VMTN Blog

EMC VP Chuck Hollis details his thoughts on the challenges facing the enterprise with their storage infrastructure and virtualization using VMware. I've excerpted a bit, but read the whole thing.

Link: Chuck's Blog: VMware Virtual Infrastructure 3 – Climbing The Mountain.

The Core Infrastructure Challenge

Simply put, the central infrastructure challenge is that server virtualization adds another layer in the stack.  As an example, instead of server / network / storage, it’s now virtual server / physical server / network / storage.  ...

Challenge #1 – Flat Name Space for VMotion

One of the most powerful and sexy features in VMware ESX 3.0 is the advanced capabilities of VMotion, managed by DRS. ... But this presents a new challenge to the storage infrastructure. You’re going to want the ability for every virtual server image to be able to see every storage object from every server. ...

Challenge #2 – Storage Resource Management

The starting point for enterprise-class SRM is discovery and visualization.  What do I have, how does it connect, and how is it all related? ... Now, insert server virtualization into this stack. ... What happens?  It breaks the connection.  Maybe I can see the virtual machines.  Or maybe I can see the VMware ESX servers.  But, unless some heavy lifting is done, I won’t be able to see that stem-to-stern view that makes enterprise SRM useful.  And you can’t manage what you can’t see. ...

Challenge #3 – Backup and Recovery

Backup and recovery – never a pleasant topic in the physical server world – gets even more thorny and problematic in a virtual server world. ...

Challenge #4 – Managing End-To-End Service Delivery

I’ve made the case before that we don’t live in a world anymore where one user uses one application.  What the user sees is a logical combination of application services that run on an increasingly complex IT infrastructure stack.  And IT finds it harder and harder to drive back to a root cause when there’s a performance or outage that users are noticing. ...

 

EMC VP Chuck Hollis has written up a nice synthesis of how grid computing and virtualization are related:

Given that IT organizations will end up having both -- applications that are smaller than a server, and applications that are larger than a server, it's a useful exercise to think what's in common, rather than what's different.  And the list is pretty similar.

• Both abstract an application from the underlying resources.
• Both need the ability to dynamically react to the need for less or more resources.
• Both need similar capabilities to be invoked, orchestrated and service-level managed.
• Both need a common data space to do their work -- no hard-binding of information to application will work here.

I also think Chuck gives the good 50,000 foot view in the first part of his entry -- it's all about the  containers. Once the compute workloads are in the logical containers (virtual machines) you can then spindle, fold, and mutilate as you need -- server partitioning, grid computing, or a future desktop made of a swarm of cooperating virtual appliances that may even live back on the server.

Continuing from yesterday's post on the Role of the OS on the desktop, two comments really brought out the vision of what Scott Lowe calls "application agnosticism", although I suppose more properly it should be called "OS agnosticism."

Scott Lowe:

Have Windows software but prefer a Linux desktop environment? No problem. Have one Mac application you love but use primarily Windows applications? No problem. The desktop OS will evolve to include the technologies (virtualization, primarily) that will make this possible so that users can use whatever they prefer without the headache of figuring out if their system is compatible.

Jessica Tanenhaus:

In the end, I think the winner on the desktop will be whoever succeeds in making the experience most seamless for a user. When virtualization allows desktop users to click an icon and run applications, without opening a new OS environment, desktop users will sign on wholeheartedly. VMware Player and Workstation have saved my sanity for running certain applications that only run under Windows; if I could run those applications in the OS of my choice without having to visit another OS environment, it would be another major leap forward.

And Kimbro touches on a few different topics (read the whole thing), but also chimes in on the desirability of the OS-agnostic desktop and the different requirements for a virtualized desktop that aren't there on the server:

With that combination, you can run all your Windows software, all your Linux software, all your Solaris software, all your BSD software and all the incredible Mac OS X software at the same time. That’s a pretty compelling advantage and something that just one year ago simply wasn’t possible. However the idea of not having a real OS like Mac OS X running on the hardware seems hard to imagine at this point. I clearly see advantages to it, but at the same time there are a lot more hardware access issues on the desktop than there are on the server. So it’s really hard for me to see replacing Mac OS X with a hypervisor, unless that was an Apple designed hypervisor that’s heavily optimized for the sharing of hardware. Today that doesn’t exist, but just a year ago, neither did the possibility of running Windows at near native speeds on your Mac.

