Sun Isn’t Relevant to Startups

Preface: I don’t write critically about a company unless I think they have some glimmer of getting better, however small that may be.

I just watched another Scoble Show interview with Jonathan Schwartz, and it reminded me how frustrated I continue to be with Sun, particularly their Startup Essentials program. He seems to be complaining about the perception of startups that the Sun platform is completely wrong for a new generation of startups, a perception I would strongly agree with.

To illustrate, here is a timeline of my experience with the Startup Essentials program.

November 3rd, 2006: Drinking the kool-aid at Startup Camp, register on behalf of Automattic for the Startup Essentials program.

November 8th, 2006: I receive an email addressed to “Mr Muellenweg” (wtf?) saying “Due to the volume of interest in this exciting new program, it is taking us longer than expected to review your application. We will let you know your status within the next few days — thanks for your patience.”

Late-November: I talk for about an hour to a nice guy named Eric Cosway doing research about what I thought of Startup Camp and Sun. I tell him about the trouble I’ve had registering for Startup Essentials, he says he’ll note that and pass it on.

3 months since: Complete silence, punctuated by frequent Sun press releases and AP stories about how they’re great for startups, probably millions of dollars of PR.

Seeing their constant press releases and hot-air filled announcements around “Web 2.0” just makes Sun seem that much more out of touch with reality and desparate to jump on hype which has already passed.

I’m not writing this, as Joyent did, with the hope of connecting with someone at Sun that will take my money. I’ve given up, I’ve lost hope, and our business has moved on. We’re growing fast and adding infrastructure faster, all on a tight budget, and I’m less inclined to depart from the LAMP-based architecture that Digg, Flickr, Wikipedia, and dozens of other sites I respect use.

This is also what I’ve started advising other startups, if we had trouble getting anything going I can’t imagine what trouble a brand new startup with a smaller profile would go through. (An exception to this is going through Joyent, which has packaged Sun’s stuff so well Sun should just buy them.)

Personally I’m far more excited about what Amazon is doing these days.

Update: In a move most would never expect of a public company CEO, Jonathan Schwartz has responded. More than that, he has apologized in a human and personal way that is utterly admirable. I’m going to work on a followup to this entry.

Update 2: I’ve posted the followup.

66 thoughts on “Sun Isn’t Relevant to Startups

  1. A number of “old names” have been dropping by the wayside as they struggle to try and keep up with the “modern 2.0 Web.” I say these in quotes because in a lot of ways they still don’t get it. They loved the day when startups poured thousands down on expensive setups that where either way more then they needed, or not flexible enough to change. I won’t say Sun is one of these businesses… but they seem to be showing that.

    Modern startups seem to be better grounded and less spend the money with no return in sight.

  2. The dangers of doing so much talking that one stops listening. To be followed, in short order, by no longer being relevant.

    With the rise-and-rise of LAMP based systems that are really starting to become plug-and-play by nature, Sun has already lost a lot of startup business.

    Your post is one of an increasing many I’ve read suggesting Sun don’t seem to be at all interested in selling hardware to startups, despite the hype, or levels of cash on the table.

  3. Hi Matt, before I got to the punch line of your post I was thinking that I should talk to you. I would love to discuss the ways that you can use EC2 and S3 for WordPress!

  4. Sun stopped being important years ago. x86-based hardware rapidly closed the performance gap between normal desktop hardware and Sun’s overpriced workstations, Java never ended up selling more of that Sun hardware, and Solaris was sent to the UNIX retirement home once Linux became the hot new kid on the block.

    What compelling products does Sun have anymore? The products that defined them, Java, Sparc and Solaris, are fading or have faded. There’s not much left to Sun but the shine of its former glory. It’s a rudderless boat trying to fight the tide of open standards and interoperability. You’d think they would have learned something from IBM.

  5. I’ve also always been fascinated by the inability of businesses to take my money. They spend millions marketing to me, then totally fumble when I want to close the deal. How many times a day do you stand in line to pay someone? Doesn’t that bug you?

  6. Jesse, in their defense they do offer x86 hardware, some of it at pretty reasonable prices. Solaris is more interesting since they open-sourced it, but honestly I think we’ll just wait for ZFS to be ported to Linux.

