So, I boot up my computer yesterday evening at home. Everything is working fine. Half an hour later, I come back, and realize my 9 month old son (the one with an uncanny ability to find keyboard shortcuts I never knew existed) has been banging away on the keyboard. There are dozens of windows open, and there is clearly something wrong with my video driver. I’m not sure what, exactly, he did, but my computer was almost unusable.
Wanova Mirage to the rescue! I called our IT guy, and – instead of spending a hour trying to remotely troubleshoot my machine – he executed an ‘Enforce Base Image’ from the Mirage console. In a few minutes, my Mirage client notified me that an update was in process. I had dinner, and was then notified to reboot my computer. When my system came up – everything was fixed.
So – what was happening in the background? Mirage centralizes the full contents of my desktop in our data center, synchronizing regularly to ensure my full system is backed up. Further, Mirage separates my desktop into virtual layers, so that IT is able to manage a Base Image (my OS and applications that IT manages for everyone, like Microsoft Office) separately from my user-installed applications and data. Consequently, when I was experiencing a problem that was likely somewhere in the Base Image, our administrator could just ‘Enforce’ the Base Image from the data center. Mirage’s network optimization also came into play during this re-alignment. Instead of having to package and send an entire copy of the OS and applications to my PC, Mirage simply scanned my PC to find the differences between my BI and the one in the data center, compressed those few files, and sent them to my machine. When I rebooted, all the missing or corrupted files had been replaced in my local copy of my desktop.
It’s cool when the technology you market actually works!
I had an opportunity to attend BriForum in Chicago several weeks ago (and belatedly found a few minutes to write about it.) There were some really excellent sessions by Shawn Bass, Ruben Spruijt, and many others. In almost every session I attended, the speakers or the attendees discussed how to deal with the complexity created by traditional desktop virtualization. There were sessions outlining ways to overcome numerous challenges, including storage requirements, printing, supporting WAN-based users, antivirus, provisioning servers…and the list goes on.
While we at Wanova are aware of how easy it is to install and work with Wanova Mirage, I thought it might make sense to clarify. Wanova Mirage software gives you the benefits of VDI: centralized management, data protection, and simplified desktop support, plus the benefits of traditional imaging and software distribution tools, like single image management and patching, without the complexity. And, it works with your remote and mobile workers as well.
To run Mirage, you need:
Mirage Client software – a service that installs directly into the OS on the endpoint (ANY Win XP or Win 7 PC – Mirage doesn’t need Intel V-Pro, or a client hypervisor).
Wanova Mirage server software – a standard Windows Server service that hosts and manages the desktops images (without executing them – execution runs on the endpoint). A standard quad core 8GB server can manage up to 1000 endpoints.
Wanova Management service – used to manage the Wanova system.
Low-end storage – to host desktop images (We love SATA drives).
That’s it. Compare this to a typical VDI deployment, where you need:
Thin client application or device – including RDP support and other vendor-specific client software.
Connection broker – mediates access from thin endpoints to the hosted desktop farm, including user entitlement, pool management etc. To avoid it becoming a single point of failure, a redundant broker is required, adding further complexity.
VDI agent – application that installs in the Windows OS that runs on the hosted images and interacts with the connection broker.
Image management service – application that runs in the hosting virtualization platform and manages the images.
A farm of servers – used to execute workloads of active desktops. A typical quad core 8GB server can typically host a few dozen desktops. Imagine how many of those you need for a deployment of thousands of desktops…
Primary, high-end storage – to host the desktops if you want access rates similar to the local disk access in physical desktops.
Virtualization infrastructure and management service – to provision/deploy virtual machines for the hosted desktops.
Management service– web-based management application (runs on client and server)
I was pleased to hear an IT Director tell me that Brian Madden, during an impromptu session on the show ‘demo’ floor, commented (I’m paraphrasing), “If what you want to do is improve your control of your desktop environment, you don’t need VDI. You should look at Wanova.”
If you would like to see Mirage simplicity in action, you can sign up for a live demo and webinar, and we’ll be posting some video demos soon.