With everybody in violent agreement, I'll stop here, but here's a question: are you running an OS-agnostic desktop today?

Link: Ballmer: Life after Vista | Tech News on ZDNet.

Virtualization is not going to be adopted first in the end-user market.

Yeah, especially since Vista Home EULAs prevent installing it in a virtual machine.

[Update: be sure to read the comments for more on the coming 'application agnostic' world.]

Kimbro at Virtualization Daily and Scott Lowe are having an interesting conversation on the Role of the OS from the desktop perspective. I'm knee-deep working with virtual appliances these days, which tend to be server-based at this point. There we are starting to see stripped down Linux distributions, which only include the services and complexity necessary for the application contained within it. I would expect to see profiled  operating systems (ie configured for specific tasks) from all the major players within a year or two. This is a major win for reduction of complexity, simplification of updates, and avoidance of security exploits in random unneeded packages.

But what about the desktop? What does a cooperating set of virtual machines on your desktop mean? And who profits more -- Microsoft, Apple, Sun, Red Hat, Novell ... ?

Kimbro rounds up some of the conversation and sets the stage in The OS is under attack:

If you can run Mac OS and Windows on the same machine and use whichever program you want, and drag data back and forth at will between the two, what does an operating system mean? In a sense, it just becomes a visual preference rather than a system or standards choice. And if you spend most of your time using Web apps, the operating system means even less. We’re not quite there yet, of course, but would such a world help Apple or Windows more?

Scott Lowe then wrote an extended riff on the End of the OS As-We-Know-It, but exempts Mac OS X.

I do agree with these conclusions on at least one point:  The general purpose operating system as we know it will cease to exist in the next 5 to 10 years, perhaps sooner.  I do believe that the release of massive development projects such as Windows Vista won’t be the norm moving forward and that, in fact (as others have predicted as well), Windows Vista will be the last of its kind.

Notice I didn’t place Mac OS X in that list as well.  Why?  Because I think that Apple is capitalizing on an architecture and a convergence of technology that allows it to make Mac OS X into what Windows NT was supposed to be.

I think Scott lumps together too much the technical arguments of Windows vs OS X (OS X wins) with marketshare arguments around the various unixes and Linuxes (OS X wins again). I don't think OS X has any magical advantage, and nor do I think it's uniquely privileged to run the hypervisor, as it currently does with Parallels Desktop and the upcoming VMware product.  In a hypervisor-driven world where my desktop is a cluster of cooperating virtual machines (security, storage, etc.) and where my "desktop" can live anywhere, Apple's current software-locked-to-hardware business model falls down. If I can't run OS X anywhere but Apple-branded hardware, then Mac users have to sit out on a lot of innovation going forward. However, virtualization is disruptive enough that it's still anybody's game to win, if they put their chips on the right strategies.

Scott concludes with the question of whether the hypervisor should be inside or outside the OS.

So I guess the future of the operating system depends on your perspective.  If you’re an operating system guy, you’ll say that the OS has a bright future, and point to developments such as built-in paravirtualization and bundled hypervisors to prove your point.  If you’re a virtualization guy, you’ll say that the OS is dead, and you’ll point to developments such as third-party paravirtualization and independent hypervisors to prove your point.  Which of these two is correct?

And that's the forking path coming up for virtualization -- do you get the hypervisor layer from your OS vendor, effectively locking you in to a single operating system? Is your ability to run a guest dependent on the business deals between the various players, which may be what we're seeing in the recent Microsoft deals with XenSource and Novell? Are you willing to run a hypervisor that is locked to one brand of hardware? Or does the hypervisor sit outside the OS, allowing you to use the OS of your choice? The latter future seems a lot more appealing to me, although I'm admittedly looking at it from a VMware perspective.

But we digress a bit from the central question here -- what about virtualization on the desktop? When two or more OSes are sharing your devices and now your display, what is the role of the OS? Who are the winners and the losers?

If you clicked through on the last post (There is No Spoon), you'll see that Paul, our intrepid blogger, finished his rant with:

The Von Neumann Architecture is not so much going to die as it is going to replicate itself so many times it is going to force us to consider other, more simple and basic ways to configure and run these things we call computers.

Apropos of this via virtualization.info comes an interview with Kate Keahey of Argonne National Laboratory in GRIDToday:

Converging Virtualization with Distributed Computing.

