Anything But a Flash in the Pan

There are only two kinds of storage devices – those that have failed, and those that are about to fail. That’s the view most datacenters have about the traditionally mechanical devices pejoratively referred to as "spinning rust." All disk drives fail, cheap drives fail faster.

If the average time to fail is five years, you and your laptop can make do with the occasional backup. But when an average enterprise has 100, or 1,000, or increasingly 10,000 or 100,000 individual disk drives, failure is a daily, if not hourly occurrence. Mechanical devices fail.

And with failure comes the potential for losing data – using commodity disks to save your boss $500,000 does her no good if she’s fined $50,000,000 for violating data retention regulations. Stock transactions, medical images or feature length movies – take your pick, some data has to be perfect. Not a decimal point or pixel out of place.

That’s exactly why, years ago, Sun invented a storage platform called ZFS. ZFS makes a powerful assumption – that a reliable system must be built from unreliable parts. By using surplus computing cycles, ZFS constantly runs powerful integrity checks, giving data corruption no place to hide. With ZFS, customers can use the cheapest disks and simplest systems, and get exceptional data integrity, along with massive reductions in cost and complexity.

But there’s a new option on the table, known to many by the memory cards they use in their phones, iPods or digital cameras – called Flash memory. Flash is very fast at reading and writing data, like DRAM (the memory chips in your computer). Its price sits squarely between DRAM and traditional disk drives. But unlike either alternative, Flash requires no power to remember data. And with the price of electricity escalating across the world, keeping 10,000 disks spinning at thousands of rpm can cost you in power what you pay for your storage. Power has become the dominant factor in high scale hardware decisions – and Flash is set to disrupt the industry.

Historically, there have been two impediments to using Flash in the enterprise.

The first was cost. Flash is more expensive per gigabyte than a comparable disk drive. But with every increase in the cost of electricity (and decay in the price of Flash memory), Flash’s relative cost per available gigabyte is quickly improving – remember, disk drives have to be powered on to be available.
And although a gigabyte of mechanical disk might cost you less than a gigabyte of Flash memory, the latter is at least an order of magnitude faster in reading and writing data – so the cost per gigabyte served is exceptionally low.

But simply introducing Flash as yet another tier of storage in a datacenter isn’t the real opportunity – that adds new costs and a set of new management hassles. To truly change the industry, adding Flash would have to be completely transparent to users and operators, alike, with no switching or operational cost. And that’s exactly what we’re doing with ZFS. ZFS will transparently incorporate Flash into the storage hierarchy of a running system, using the microprocessor cache for the most performance sensitive tasks, DRAM for the next, then Flash, then disk (then ultimately tape). ZFS will allow Flash to join DRAM and commodity disks to form a
hybrid pool automatically used by ZFS to achieve the best price, performance and energy efficiency conceivable. Simply put, our storage and server systems will get enormously faster – without any upgrade to the microprocessor. Adding Flash will be like adding DRAM – once it’s in, there’s no new administration, just new capability.

That’s one reason we’re so excited by Flash – the cost per available gigabyte served (the total operational cost of storage) plummets with Flash in the mix, especially for data or performance intensive applications (like MySQL, Postgres, Oracle or SQL Server). From the right systems designs, Flash has the potential of providing orders of magnitude improvements in economics and performance – and with the advent of Sun’s xVM hypervisor, we can bring this performance benefit to any host operating system (running atop xVM, Windows operators can inherit the benefit of ZFS+Flash, too). 

The second problem is trickier – simply put, although Flash memory can be read an infinite number of times, writing to Flash more than a few hundred thousand times can wear it out. Now, most normal humans will never hit 500,000 writes in a digital camera. But you might in your enterprise. What to do?

Again, ZFS to the rescue.

ZFS treats Flash memory like any other storage medium – remember, all storage devices fail – and manages data integrity whether failure’s induced by a bad hard drive motor, write-fatigue or a hammer drill.

Increasingly sophisticated "wear leveling" algorithms are also maximizing Flash lifespans – evening out write activity to avoid hot spot failures. But the bottom line is this – with ZFS at the helm, wear is a non-issue (on hard drives or Flash – they both have wear limitations, after all).

These are the premises behind Sun’s systems approach to Open Storage. We’re integrating ZFS, Flash memory and some exceptional hardware/silicon innovations to deliver high performance, low power, general purpose storage and server appliances – accelerating any software that runs on our SPARC or x86 systems (MySQL users, in specific, will see a big turbo charge). For a fraction of what proprietary NAS storage ordinarily costs. Expect our first Flash systems to ship toward the end of this calendar year.