Brian Madden wrote a blog post today revisiting his prediction of two years ago that VDI would be ready for wholesale deployment by now, and was contingent on solving four key problem areas.
Single disk image for many users
Remote display protocols that are indistinguishable from local
Local / offline VDI
Broader compatibility for app virtualization
I (of course) was pleased to see that he mentioned Wanova in several of these categories, but I thought it was important to clarify a few points. With the emergence of XenClient and other proposed ‘offline VDI’ technologies, I find people assume Wanova is addressing the problem in the same way, using a hypervisor with some sort of ‘check-in/check-out’. This couldn’t be further from reality.
When we built Wanova Mirage, we started with the idea of how best to address real customer pain: helping IT better manage, support and protect desktops, especially for remote and mobile workers, without compromising user experience. We were not trying to force fit a technology that made perfect sense for servers to fit a mobile desktop model. That gave us a lot of flexibility.
Instead, we engineered a software solution that works within Windows. Hypervisors of any sort, while supported, are not required. Wanova Mirage centralizes the full contents of the desktop in the datacenter for management and protection, but distributes execution to the endpoint for optimal user experience (added benefit: no need for a massive server farm to run each desktop). Mirage’s virtual layers allow IT to create a single image and apply it to hundreds of users, easily addressing Brian’s first point. Because users are accessing a local cache of the centrally stored desktop, we can stay out of the ‘protocol’ war. Point 2 – Check.
And while connected, Mirage continuously and bi-directionally synchronizes the endpoint cache with the centralized desktop: IT changes to the managed image layer propagate down to endpoint caches, and user updates to the personal layer propagate up to the central desktop instance. Further, Mirage employs in-data-path and sophisticated network optimization technologies that enable extremely efficient data transfer. The bottom line is that with Mirage, the desktop is “always checked-in” and resides at the data-center. It just leverages the available CPU cycles of the endpoint cache to execute desktop workloads. This architecture makes it possible for users to easily work online as well as offline, with full performance and complete access to their entire desktop. Point 3 – Check.
Finally, in the Mirage layered model, IT can define core applications that can be managed and distributed as part of the Base Image (typically, Office, antivirus, etc.). But in the second layer reserved for user-installed applications, Mirage is agnostic to how these are provisioned. They can be installed by the user, or provisioned centrally through SCCM/Altiris, or virtualized. Whatever works best for your organization. Point 4 – Check.
In my view, easier is better. Why create a whole infrastructure with connection brokers, server farms, high-performance storage, new clients, synchronizers, hypervisors, and the list goes on…to virtualize everything, when you can just as easily manage what you have today?
Last week, we had a chance to go visit Brian Madden at the TechTarget offices in San Francisco. Ady, our director of product management, and I were planning to do a tag team demo of Wanova Mirage. Fortunately, we are flexible. Brian indicated that he would prefer us marketing types stay away from the camera, so Ady took on the demo solo.
Ah, the challenges of marketing in tech.
Anyway, Brian asked a lot of great questions and provided some good insights as Ady gave him a great deep dive into Wanova Mirage distributed desktop virtualization software. Highlights include a tour of the management interface, and a full restore of a complete desktop to new (and different) laptop hardware, complete with all user-installed apps and data. Oh, and did I mention the Mirage server is located in Israel?
In the nicely written article, “’Zero clients’ promise to replace fat clients but have downsides,” Bridget Botelho points out that while Zero clients promise to improve the manageability and reduce cost for IT, they also have some significant disadvantages. Most notably, she refers to user-experience demands that are in direct conflict with this approach. Tim Garland, CIO and director of management information services at the Mississippi State Department of Health is quoted, further highlighting this inherent conflict: “Users want all the bells and whistles of the PCs they are used to, but we don’t want them storing a lot of stuff on hardware drives and turning their thin clients into PCs,” Ragland said during a recent interview. “The whole point of a thin client is to have them store on the server, where we can manage everything.”
