Of Wine, Virtualization and xVM

A few years back, I remember sitting with a group of customers talking about wine, and virtualization (a natural pairing, if ever one existed). Wine, because we were at an event Sun was hosting in Napa Valley, the heart of California’s wine country – virtualization, because the attendees were data center professionals who’d come to talk about the future.

The customers in attendance all ran very high scale, high value data centers, who would deservedly respond to the accusation that they “hugged” their servers with “and what of it?” They were the individuals who kept some of the world’s most valuable systems running with exceptional reliability.

But they were all starting to see and worry about the same thing, running applications in “virtualized” grids of networked infrastructure (“cloud computing” wasn’t yet in vogue, or I’m sure someone would’ve used the term).

Now, virtualization is a simple concept with a fancy name (abbreviated to “v12n” by the cognoscenti – by that method, I am “j14z”). It’s simply slicing up physical computers into many smaller “virtual” computers, each of which can be outfitted with its own OS and application stack.

That is, not only does a virtualized computer take on the task of running multiple OS’s (running atop a hypervisor, described below), but the OS’s themselves might change over time, responding to load or schedule. The traditional view of “computer A runs OS/Application B” can now give way to a more responsive “these computers are available for high priority work,” without regard to operating system or architecture. A spike in on-line shopping might reallocate more “virtual” machines to transaction processing during peak shopping hours, shifting to a different OS/app stack when the frenzy dies down. Capacity moves from fixed to fungible.

Although desktop virtualization wasn’t the focus of these customers, most live in a world with multiple desktop OS’s, too – it’s not that they all (like me) run five different desktop OS’s, most don’t – it’s that they have multiple generations of Windows, or no longer have the source code to legacy applications, a condition that dictates you keep old OS’s (and hardware) around. Desktop virtualization enables users to run multiple OS’s side by side on a single desktop, and divorces software upgrades from hardware upgrades (an innovation keeping CIO’s and developers smiling).

Back to the datacenter, virtualization can enable extreme infrastructure consolidation – decoupling applications from hardware drives more efficient capacity planning and system purchases. And as exciting as that was to everyone, if things went wrong, you could also tank the quarter, blow those savings and end your career. So, why all the anxiety?

If I could sum it up, these customers worried that virtualization would dissolve the control they’d carefully built to manage extreme reliability. In essence, they could hug a virtualized mainframe or an E25K (hugging is the act of paying exquisite attention to an individual machine), but it’s far harder to hug a cloud. Nor can you ask a cloud why it’s slow, irritable, or flaky, questions more easily answered with a single, big machine.

As the wine soothed their anxieties, a few of them began to draw out their vision of an ideal cloud environment (our laptops were open to take notes). Summarized, here’s what they wanted:

Extreme diagnosability. Datacenter veterans know that things rarely run as planned, so assuming from the outset you’re looking for problems, bottlenecks or optimization opportunities is a safer bet than assuming everything will go as expected. They all wanted ultimate security in responding to the question “what if something goes wrong?” – their jobs were on the line.

Second, they wanted extreme scalability – they all believed the move toward horizontally scaled grids (lots of little systems, ‘scaled out’), would give way (as it always does) to smaller numbers of bigger systems (‘scaled up’). We’re seeing that already, with the move toward multi-core cpu’s creating 16, 32, 64 even 128 way systems in a single box, lashed together with very high performance networking.

But scalability applies to management overhead, as well – having 16,000 virtualized computers is terrific (like 16,000 puppies), until you have to manage and maintain them. Often the biggest challenge (and expense) in a high scale datacenter isn’t the technology, it’s the breadth of point products or people managing the technology. So seamless management had to be our highest priority, with extreme scale (internet scale) in mind.

They wanted a general purpose, hardware and OS independent approach. That is, they wanted a solution that ran on any hardware vendor they chose, not just on Sun’s servers and storage, but Dell’s, IBM’s, HP’s, too. And they wanted a solution that would support Microsoft Windows, Linux and not just Solaris. Ideally embraced and endorsed by Microsoft, Intel, AMD, and not just Sun.