And as you might expect, ZFS and all the underlying software will be free without commercial support – OpenSolaris, ZFS, MySQL and Postgres are already available (click the image at left to get a free LiveCD – or check out Apple’s release of ZFS in Mac OSX). Software revenue will come from those enterprises that want Sun’s technical support for mission critical deployment. On the hardware side, Sun’s Try and Buy programs allow any partner or customer to order one of our systems for free 60 day trial – if you love it, buy it. If not, we’ll cover the return postage.

If the above doesn’t make it obvious… our view is Flash is anything but a flash in the pan – as the price of power continues to increase, and the price of Flash continues to plummet, the combination of Flash, ZFS and true systems innovation will have an even bigger impact on datacenter economics than virtualization.

It’s that big a deal.

UPDATE:  You may have seen we added yet another company to the list led by Intel, IBM and Dell supporting Solaris (and thus ZFS) as OEM partners – this morning, we officially announced the addition of Fujitsu-Siemens. Congratulations, all…


Filed under General

41 responses to “Anything But a Flash in the Pan

  1. ilovezfs

    This is so cool. I hope I can use this flash technology on my Mac someday.

  2. Sun Employee

    This blog is what your voice should be, J.S.
    You make money on the Hardware, you should passionately write about the Hardware…THAT’s where the money comes from.
    The posts about open-source, software, vision makes me yawn. Write more about your money-makers…BABAY!

  3. Brian Scully

    I’ve been a Sun shareholder for the past 2 months. I am a firm believer that your strategy is perfect. You are doing a wonderful job.
    ZFS, OpenSolaris, DTrace, Glassfish, MySQL…the list of Sun software brands just grows. Companies are shifting their resources to open source at a phenomenal speed. I am a CIO at a small company and appreciate the support plans behind these products.
    Can’t wait until your detractors are proven wrong!

  4. Daniel McDonald

    I have to agree with Brian. I am a firm believer in Sun Microsystems current strategy. From a consumer’s standpoint, there is little risk in purchasing Sun systems because the software I deploy on the server does not lock me in to a particular vendor or even operating system. There is not even a proprietary look and feel they are pushing. They have made all their work freely available. When buying a sun server, the worst case is that it is simply a server that is innovative, fast, and works well with other equipment. Likely case though is that I now get to deal with one systems vendor for hardware (servers and storage) and software (simplification, yeah) and suffer NO negatives from my commitment. Sun can handle all your SYSTEMS needs without fear of being limited in the future.

  5. taipang

    I have been a SUN shareholder for the past 7 years – and have been in negative territory since..
    I have faith in SUN’s technology and you to turn the company around. And that is what is holding me back from selling my shares (though not a big chunk)
    But it gets a little irritating that SUN tech support cannot grab 600MB core files from SUN’s ftp site – and keeps on coming back to me to ask me to upload to the http site.

  6. Kevin Hutchinson

    Agree completely with the above posts (Brian and Sun) – hope you can execute on time. xVM is going to prove crucial to Sun’s future, so I hope you can get big adoption and kudos as early as poss. Keep going mate, you’re doing the right things!

  7. Miguel Perez

    Since you arrived, my image of Sun has improved considerably. Not only that, but your blog is a good read and you make good points.
    Almost every software technology I work with is free software, so I’m not interested in "big enterprise crippleware". However, free software runs on hardware, and this is where Sun comes. I’m far more inclined to buy hardware and trust software technologies from Sun if I see they are as open as possible. A CEO who actually understands the technology I work with and writes insightful articles which are not a buzzword soup of enterprise babble straight out of magazines is a very good sign in this direction. This is already impacting my decisions and recommendations with regards to Sun products, and I couldn’t have picked a better day to say this because I’m just about to install two of our new Sun Fire servers.

  8. Ewen

    While ZFS is an important innovation and a complete degression from the traditional way of thinking about storage systems and management; I still think that it is an "in the works" type project. There are a lot of things that it CAN do, and that it is designed to do, but at the same time, there are a lot of things that it CAN’T to (yet). I’m currently using it only out of necessity, but at the same time weary about what it can’t do. I think that it would help if some of your employees write a "best practices" for implementation and also for preparing a system as a ZFS file server since the demonstrations and the knowledge of HOW to use ZFS would help users like myself make the best out this wonderful technology. Otherwise, it’s a missed opportunity that someone else will fill.