This quote exemplifies the dilemma of IT – how to bridge the gap between user experience and cost-effective endpoint management. Furthermore, the challenge of supporting adequate user-experience is far greater when the managed endpoints are mobile laptops, especially those used by knowledge workers or power users. These users are accustomed to rich user-interfaces. They need to execute their advanced applications quickly, and must be able to work offline. Further, they do not want to depend on network latency for every keyboard stroke or mouse click they make.
What if there was a solution that could, as Ragland says, “store [the desktops] on the server, where we can manage everything,” but do this without degrading the user-experience? This is the objective that we challenged ourselves to address when we founded Wanova. Now — 2 years later – this vision is fully realized with the Mirage product. The key idea is to centralize the desktop contents in their entirety in the data center for total manageability, protection, and support, but turn the endpoints into caching devices that execute the workloads locally. This way, IT has all desktops in the data center, while users get the rich experience they want. What’s more, IT does not need to forklift upgrade their data center infrastructure to centrally host and power these thousands of desktops. With Wanova’s distributed design, only a small fraction of server and storage resources is required. We leverage extensive deduplication to reduce storage requirements, and because the applications execute on the endpoints, the server is used primarily for management. In fact, the same server that could host maybe a few dozen thin clients can support a thousand Wanova-managed endpoints.
So – IT enjoys all the benefits of centralized desktops – central management, backup, continuity, migration, provisioning and support — while end users stay fully productive. Sound too good to be true? Why don’t you try it out for yourself ?
A few weeks ago, we reviewed a few of the key storage considerations necessary for a VDI implementation (and, more specifically, why Wanova Mirage’s distributed architecture doesn’t need high performance storage networks or heavy server farms.) But there’s another aspect to storage efficiency that is important to consider: data redundancy. While image management can reduce duplication across centrally provisioned applications, what about the thousands of user-installed applications, or the 600 people who have all saved the same copy of a Powerpoint presentation? Centralizing desktops in the datacenter can simplify management and protection of these endpoints, but for the solution to be cost-effective, it is critical that the solution also ensures an efficient use of storage.
Wanova Mirage leverages a global single instance store across all users of the system. Consequently, Mirage can deduplicate both user data and user-installed applications, as well as the OS and applications included in the Base Image.
Let’s consider a fairly conservative example: An organization has 1000 engineering end users. The Base Image for their systems is about 12 GB in size, and includes Microsoft Windows XP, Office, Outlook, Norton Antivirus, corporate agents and IT utilities. 500 of the engineers use the company’s software catalogue and install Visual Studio (2 GB). 100 users also install Visio (1.5 GB), and 50 users install Microsoft Project 2010 (2 GB). Using Mirage, the data comprised in the Base Image can be reduced from 11.7 terabytes to 12 gigabytes. Further, Mirage’s deduplication of the user- installed applications will reduce the storage requirement for this data from 1.2 terabytes to 5.5 gigabytes. Mirage will also deduplicate user files, which on average have about 30% redundancy, depending on the organization.
This indexing scheme results in a massive reduction in the amount of data that must be stored. The deduplication happens automatically — Mirage doesn’t require an administrator to do anything special in order to realize the savings. And what’s interesting is that the larger your deployment, the more efficient it gets.
In looking at the TCO of any desktop virtualization solution for a distributed organization, it’s important to consider how the system addresses all aspects of the infrastructure: endpoints, the WAN, LAN, servers, storage network, and of course, back end storage. At Wanova, we have invested – and will continue to invest – in all of these areas to create a solution that is optimized for the way global enterprises need to work.
If you’re like 90% of the companies we talk to, you’re most likely trying to figure out how to move thousands of PCs from Windows XP to Windows 7. Any way you slice it, the scenario is not pretty – especially if you have a large percentage of remote and mobile end users.
In fact, the challenge is so daunting for some that they have thrown up their hands and postponed their migration plans until a clearer solution emerges.
This week, we’re opening our beta program for Wanova Mirage 1.5, which adds support for Windows 7 endpoints. In addition, we are thrilled to provide a sneak preview into some technology from Wanova that will significantly ease the XP -> Win7 transition.