And finally, they wanted open source. After years of moving toward and relying upon open source software, they didn’t want to reintroduce proprietary software into the most foundational layer of their future datacenters. Some wanted the ability to “look at the code,” to ensure security, others wanted the freedom to make modifications for unique workloads or requirements.

And with that feedback, the answer to the above seemed obvious to one attendee, “why can’t you guys just use Solaris?” They all ran Solaris in mission critical deployment, all appreciated its performance, they loved the diagnosability (via delivered via DTrace), and the capacity to scale to the largest systems on earth. It was the perfect answer until one of the customers asked, “do Windows customers want to run Solaris? I don’t think so.” The “Solaris” brand didn’t convey OS neutrality – and that neutrality was core to what we were thinking. But we knew the underlying inventory of OpenSolaris innovations would certainly give us a fabulous headstart.

That’s the rough backdrop to what drove our virtualization announcements last week – a desire to solve problems for developers and datacenter operators in multi-vendor environments. If you look to the core of our xVM offerings, you’ll see exactly how we responded to the requirements outlined above: we integrated DTrace for extreme diagnosability. We leveraged the scale inherent in our kernel innovations to virtualize the largest systems on earth. We’ve built a clean, simple interface to manage clouds (called xVM OpsCenter, click here for more details), to address management and provisioning for the smallest to the largest datacenters. And everything’s available via open source (and free download), endorsed by our industry peers (watch these launch videos to see Microsoft and Intel endorse xVM – no, that’s not a typo, Microsoft endorsed xVM). We even leveraged ZFS to get a head start on storage virtualization (the next frontier).

And why call it xVM? To make sure everyone knew we weren’t simply targeting Solaris – xVM virtualizes Microsoft Windows, Linux (Ubuntu, RHEL, all other distros) alongside Solaris (8, 9 and 10). Customers can consolidate those operating systems, and similarly consolidate their hardware infrastructure – and use xVM OpsCenter to manage and maintain the whole plant.

This week, we’re unveiling a full line of desktop to datacenter virtualization offerings, covering desktop virtualization (xVM VirtualBox), datacenter virtualization (xVM Server), high scale management (xVM Ops Center), and Virtual Desktop support (xVM VDI and SunRay). All endorsed and supported by the industry, and all in use by some of the most powerful customers on earth.

And to that end, I’d like to offer my thanks to the customers who were present at that event a few years ago, and offer my sincere congratulations to the teams involved in bringing xVM to market, across Sun and our partner community.

With all the celebration around xVM, perhaps our next customer event should be held in Champagne…


Filed under General

34 responses to “Of Wine, Virtualization and xVM

  1. wariola

    way to go Jonathan.. cant wait to try xVM

  2. Nice stuff 🙂
    Hope you succeed at what you’ve envisioned.
    BTW, since you’re also focusing on the desktop now, how about coming out with a decent mobile platform? A SPARC laptop "from Sun" would be great for most people running [Open]Solaris on AMD64 laptops.

  3. Innovation pays. It will take Sun a few years to get to where it was but your getting there. keep on innovating.

  4. I’ve been reading your blog since you first put it up. You’re a very dynamic writer, and I love reading your posts. Usually, I just read, but I couldn’t help but mention something here.
    Xen, as awesome and stable as it is, is not the future of virtualization. What I hope you realize, is that by creating xVM Server on top of Xen, you have opened up your corporation to supporting a 2nd operating system besides Solaris. Maybe this bothers you, maybe not. After all, Xen is just a hypervisor, so there shouldn’t be much to support. However, it still is additional support, which means additional energy and income.
    Now, switch gears to turning your existing kernel into a hypervisor, such as KVM is doing to the Linux kernel. By taking this path, you keep one operating system running, namely Solaris, utilizing the power and technology of your existing kernel. You only support one operating system. You save time, energy and money. Everybody wins.
    Now, don’t get me wrong. I’m not putting down Xen, or your decision to run it. Xen brought about Open Source virtualization, which was sorely needed after VMWare took the reigns. VirtualBox is nice tech too, of which I run daily on my laptop. I’m just saying that Xen has seen its days. It was necessary to get us into the virtualization market, but it won’t last, due to the decisions it made regarding its implementation.
    You’re putting a lot of eggs in one basket, so I hope it turns out good for you. Good luck.