  9. David Holden

    >>By using surplus computing cycles, ZFS constantly runs powerful integrity checks, giving data corruption no place to hide.
    surely using "spare" cycles runs counter to the energy efficiency claims?

  10. Dunstan Vavasour

    Fitting Flash into the stack at the correct point.
    We’ve seen interest in flash from the storage vendors as well. But, of course, if you’re EMC then the logical place to use flash is in the storage subsystem, meaning your data still has to go through the volume layer, down onto an HBA, trundle across a SAN, possibly via a storage virtualisation engine, through the storage controller and into cache … and then the acknowledgement has to make the return trip to the server before the transaction can be marked as committed.
    ZFS is the clever bit. As far as the application is concerned, it’s writing a data block using a filesystem API. ZFS can then stash the write onto Flash and return without involving any other layers. It will be interesting to see real life performance comparisons of server based flash versus a fast storage subsystem.

  11. JC

    I’m not a shareholder but a SUN user in Brazil, and follow SUN and OpenSource paths for about 12 years, and I also believe that the strategy is perfect.

  12. David Emery

    Apple announced this week that the next release of OS X Server ("Snow Leopard", X.6) will support ZFS.
    (p.s. for a ‘technology blog’, it’s probably appropriate to add the base indication for the simple math question. 5+35 could well be ’43’ if interpreted as octal or ‘3A’ in hex. Me, I now report my age in hex, 34 hex sounds a heluva lot better than 52 decimal 🙂

  13. BuzzCut

    I’m just an investor, but I don’t understand why you care at all about free software when all the money comes from boxes. No real business is going to use open source products when you have corporate options. And instead of writing this blog, you should spend your time fixing the share price.

  14. @BuzzCut
    What on earth makes you think that "no real business is going to use open source software"? Business is *already* a big user of open source software. You think that Linux and Solaris aren’t used in corporations? Or MySQL? Or Java? Or PHP? Or Firefox? Or WebKit? Or Eclipse? or Netbeans?
    The question you *should* be asking is – What business is going to use *closed* source software when there are genuinely compelling open source options available?

  15. davefromnz

    +1 Simon
    And, go Sun!!

  16. @Buzzcut:
    The opensource strategy makes sense in many areas, but from a business investors perspective look at it like this:
    All the upcoming 20 somethings starting to work in the enterprise will have 5-10 years experience with Linux or Windows, why would they choose Solaris when they can use what they know? And why use Sun hardware when they can use Dell or HP which target MS/Linux? By opensourcing, it encourages young people at home to go and learn and play with ZFS, Dtrace and whatever else Sun cooks up.
    Sun will become the only player in the game who offers a truly open COMPLETE stack with no vendor lockin. For me, going with Sun is a no-brainer – there’s almost no downside and plenty of upside.
    It’s similar to McDonalds targeting children with happy meals, playgrounds.. get ’em while their young and allow them to grow up with McD…except Sun isn’t as evil 🙂
    or Microsoft allowing their software to be pirated for 10 years and then suddenly clamp down licensing, people now prefer to pay the premium than go and learn something else… except Sun ain’t so evil.
    Now hopefully Sun won’t ever do a Microsoft to us, but they now offer a VERY compelling technological stack / solution.
    They just need to get everyone to hear, hear..

  17. Mark Halverson

    As an investor, Sun is my largest holding. Hence I am distressed with stock price and earnings/revenue deterioration. Is Sun’s "free software" business model working? I have no idea. Impressive new software is reported daily (it seems) by Sun. But never is there an inkling as to quantity of imcome generated per monies spent on software development, so we can intuit if business model is working. In short one cannot see the forest (where bottom line is headed) for all the trees (constant new software announcements). Pease enlighten us re the bottom line.

  18. W. Wayne Liauh

    When you listen to Jonathan talks, you think Sun is a great company, full of hope, full of energy and full of potential. And if you, like myself in the development of Open Solaris/OpenSolaris, have had a chance to be somewhat commingled with Sun’s developers, you very likely will develop a strong conviction, an urge to do everything for this company.
    However, somewhere in between, that’s where you may feel your bubble gets pinched, burst. Most of Sun’s management/executives have their hardware quotas to be filled. As far as I can tell, very few, if any, in Sun’s higher hierarchy give a damn about the open source movement.
    Tell me which of Sun’s employees at the "executive" level is using OpenSolaris or Solaris X as his/her primary OS? Mr. Schwartz included. Somehow such a showing of "strong" interest within Sun begins to make me feel like a stupid idiot.

