiculous. Fast forward to today and you won't find a major enterprise organization that can't live without x86 virtualization.
If we are going to achieve the industry's collective vision of the future dynamic data center, having an intelligence layer between applications and the physical infrastructure is critical.I'm not going to get into all of the technical details of VDC-OS (you can read them on the VMware VDC-OS web site). Instead, I want to highlight a few critical points that need to fall into place in order for VDC-OS to grow as VMware intends.
First, VMware has to make it just as easy for software vendors to deploy both virtualized Windows and Linux applications to their virtual infrastructure. Applications available on the VMware Virtual Appliance Marketplace practically all run a Linux OS. So while the current offerings on the Virtual Appliance Marketplace represent a good start, there are still upwards of thousands of Windows applications that cannot be deployed using this model. Vendors cannot deploy a Windows application as a virtual machine appliance due to the nature of Microsoft's licensing policies. So instead, I see the future of Windows client and server applications as being deployed using application virtualization, such as Microsoft Application Virtualization (organizations wishing to deploy Windows applications as virtual appliances will need to package them internally using VMware's vApp in order to do so). VMware's ThinApp represents a good start and covers client applications, but I'd like to see VMware make it easy for ISVs to package Windows applications using vApp as well. I've called Microsoft's System Center Virtual Machine Manager support for ESX Server "Gateway Services for NetWare 2.0" on numerous occasions. To win the battle with Microsoft, VMware's going to need their own gateway; let's call it Gateway Services for Microsoft (GSMS). GSMS would be a very simple tool that allows an ISV to package their Windows apps that can be downloaded from the VMware Virtual Appliance Marketplace and deployed to an organization's VDC-OS. Software vendors have yet to distribute pre-virtualized Windows applications (no software installation required), but if that day comes, VMware will need to ensure that they can support the virtualized Windows applications, especially if they are packaged using a Microsoft product.
Upstream management integration is another key piece of the VDC-OS puzzle. VMware can't succeed alone. Instead they must encourage their clients to leverage best-of-breed enterprise management tools from vendors such as BMC, CA, HP, and IBM. It's clear that VMware is working quickly to fortify partner solutions, as evidenced by the announcement "BMC Software and VMware to Expand Business and Technology Partnership to Improve Automation and Control of the Virtual Datacenter." The agreement calls for BMC and VMware to deliver "Best Practices and Reference Architecture to manage the Virtual Datacenter Operating System (VDC-OS) from VMware and run BMC Remedy IT Service Management on the VMware platform." This is how VMware needs to partner with enterprise management vendors. Don't just announce a partnership. Instead, provide a reference architecture and technical details of how the products integrate. This is a crucial step for VMware to combat Microsoft's position on enterprise systems management. Vendors that are part of the VMware Ready Program for Management Solutions will have the tools they need to tightly integrate their products with VMware's product portfolio. I see this as a good move for both VMware and its partners.
Ask any network administrator how much they like the fact that it is often server administrators who manage virtual network switches and you'll likely hear a few four-letter words in the reply. So to me, it's no surprise that the VMworld floor was buzzing with interest of Cisco's forthcoming Nexus 1000V virtual switch. I believe that in time, many traditional network devices will transition to virtual appliances residing in the virtual infrastructure. 10GbE physical network interfaces offer the upstream bandwidth to make this possible. In time, the new network access layer will reside in the virtualization stack (via switches like the Nexus 1000V). Also, special purpose security appliances such as firewall and IDS will also reside within the virtual infrastructure. Moving these roles from proprietary purpose-built hardware to virtual appliances on commodity-based hardware will improve the elasticity of the dynamic data center.
VMware announced that the next vCenter Server (formerly Virtual Center Server) will be a Linux-based virtual machine appliances. In addition, the virtual infrastructure client software will be able to run on Linux and Mac operating systems. These two steps are placing VMware on a path to remove its dependencies on Windows operating systems to host virtual infrastructure management products. I expect all products in VMware's management portfolio to follow a similar path.
The vClient Initiative is also a critical part of VMware's strategy moving forward. To realize the benefits of anywhere access to any application, client-side hypervisors are critical. The reason for this is that they allow organizations to securely deploy applications or present remote applications to traditionally untrusted endpoints, such as a user's personal laptop or mobile phone. The desktop hypervisor allows organizations to fully isolate a user's personal and work environments, thus mitigating traditional risks of the mobile workforce. For example, a user that uses the Citrix ICA client to access work resources from his personal computer could place his organization's data at risk if a software keylogger is installed on his system. The desktop hypervisor could isolate multiple virtual desktop environments, thus preventing malware on the user's untrusted personal virtual environment from impacting his work virtual environment. While security is important, users will appreciate the following benefits of the future desktop hypervisor:
- Having the ability to run isolated "work" and "personal" virtual environments on the same system
- Seeing both personal and work environments presented under a single pane of glass
- Accessing a virtual desktop from a mobile device such as an iPhone or smart phone
- Providing offline caching that will allow users to have anywhere access to their data without the need for an Internet connection
- Synchronizing the client-side virtual desktop with the user's corporate virtual desktop image (at the organization's data center or via a cloud-based service).
The desktop has to be a key part of VMware's strategy if it is to hold off Microsoft long term. Paul Maritz has made it clear that VMware is going to take the fight to Microsoft, and not sit idly by as Microsoft chips away at their market share. VMware wants to redefine the traditional desktop, including how the desktop is deployed, managed, and presented to users. This will be a very tough task, but if they are successful, Microsoft's dominant foothold in IT will be at risk. Of course, VMware will not be able to redefine the desktop alone. It will need the help of a lot of partners, such as Apple, the IHVs that will embed the VMware desktop hypervisor on their desktops, laptops, and mobile phones, as well as cloud and service providers.
VMware is embarking on a battle that will be fought on several fronts: virtual infrastructure, cloud, and desktop. They are clearly ahead in the battle for the virtual infrastructure, but cloud and desktop are two hills in VMware's long term strategy that must be taken as well. These battles will be fought over the next several years and certainly will not be won overnight. However, it's clear to me that VMware has found the weapon it needs to fight these battles - its mojo.
Posted by: Chris Wolf