  5. Alex Lam

    Out of interest – any roadmaps for VirtualBox to have everything under an Open Source license? The current state (http://www.virtualbox.org/wiki/Downloads) looks a bit funny to me, to say the best.

  6. Alan

    As a Sun shareholder whose shares now trade for the price of a good cheeseburger, I again feel the need to express my opinion.
    8 of your last 15 blogs directly had to do with free software. I understand the ‘halo-concept’ you are using with this free software. We used a similar tactic in the airline industry too. Travel agents using your reservation system tended to sell your flights. Likewise, businesses using your free software may eventually buy your servers. I get it.
    But I believe this approach only works with larger companies. Except for OpenOffice, smaller companies don’t have as great a need for your free software, so another approach is needed for the SMB market. Sell them other hardware they need right now like switches, firewalls, routers, mobile devices, etc. so they become familiar with the Sun brand. When they eventually need their first server or another server, they think of Sun instead of Dell and HP.

  7. >BTW, since you’re also focusing on the desktop now, how about coming
    >out with a decent mobile platform? A SPARC laptop "from Sun" would be great
    I would buy a Sun laptop with a multi core SPARC processor in it. I’m also wondering where those Sun Ray laptops went? I’ve heard about them but can’t buy one.

  8. To Comment from Alan

    Alan, I’m not sure there is a recipe for success in the world you describe. Yes, Sun sells to companies that will pay for software/services, and those that don’t can use a lot (not all) of their technology for free.
    But if Sun didn’t provide the technology to those firms, someone else would beat them to the punch (as MySQL is doing to Oracle) in SMB, and out maneuver them in big customers. So I think their strategy makes a lot of sense if people care about the IP (and a lot of people seem to).
    The real question is in the current environment, does anyone’s strategy look like a recipe for success. Dow down 500 points, ouch, even GOOG’s at a two year low.

  9. Opteron

    I have only one thing that I do not understand with SUN and I guess Jonathan can help me understand it. I may be dumb but at least I can read a clear explanation in English. Here is my question:
    How can SUN have unquestionably the best server technologies on the market (verified by third-party benchmarks and reviews) and its stock has been trading at or below the revenue price since 2000? I’ll also note that during that time all your competitors did well with their junk boxes. Some of them are no better than assembly shops with no real R&D budget and SUN spends $2 billion in R&D.
    Now a comment:
    When I was a graduate student doing research I was using SUN workstations but when I left the academia I never entered an office where SUN trademark can be found and I have been using Windows and Intel boxes ever since! Does that mean that SUN does not understand that after training students on their tools they should use the opportunity to ensure that they continue using SUN technologies after school? Yes, SUN can offer cell phones, desktops and other hand-held tools to the mass market not to make a lot of money but to be present all the time if their technology has to endure. That’s a marketing technique that SUN does not get. Imagine that in a connected world Java is running on billions of devices for free and SUN has no presence on the mass market? I don’t know how you guys will ever make money. Are then doing R&D for the sake of it?

  10. Bob

    When are you going to reduce the company to just Fowler’s group, you know, the one that makes all the money? Sun should be a 12,000 person company at this point, at best.

  11. Isaac Rabinovitch

    You run five different desktops? I’m one of your more technical employees (senior tech writer, x64 West) and I only run three. Inquiring minds want to know which five!

  12. VMware x’ed…:) "n" is for new monitor in the SCSA lexicon…

  13. Terry Hunt

    The only SPARC laptops I am aware of can be found here:

  14. Bill W.

    Sun could challenge mighty VMware in the desktop.
    Oooops!! i see we joined with them. Well it’s all for the better.8>)

  15. systems employee

    Hey Bob – the guy who said we should reduce the company to Fowler’s group? I work on that team. So let’s see, we could do without the Microelectronics team, and that’d jettison $3billion in gross margin dollars, and $6billion in revenue. And we could do without OpenSolaris, and that’d kill of the rest of the business. Wow, using that logic, we could be a lot smaller, you’re right? We wouldn’t even exist!!!