  19. To Mark Halverson

    Who is your financial advisor? Sun should NOT be your largest holding. that was a BAD idea.

  20. Way to go! Understood this ZFS thing. The approach sounds smart, energy crunch is the next big hassle world will face. Products low on energy consumption will soon be the preferred ones. This ZFS maps with reduce-reduce-reuse principal. This Sun is cool! Does not add to the ‘global warming'(oxymoron?). Thanks for the knowledge you share.

  21. Mike B.

    We are a small Sun Partner.
    We’re facing problems selling Sun products to MSWindows only shops.
    When we are selling the whole package as a mangaged service (think appliance), we don’t face any huge problems.
    Maybe Sun should repackage their stack and sell e.g. Storage Appliances, Desktop Virtualization Appliances etc. for companies not having any Unix know-how. This is something that others have done for years with success (Like spamfilters, firewalls running on Linux Appliances).
    For the rest of us, we can still build our own solutions.

  22. Kevin Hutchinson

    Regarding ZFS, MySQL and your other excellent open source software, how come Sun doesn’t have the equivalent of Redhat Network (RHN) where I could pay a subscription to run an update program to keep all my Sun open source software patched and up-to-date? I’d pay a simple yearly or monthly subscription and know that my software was patched and secure. If you pitched it at a reasonable price, with a simple download for all modern OSes you’d get a lot of interest I’m sure.

  23. Billy

    A couple of questions/observations:
    First, Several reviews of systems using flash-based memory have indicated performance (write speeds, in particular, but also read speeds) that place flash SSD’s below rotating disks. For example, see this Computerworld article:
    Performance showdown: Flash drives versus hard disk drives
    Second, is the power consumption advantage of SSD being overstated in considering life-long cost? A $200 1TB Western Digital GreenPower drive uses 7.5w at maximum (during read/write). Assuming it’s used 24x7x365, it would use 65.7 kwh per year — roughly 197 kwh over a 3 year life. At 10 cents/kwh (approx. the current national average commercial electricity rate), the drive would cost roughly $20 to power over its life — approx. 10% of its purchase price. Even taking account for the cost of cooling (not incl. in the 7.5w spec and expected increases in electricity costs, it seems that energy costs do not constitute a substantial part of the operating cost?

  24. VoC

    Hello Mike B – that is a fascinating and valuable comment. If I read it right, it seems that a more current product presented in a bundle similar to the former Cobalt products may be of interest to you. What would you think about something like an "appliantized" version of the Sun Fire X2100 M2, low cost rack server? Do you think something like that would be of interest to the companies you are supporting?

  25. Mike B.

    That’s basically what we’re doing. We are setting up small X86 servers with our own Solaris build and Sun software-stack. Integration is done by us onsite.
    The missing part is a front end, which would enable a customer to be able to integrate the solution by himself. Just like any other appliance around there…
    This is something that Sun should look into. The technology is certainly there…

  26. @Billy — it’s true that SSDs offer little to no advantage for sequential I/O. Where they shine is random IOPS, which often limit the performance of databases and file servers. The better SSDs combine Flash with DRAM and a supercapacitor to buffer synchronous writes. With samples of such devices from one major vendor, we’ve measured 10,000 IOPS per SSD, versus around 200 for a fast mechanical disk drive. Random read latecy is similarly excellent, at around 100us rather than the 4ms or so typical of hard drives.
    It’s easy to achieve a 50x improvement in throughput — just have 50 things running in parallel. But a 50x improvement in latency is much harder to come by. When it does, it really changes the landscape.

  27. Toby

    What are your going to do to fix your sales issues? SUN has wonderful technology but people don’t know about it. That’s my experience from the field.

  28. ilovessd

    @Jeff Bonwick:
    Most database applications (and OS filesystems) are optimized for sequential access to non-volatile memories — it is by design that data are arranged contiguously and sequentially. The probability of fully randomized access patterns is still quite low, so SSDs would show improvements a small percentage of the time. Would cost benefit suggest little overall value with SSD (given current prices/capacities)?

  29. Sun User

    ZFS is definitely interesting technology. Even more exciting is that multiple vendors are adopting it- Apple has announced it will be using ZFS in Mac OS X! Way to go!