So that we’re all on the same page, the basic premise of Wanova Mirage is that it separates the PC into layers: an IT-managed Base Image (OS and core apps), a layer for user-installed applications and machine data, and a layer for user data and settings. These three layers (we call this a Centralized Virtual Desktop, or CVD) are stored in the data center, though we maintain a copy in a cache on the endpoint for local access and execution.
Now, CVDs are hardware agnostic, so one can imagine the following scenario:
An end user (let’s call him Spud), has a laptop running Windows XP. He’s an engineer, and he’s got years of his work stored on his PC: pictures of his kids, a hundred bookmarks in Firefox, Microsoft Project, Skype, and gigabytes of files. Furthermore, Spud is remote from headquarters and from IT, and has no access to nearby corporate images. His IT admin is trying to figure out how to get him (and thousands of his peers spread across the world), upgraded to Windows 7.
Fortunately, the company recently acquired Wanova Mirage. Spud’s CVD – and those of his colleagues – are already stored in the data center with a Base Image for their XP machines. With the click of a mouse, the Administrator introduces a new Base Image, which includes Windows 7 and IT-managed applications like Office, into each CVD. When Spud connects to the data center, the new CVD begins synchronizing with his PC. Because this transfer happens in the background, Spud can keep working, hopping on and off the network until the new Base Image is resident in his laptop cache.
Then, Spud is directed to reboot. Mirage instantly takes a snapshot of his XP desktop in the data center –storing the complete, bootable instance as a backup to make sure the current state is fully recoverable. Then, Mirage does its magic, transparently switching the operating systems. When Spud’s PC comes up, he will now be running Windows 7, with complete access to all his files, preferences, and machine state, so he can quickly reconnect to the domain without further configuration.
While Mirage will automatically provision applications contained in the Base Image, user-installed applications will need to be reloaded. Note: this is only true for XP -> Win 7 migration. In all other cases, such as moving a user from an old PC to a new one, or updating an OS service pack, Mirage will automatically preserve user-installed applications, as well as personalization.
So – all these benefits can be realized with a solution that doesn’t require a forklift infrastructure upgrade, or a complicated image distribution network, and it works well for WAN-connected users who are remote and mobile. Customers are telling us this capability will be tremendously useful. What do you think?
There are many ‘registry cleaner‘ utilities out there designed to remove junk from your registry and get your PC to run better, faster and with fewer issues. While some of the actions they perform are safe in nature (e.g., cleaning histories), other procedures can do serious damage if used incorrectly. These utilities can delete many items that a typical end user might not even consider. I was toying with the question – if an end-user ran such a tool with the wrong advanced options and damaged their system, how quickly could IT repair it with Mirage?
So I picked up one of the freeware tools and ran it on my work laptop (yes, sleep deprivation probably had something to do with the questionable logic of this decision). I ran the tool to create the most damage possible – letting it delete over 600 different registry keys ranging from file type associations to ActiveX and COM classes it deemed unusable. The system was still up after that, but many applications had serious issues: some crashed, some gave errors, and some simply wouldn’t launch at all. Internet Explorer had problems opening pages, and Flash froze often.
If I had to call a helpdesk to try to fix these problems, it could take hours or days to troubleshoot the issues. However, all our systems are Wanova Mirage-managed, so I put on my ‘Admin’ hat. With a few mouse clicks on the Mirage Management Console, I initiated an ‘Enforce Base Image’ command. A few minutes later, I was able to reboot my PC back to its fully functional state, without losing any of my data or user installed applications. And all this was accomplished via a Mirage server that was thousands of miles away from my crippled laptop.
Mirage’s image management, combined with its unique desktop streaming capabilities over the WAN, solved hundreds of issues in minutes. I didn’t have to touch anything on the laptop and didn’t spend any time troubleshooting. Think about the number of hours helpdesk employees spend trying (often unsuccessfully) to troubleshoot PC problems. How much could your organization stand to benefit by reducing helpdesk costs, improving support SLAs, and keeping end users productive?
Aloha! I was able to duck away for a short holiday in beautiful Kauai. While I was out of the office, our IT folks wanted to do some maintenance on my PC, so I was planning to bring only my Mac on my trip. This gave me a great opportunity to put Wanova Mirage into action.