  16. Guido

    People doesn’t want open source software. They simply don’t care about it, they don’t care about technology. Outside USA, apart very big companies, there’s plenty of companies that don’t take advantages of Open Software. They’ve a little experience as end user with pc, and they reflect this as technology for their business; so there’s no innovation, very low functionality, higher costs. Bottom line, we should enter the end user market by educating end user with new products. Sparc laptops is an example (as someone already posted), imagine it with fully (and only) open software and great hardware support. Being able to use Sun, Open Software at home, they will be start looking for high level solutions for their companies too. so could finally speak about technology they want.

  17. Kevin Hutchinson

    What will be Sun’s response to VMWare’s Virtual Datacenter? And congrats on your excellent http://www.kenai.com project management site – awesome stuff!
    @Isaac Rabinovitch, my guess is that Jon runs Mac OS X, Ubuntu Linux, OpenSolaris, MS Windows XP and MS Windows Vista. Java SE 10u6 needs to look great everywhere.
    PS GregP hasn’t blogged since January – does he still work at Sun or did his red-shift fall into a black hole? 😉

  18. Ryan,
    Both Accutech and General Dynamics make Sun Ray laptops. 3G connectivity even. Accutech makes the Gobi7 and Gobi8 and GD (they acquired Tadpole) makes the Meteor. And just so we are clear on branding, Sun Ray is two words. 🙂

  19. @Aaron You seem to misunderstand Xen. It basically does turn your existing operating system into a hypervisor; the "administrative UI" for xVM is Solaris.

  20. Peter

    In reply to Opteron,
    I used a Sparc Classic during Uni, not long after Graduating, I came across Ultra 80’s in Engineering, CAD & FEM, yes there are Windows boxes everywhere, no, they’re not as good as the Unix boxen, just good enough and cheaper, accounting and management were easy pickings for salesman. Today I use Windows, Linux and Solaris while developing in Java for cross platform portability.
    Unix became marginalised due to clever marketing and low up front purchase costs of high maintenance Wintel gear, but that platform is nearing the end of it’s life cycle, it’s just too costly to maintain for MS and customers alike. Later Linux ate into Unix’s server market share, open platforms have less restriction to cross vendor pollination and migration. Open Source is better marketing, it appeals to accounting with lower up front costs, management get’s the vendor independence and tech’s love its unrestricted access, today businesses are squeezed by high costs of power and maintenance, labour shortages and downward pressure on budgets which is impacting Windows market share.
    Perhaps Sun took a little long to react to market changes, however I wouldn’t discount them yet, recent innovations, coupled with open source, Sun’s gaining traction, while Sun’s Market share has expanded, their revenue has remained flat due to server consolidation and virtualisation, but instead of resisting it, they’ve embraced it.
    Most of the Microsoft platform developers have now learnt .NET, Windows is loosing client market share to smart phones and other platforms, transitioning to Java is easy, its the largest OS platform today, its everywhere you look. Java is one of the least difficult platforms for developing multi threaded apps, the new concurrency libraries are great. Sun’s riding the crest of the next hardware wave with Niagara. More innovation, openness, customers and service contracts, more income? Must be time for more advertising?

  21. nick

    Keep up the good work Jon and everyone at Sun. I can’t wait to see VirtualBox mature some more; with all-in-one solutions that technologies like this offer I can see a bright place in the computing market for Sun.
    I’ve been running OpenSolaris for a while now and have seen it progress in leaps and bounds. There was a number of things that sold me on OpenSolaris. It has the commercial backing of a large company; it provides unrivaled technologies such as Dtrace and ZFS (ZFS is the bomb by the way); it has the ability to run virtualisation software such as VirtualBox.
    With the last point – this is very important. I’m a student, which means that I inevitably have to use a range of operating systems. If I am able to run them all from the same desktop then I think that’s a win.
    For those who don’t get the open source idea, perhaps this may clear things up. Open source provides a loyal community backing which can not be bought no matter how much money a company has. This then will of course lead to the popularisation of the technology among the computing industry and allow it to grow further. It also allows the adaption and interoperability of the technology to other areas, which will of course aid in the products growth. A product grows enough and you then have an industry standard.
    If I may comment on OpenSolaris for a minute though. I know this isn’t really the right place to do this but who knows, it may be read by Jon or someone else high up in the food chain and be taken notice of. OpenSolaris, to get the real community backing, needs a better method for community packaging. Look at openSuse’s Build Service, Arch Linux’s AUR, FreeBSD’s ports service, all excellent services. Sunfreeware.om and blastwave are all over the shop. Most of the packages are outdated and are compiled for the likes of Solaris 8.