Virtualization introduces a layer of abstraction that turns the question around from "let's see what resources are available and figure out if we can adapt our problem to use them" to "here is an environment I need to solve my problem -- I want to have it deployed on the grid as described." For a user this is a much simpler question. The issue is whether we can implement the middleware that will map such virtual workspace onto physical resources. One way to implement it would be to provide an automated environment installation on a remote node.

But what really gives this idea a boost is using virtual machine technology to represent such a workspace. This makes the environment easy to describe (you just install it), easy to transport, fast to deploy and, thanks to recent research, very efficient. Best of all, virtual machine management tools nowadays allow you to enforce the resource quantum assigned to a specific virtual machine very accurately -- so you could for example test or demo your application in a virtual cluster making sparing use of resources, and redeploy the virtual cluster on a much more powerful resource for production runs. This is another powerful idea behind virtualization: the environment is no longer permanently tied to a specific amount of resource but rather this resource quantum can be adjusted on-demand.

See also the Grid-Appliance from the University of Florida, which won an Honorable Mention in the Ultimate Virtual Appliance Challenge earlier this year. Expect to see much more about resource pools, utility computing, and virtualization over the next year.

Link: Brownian Musings: Operating Systems never did exist, and the ones we have are doomed....

This is one of my very favorite rants. And it is one that generally has traction with people that just want to use computers, and has no traction with people that have been sucked into the way we build and define software.

I am not saying we don’t bundle up software and call that software an Operating System.

I am saying what we call an Operating System is nothing more than discription given to a bundle of software.

Background reading for today's opening keynote at VMworld. There will be no quiz following the session, but there is a tectonic shift going on in IT -- it pays to pay attention.

Overview page: Freedom

There is a significant change underway in systems infrastructure. The traditional infrastructure model of a single monolithic system running a single, monolithic OS and a single application at wastefully low levels of utilization is obsolete.

Karthik Rau: Changing Role of the OS

As the market for virtualization rapidly evolves over these next few years, customers need to ask themselves the following key question: Is it really simpler to have virtualization integrated into the OS and follow the same pattern of lock-in that has dominated the past 20 years of computing, or do I want a world where I have choice and can focus on running a best-of-breed technology stack for each of my applications?

Raghu Raghuram: Hypervisors, Operating Systems and Virtual Infrastructure

With virtualization, there is now an opportunity to implement security, availability and reliability outside the OS, through the virtualization layer. Implementing these services outside the OS delivers significant benefits.  First, the implementation is global in scope - independent of any OS or any application. Second, implementing these capabilities once at the virtualization layer benefits every guest OS and application on every VM. You no longer have to implement and manage agents or software for availability or security or system protection per application. Third, since the implementation is not dependent on the OS, it is inherently less susceptible to attacks on the OS and therefore leads to a simpler, more robust infrastructure.

Dan Chu: Virtualization and Licensing: What Customers Need

Vendors can evolve their licensing to allow customers to take advantage of new technology, or conversely vendors can hold back and seek to inhibit and restrict how customers can use new technology because they feel threatened by it.  Customers have adopted virtualization broadly and made it mainstream, and have been able to drive some significant changes and improvements in licensing and openness.  However, there are also a growing number of areas where specific vendors (Microsoft in particular) are threatening to use licensing to restrict and undercut the benefits that customers and the industry are gaining from virtualization.

Steve Herrod: Virtualization: Open Standards, Interfaces, and Formats

For virtual appliances to achieve their full potential, openness in virtual machine-related interfaces is critical. The real promise is "any software on any virtualization layer". We believe customers should be able to choose and/or purchase a virtual machine consisting of any application running on any operating system and then run it on their virtualization layer of choice.

VMware co-founder Mendel Rosenblum will be speaking at VMworld next week. At some point in the recent past he was sitting on a plane beside Dan Farber of ZDNet and had this to say:

The biggest misconception is server consolidation. It was the first big application in the enterprise. Part of the problem with being the first successful application is that it misses a lot of the benefits. A virtualization layer is a different way of thinking about hardware, and server consolidation is just one tiny sliver of it.

People deploying virtualization infrastructure now view hardware differently–you have a pool of resources, virtual machine units, and you allow the infrastructure to map onto computation. If you need more capacity, just add more hardware.

He goes on to talk about capitalizing on software services, the TCO of automation and 'utility' computing, and the changing role of the OS. A good teaser for the keynotes at VMworld, and worth reading.

(via Virtualization Daily)