Quick tutorial for those of you not familiar with Mirage: Mirage centralizes the complete contents of your desktop into the data center (including data, files, personalization settings, and any applications you’ve installed.) This is combined with an IT-managed Base Image, which includes the Windows OS and core applications like MS Office. We call this combination of ‘layers’ a Centralized Virtual Desktop, or CVD.
The CVD is a fully-bootable image and is hardware agnostic. To give me remote access, our admin simply provisioned a virtual machine in the data center, and assigned my CVD to it.
So, while I was sitting on the lanai, sipping a Mai Tai, I was able to securely access my full desktop through our VPN from my Mac using Microsoft’s Remote Desktop Connection. When I returned to the office, my desktop was moved back onto my laptop hardware – complete with all the changes I made while working remotely. Further, I now had an updated corporate image, yet all of my apps (like Photoshop and Illustrator) and personalization were still there. I was back to normal – with only a slight sunburn to show for my travels. All hail the flex desktop!
For all the benefits desktop virtualization offers for certain applications, there is an ongoing discussion around the back end storage requirements – and the potential costs. Ruben Spruijt penned a lengthy tome about it back in December, and with Win 7, we can be sure that file and applications sizes are not going to be getting any smaller.
Storage Performance Requirements with VDI
There are some slick demos out there these days that demonstrate how much better user experience can be for rich media applications over VDI. But what you don’t see is the price tag associated with the storage and server infrastructure required to support this kind of performance. Here are some things to consider:
As Ruben points out, because VDI requires a synchronous connection with storage, the quality of the storage network is going to have a direct impact on the quality of the user experience. High capacity, high performance drives = big $$$.
A good storage network isn’t enough. With VDI, user data is stored separately than the base operating system and applications, so you’ll need some beefy servers to prevent lags there.
The actual storage network can become the bottleneck in larger deployments, so you should make sure the SAN and Fibre Channel infrastructure can handle the traffic. For a 1000-user deployment, you’re going to need 20-40 Gbit of network connectivity to storage, with multiplexing and other costly features.
Yikes. No wonder the storage companies are excited about VDI.
Storage Performance Requirements with Wanova
With Wanova, you can get the same centralized management and data protection, but at a dramatically lower cost. Each endpoint runs a cache of the centralized desktop, leveraging local (and cheap) compute power and storage. Consequently, we don’t need an expensive server farm to handle execution. Plus – you get great side benefits – such as the ability to work offline. Since our architecture is distributed and asynchronous, we don’t need Tier 1 storage – you can use Tier 2 or Nearline storage that is 1/5 the cost.
To summarize just some of the methods Wanova uses to ease storage requirements:
Wanova leverages the cheap endpoint disk for executing the Windows and application workloads locally. Storage cost for workload execution on the client is $0.2/GB vs. $5/GB on the server side in VDI. And that’s excluding server side networking, NAS/SAN equipment, etc.
With Wanova, the endpoint acts as a “shock-absorber” to dampen any IO bursts and load during execution – for example, filtering page file access, temporary file access, log and repeated file read/writes, etc. This reduces the load on the Wanova Mirage server to handling only persistent changes on the desktops. So, with Wanova, the server and storage subsystem must only be sufficient to handle the averaged and deduped volume of changes. It is interesting to see Daniel Feller’s comment to Ruben’s article, highlighting the issues with sizing a VDI deployment using average IOPS. These issues do not exist with Wanova.
Another interesting fact raised in Ruben’s article is that with VDI, the typical IO workload is write heavy, with ratios mentioned in the order of 80:20 writes to reads. Write throughput is much more expensive than reads in RAID configurations, as Ruben explains. With Wanova, not only is less IO required at the server side, but the write:read ratio is actually reversed, allowing cheaper Nearline storage to be used with fewer, larger capacity disks – further enhancing the cost efficiency of the solution.
We recognize that there are applications where VDI makes great sense. However, for organizations that want to centralize control over their roaming force of laptop users (while still keeping them happy with good performance, personalization, and offline use), Wanova offers a far more cost-effective solution.