  22. Carles

    Me as customer have 2 servers at home for doing my developments.
    Me as shareholder have had invested several times in this company and now I see this company very very very bad. Someone has said Dow lost 500 points, GOOG at 2 year lows, Sun is at more than 10 years lows now, beeing its main bussiness big companies like banks, for example, the perspective for the following quarters is really bad.
    Sun needs to reinvent itself, leave server as a division and focus on cheap products for the general public like AAPL did with iPod and i-Something.
    StarOffice (theorically the extended version of OpenOffice) is for FREE with Google Pack. OpenOffice is FREE. OpenSolaris is FREE. NetBeans is FREE. Those products can be sold for a low price in stores and win market share to competitors!!! Please, it is not a non-profit organisation but a company which is focused in making money!!! Or at least me as an investor (when I owned sun shares) was expecting that. Share price at this moment is less than 3 dollars above pre-reverse split share price.
    Anyway I guess that all this is part of a strategy, but I’m not sure if it will benefit investors or general public or neither of two…
    I wish the best for sun workers. I wish for Sun the best too, please Jonathan, make this company shine again.

  23. Another Jon

    I’m a big fan of Sun. I like the products, I like the employees, I like the culture. I like Sun so much, I bought a bunch of stock a few months ago in my retirement fund (ouch). I really, really want to see you guys succeed.
    We use some of the Sun software at my company. Java, IDE’s, etc. Aside from the fees paid for the JavaOne conferences, I don’t think we’ve ever paid Sun a nickel to Sun to license anything. Maybe someday that will change, but for now our license budget is allocated to other companies just on the momentum of legacy platforms. If I could, I would gladly redirect more of that expenditure towards Sun. You guys rock.
    Jonathan, I’m sure everyone and his brother has suggestions for the secret sauce for the Sun turnaround. I’m no different. I have a suggestion. I’d like you to succeed. When you win, I win as a user and as a shareholder.
    Sun needs to increase the number of internet hosting providers offering Solaris in their technology stack. Oh sure, there’s a few, but not many, and their are not as price competitive as Linux. Try Googling for hosting services offering Solaris, Java, and Glassfish in their offering at a good price. You won’t find many.
    I know it is a catch-22. If there were customers (demand), there would be more providers offering the service (supply). But without a supply of providers offering that stack, no one is going to be developing software for those technologies to create the demand. It’s a marketing race condition.
    What I’m saying is you need to prime the pump. Create a program to offer the hosting companies a huge incentive to make the offering. There are several things you can do that won’t cost Sun much money, but will build market share. Offer pre-built servers with Solaris, Java, Glassfish, etc. to the hosting companies for exclusive use for customers using Glassfish, Solaris, etc. You can offer them free for a year or two provided they sign a pre-set number of customers for these platforms. The host providers will market it heavily because without the overhead, they will make more money. You can also offer free (or discounted) classes to their administrators. Whatever. But the idea is to increase the user base. Spread the word. Build the user base. Increase adoption.
    I understand your strategy with free and open source. I even agree with it. For it to succeed, though, you need to grow the user base to a critical size, and you’re not there yet – not even close. Whatever you’re doing now isn’t doing it fast enough.

  24. Peter

    Carles comment on StarOffice, it’s quite a good marketing strategy, sold next to Microsoft Office on retail shelves, next to all those new PC’s, exposure to non tech types who haven’t heard of StarOffice or OpenOffice or get confused with OfficeOpen formats marketed by Microsoft. That would have an impact.
    There are a large group of consumers, who, while very likely to buy StarOffice/OpenOffice off the shelf, are also the least likely to download due to malware / virus fears, it’s a trust issue.
    In fact, I’d investigate boxing up other Open client software (including VirtualBox) and sell it via retail channels too. You might even consider extending this as a form of Sponsorship for Open Software outside of Sun, by packaging and on selling other open software via retail chains and channeling revenue back to those projects.