  30. Customer

    As a customer owning two 9990 arrays and finding myself unable to locate anyone who can provide the password to open this file: (go ahead, download it and then try to open it.)
    I have to ask, will you provide documentation for these flash devices to your customers?
    I’ve opened a support case re. this issue and spoken with our sales rep. No luck so far.

  31. Denis Vilfort

    Right on Jeff!- SSDs do not spin.
    Let me say that again: SSDs DO NOT SPIN! So no rotational latency, no seek latency, no motor that eats electricity. The only similarity to HDDs is that FLASH retains data after we turn off the electricity. We must therefore be careful not to evaluate FLASH based technology with a HDD mindset.
    Performance characteristics of FLASH based data retention technology – like Solid State Disks (SSDs) – are vastly different than that of mechanical disk. A mechanical disk has to spend time moving its actuator arm to the desired track (seek latency) and then wait for the rotating platter to come around with the desired sector (rotational latency). This ‘latency tax’ is paid for every IO.
    Average rotational latency is a function of how fast the platter is spinning, whereas average seek latency is a function of how fast the actuator arm’s voice coil is moving the read head in place. A 15,000 RPM drive will have an average rotational latency equal to half a revolution per time unit. A 15K RPM drive will spin 15,000/60 = 250 RPS so the time it takes for one revolution of the platter is 1/250 = 4mS. Then average rotational latency is half that: 2mS. Add to that the average seek latency (example here: ) of say 3.5mS we get 2mS + 3.5mS = 5.5 mS total latency tax paid when reading a 512 byte sector from rotating platter. This tax puts a limit on how many 512 byte random reads we can accomplish each second.
    Moving the data takes time too. Again the vendors is sharing specs of HDDs with us: 125MB sustained for sequential reads. The trouble is that modern server workloads are rarely sequential in nature. They are highly random. The servers read and write behavior is a function of all client/server requests and the many parallel procesees running in the CPU complex.
    So how many IOs can a HDD do? We know the average seek latency. But how long does it take to move a 512 byte IO across a the wire? If we take the 125MB/S number and express this in terms of 512 bytes, we would expect 125,000,000/512 = 240,140 IOs assuming no other overhead. Each IO would then take 1/240,140 = 4uS (roughly). The disproportional misalignment between bandwidth and latency is what disqualifies HDDs as sensible option for high performance data retention and retrieval. The latency tax is per the HDD vendor’s own data is 1375:1 !
    This is indeed the dirty little secret in the HDD industry. A mechanical disk drive cannot perform random IO well! A 512 byte random read from its mechanical platter takes an averaage 5.504 mS equal to a measly 181 IOs per second. Not good when we have disk interfaces that can consume data at 140,000 – 300,000 IOPS per HBA. To satisfy just one SAS 1.0 HBA would take the aggregate IO capability of 733 disks! The power consumed would be 733 * 17.4W = 12,754Watts. And at $.10 per kWh running these drives equals a yearly cost of 24*365*12.754*.1 = $11,172. Nuts!
    So how about FLASH? – How many SSDs will it take to do 140,000 IOPS? And how much power will that consume?
    A modern FLASH based SSD consumes 2.5W and sports 50,000 read IOPS. So it will take only 3 such devices to saturate a modern SAS HBA. Power spent: 7.5W. Yearly cost: $ 6.57. It is easy to see why SSDs are the right choice here. SSDs are 1,700 times more cost effective than HDDs for the same performance.
    ROI is also in SSDs favor. Thanks to consumer electronics, FLASH cost is dropping 50% to 70% year over year. By the end of this year, a 32GByte enterprise SSD is expected to sell for $1,200. Since we only need 3 SSDs for 150,000 IOPS and can save $11,172 – $6.57 = $11,165 in one year, the SSD cost of $3,600 will break even in less than 4 months.
    FLASH based SSDs are emerging as the performance choice for IT. HDDs are still offering the best capacity cost. So when we are talking SSD vs. HDD, I suggest we look at the whole picture:
    300GB capacity, 181 IOPS, $500 list. 17.4W
    $/IOPS = $2.76, $/GB = $1.6, IOPS/W = 10.4, GB/W = 17.2
    32GB capacity, 50,000 IOPS, $1,200 list. 2.5W
    $/IOPS = $0.02, $/GB = $37.50, IOPS/W =20,000, GB/W =12.8
    The SSD clearly is the smart choice for best performance cost – both CapEx and OpEx. HDD should be used only when CapEx per GB is the criterion. I would think twice when looking at HDD OpEx. A few more months of falling FLASH prices combined with increasing electricity cost and HDDs are no longer attractive.