  25. Brazilian Java Developer

    Lots of companies are making millions of dollars everyday creating and maintaining huge java applications here in Brazil. Why doesn’t Sun do the same? Why don’t you guys offer development and consulting services here?

  26. enkidu

    Hi Jonathan,
    as long as You are on top of this company I would like to see starting three things.
    1. Take SUN off the stock-exchange. Make the company independent of so-called analysts and only money-minded people. Looking at the bankruptcies in NewYork it is evident for everybody how manipulated this market was. That will not change.
    2. Make Solaris a full scale alternative for the consumer market, by supporting all needed hardware, developing and supporting all needed software. Even if You take the strategy of Apple by designing fully integrated Products. Think about a cool-thread Niagara-Laptop 🙂 Can You imagine how many poeple would gladly buy it.
    3. Take more care about all valuable SUN-employees, that You don’t loose experts to competitors and Your knowledge base doesn’t erode. Knowledge is the most important asset of SUN.
    Best Regards

  27. Thanks and welcome back Jonathan. This is really all very relevant information.
    Recently,I had massive problem with the new Windows Office 2007. I had brought some work from office and the word got corrupted. I tried to work on same files in my desktop and the compatibility problem did not allow me to edit in the Office XP.I got some clarity about the working of different operating systems after reading this post.
    My son is insisting on telling you that he has cracked the Lime wire pro version.Hope Sun is not affected by the recession in the market.

  28. Sun wont change

    Unfortunately, Sun management seem to have made a classic business mistake
    -come up with an idea (Open Source) .. good
    -checked it was possible to implement it…good
    -asked others (customers, downloaders) if they liked the idea…good
    -FAILED to check whether it would make money…BAD, VERY VERY BAD
    Simply repeating a mantra of "Open Source will save us/make money" will not make it happen. Yes, it’s good for the "community", but history shows it’s not made money for Sun.
    All companies have to make money. Even Google, free to many (like myself, who uses their search engine) have to SELL things.
    As to Sun’s (in)famous "culture", it seems similar to working for a charity – getting paid to help "the community" regardless of whether you make money.
    I suspect this is going to end in tears for many, if not all, of Sun.

  29. Bill W

    Just read this and would like to pass it on… Sun is a great company who needs to get it’s message straight. There’s a range of new and innovative ideas coming out of Sun if they can just delivery a clear explanation to investors on how it all ties together.

  30. Raghu

    I like the work you are doing to bringing up the business, I would like to offer you a suggestion why can’t Sun come out with Intel/AMD desktops/Laptops? by this Sun will have diversified products portfolio and still have an edge with technology with sparc/storage/Virtualization/java/..etc
    By only concentrating on Server solutions you are loosing a big opportunity to play/lead the other sector of the market.
    Thanks about it.

  31. softwareman

    Mayuresh, regarding you comment about sparc laptop, check out http://www.tadpole.com/ Tadpole laptops have been around for many years…

  32. @Ceri Davies- No, I teach Xen on Linux for a living. I don’t misunderstand it. Xen does not turn your existing operating system into a hypervisor. Xen is the hypervisor that runs on bare metal, then your operating system is virtualized in Xen as "Domain-0". Additional operating systems are installed as "Domain-U"s. Xen is a completely different operating system than your Domain-0. Xen has done a good job at blurring the line between what is running on bare metal, and what is virtualized, however. Thus, I can see why you would say what you did. However, you are technically incorrect.

  33. Kevin

    Here’s an idea for Sun. How about partnering with Akamai? So you’d blend xVM with Network.com to deliver managed cloud services, and have Akamai provide the acceleration and resilience. I believe you both already have good relationships with SAP for example. Imagine providing SAP hosting at Network.com? Oracle Apps etc could follow too? Sun + Akamai = Enterprise-grade AWS

  34. Z

    I believe the thin client time has come, there are millions of pcs and macs out there that are waiting to be replaced (Hmm…, pc guy, mac guy, can be joined by ray 🙂 )… they consume less power, less IT overhead, less upgrading requirements, security…
    Thin client is a big oportunity for sun, try not to miss it …

Leave a Reply to To Comment from Alan Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s