  32. Billy

    Dennis, you definitely make some strong points when addressing applications where data in small chunks are written/changed very infrequently and read on a random basis very intensively.
    I’m curious where DRAM would rate in the metrics that you’ve used for HDD and SSD? A DRAM-based SSD (or maybe just a very large, well-managed cache in a storage system) would have the disadvantage of requiring battery backup, but would have the advantage of being able to address applications where data changes more frequently (i.e. the "write fatigue" issue described by Jonathan).
    I’m also curious about what % of the enterprise storage market this performance segment constitutes?

  33. To Customer

    i couldn’t open the file either. usually documentation problems are corrected at the next release of the documentation. hopefully this is not one of those types of documentation problems. did you try couldn’t hurt to attack this from multiple angles if necessary.

  34. Wes Adams

    Regarding the posts made by:
    "Customer" June 13, 2008 at 09:24 AM
    "To Customer" June 14, 2008 at 04:54 PM
    You are trying to access an outdated location/file. Please contact and we will direct you to the appropriate resource.
    Best Regards,
    Wes Adams
    Sun Microsystems, Inc.

  35. Mark

    Interesting. Definitely gives me some new insight on flash disks. But at the same time, if flash replaces disk at the near-line storage level, will power-down SATA drives replace tape at the offline archival level?
    I would be much more interested in hearing what Sun plans around archival storage.

  36. For write-biased IOPS performance, don’t forget that the DRAM/supercapacitor
    combination may soon give way to phase-change PRAM, which sits between
    DRAM and Flash in the memory hierarchy. Coming soon from Samsung
    and Intel (and invented by Stanford Ovshinsky, who also innovated
    nickel-hydride batteries for the sadly-cancelled GM EV1), you can be sure
    that Sun is evaluating nonvolatile PRAM chips, now in sampling … Oh, and ZFS
    can undoubtedly handle this too!

  37. ilovessd

    @Denis Vilfort:
    Your arguments for SSD seem to be based on changing flash technologies, but no changes in HDD technology. HDD performance and energy efficiency have drastically increased with 2.5" drives. Many of your arguments require flash (of the future) to be compared to 3.5" drives (of the past).
    Furthermore, there is your explicit assumption that random IO performance is a highly valuable metric. The problem with this argument is that most software (databases, filesystems) are already designed to sequentially access data in non-volatile memories (e.g. prefetching from HDD), while random access patterns are used in volatile memories (why it’s called RAM 😉 The behavior of non-volatile memories have been taken into account in software for ages now, this is not a new discovery.
    The advent of 64 addressing allows even more non-volatile memory to be cached in volatile memory — and volatile memory is almost always used for applications where random access patterns happen most frequently (transaction processing).

  38. Software Guy

    About ZFS on Mac OS X. Does Sun provide ZFS phone support to Mac user ? Or Apple pay license fee to Sun ? Or you expect Mac user will buy a Sun box due to ZFS ? What is your revenue plan ?

  39. Jan Flodin

    @buzzcut, Mark and other shortsighted investors
    I guess you dream that every company around the world will buy Sun products and the stock price will increase. And then you look at the hardware which right now generates the revenue and suggest that Jonathan and the rest of Sun put focus "where the money are".
    What you don’t understand is that if Sun not had the Open strategy and such wonderful technologies as Solaris and ZFS then Sun would lack the most important key differentiators and a customer would have less reasons to buy a Sun server and the revenue would drop. Also you should remember that technology buyers are conservative. Nobody gets fired from buying IBM or Microsoft is still a thruth – well, in fact everyone buying IBM and MS should be fired because they buy the wrong solutions compared to what Sun can offer with respect to usability, administration, cost of ownership and any other relevant metric – Sun’s only problem is that they are Sun and have to fight uphill against the more dominant players in the industry (dominant in size, but not in mind).
    My true belief is that Sun has the right vision, but as you might now it’s not always the best and the right that wins, especially not if the competitors has more money. Anyway my advise would be to take a long term position in Sun, because the product vision by far is the most compelling and logical in the industry and in the long run IT buyers simply has to realize this – they can’t afford not to.

Leave a Reply

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

You are commenting using your 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