  7. “Web 2.0”, huh? Welcome to 2003, Sun!

    At this rate, they’ll market SOA in 2010. By 2020 they might even start talking about BI.

    How the mighty have fallen…

  8. Sun’s got a huge hill to climb to get their hardware and Solaris back on the radar of startups.

    ZFS, though, is going to be relevant to a lot of people, especially startups wanting to store huge volumes of content on dirt cheap hardware. Once it’s on Linux, ZFS is going to clean house of ext3 and reiser.

    I’m not sure it’s going to have a big impact on Sun’s bottom line or adoption of Solaris, but they’re still innovating – and shipping the results.

  9. I sympathize with the issues you are having with the Sun program, but my company uses Sun in a number of key areas and we are very happy with their support. As for the prognostication of some of the comments, Sun’s recent number point to just the opposite of the uninformed comments here. Java is in fact on more devices than all the major operating systems combined. In addition, Sun’s T2000 server just won server of the year award for their new CMT technology. Finally Sun is innovating like crazy not only in SPARC and Solaris, but also with their X86 based systems, which have also been winning awards right and left.

    So again, hopefully Sun will get their act together on the Startup program, but folks, please do your homework before making generalizations.

  10. I worked at Sun for a while. I think that Sun got lucky twice (what’s the old saw? “Better to be lucky than good anyday”), once with “The Network is the Computer” and again with their positioning (Telcos) for the first dot-com boom. I believe their now desperate to try to get lucky again and are shovelling ideas out there to see what sticks.

    The mistakes in strategy (Dawdling about Solaris on x86 for 15 years, focus on “Big Enterprise Iron”), a total failure to understand what customers wanted (Scott once ripped someone who called Sun a “solutions company” and told him that Sun’s a technology company), a failure to get any viable software strategy going and the leakage of their good people have really crippled that company and many business units are adrift, disconnected from the other parts of their business, which compounds the problem.

    Sun’s a good company, with some great people and I really would like to see them do well but I’m afraid that they have little expertise and ability to execute.

  11. Yawn.

    I don’t get people who bitch about Sun, but don’t get near the marketing attention from say a Gentoo or Canonical. Most who use Gentoo or Ubuntu simply download it, use it, and self support it. But Sun is supposed to come to your house, pull down your pants, and kiss your ass in person? Please.

    I am also tired of people fawning over some killer feature in some OS or distro and switching all of their systems to it, rather than just saying “wait until they port it to my favorite distro”, but holding other OSs to a different standard.

    The double standards become boring, and give you guys zero credibility. You would rather bitch about something than try to improve things.

    “I downloaded it and it wouldn’t install on my XYZ motherboard.” I remember those days on Linux. And back then if you complained, you got flamed on Usenet for being a Linux poser. “Just write a script”, or “If it doesn’t work, write your own driver”. My, how times have changed.

    Download Solaris and try it. You don’t need Sun to come and do it for you. If you like it, use it. But like any OS change, it will take some work to change over to using it. If you don’t like it, then fine.

    ZFS, Dtrace, Zones, Fire Engine IP stack, all awesome stuff. Zones does not get enough attention as a security feature. And more stuff coming like Crossbow. If you want innovation at the OS level, this is it. If you want compatibility with 1,000 of systems, Linux is good.

    And by the way, most Sun x86 hardware ends up running Linux, and running it very well.

    As Bill Maher used to say, “Get over yourself”.

  12. Sun may be most relevant to startups in the following way – people leaving Sun to create startups and put ideas into practice more quickly than they could do while inside Sun. And then Sun sometimes bringing those back into the fold through acquisitions.

    Our founder and myself both came out of Sun and we have a rapidly growing software business for optimization of data-intensive cluster and Grid computing.

    We have had success partnering with them, not always easy, it depends greatly on the individual relationships you form in the business units and field, much more so than the programs.

  13. Hey Mark, I totally agree they have some great technology and some brilliant people there, and I’m not asking to be hand-held through Solaris, but I’d rather they make no promises at all than make promises and repeatedly break them.

    It’s totally possible my case was an anomaly and I was the only one of thousands of startups that didn’t have a good experience with them, but I’m guessing it’s indicative of a deeper problem, a disconnect between the tech gurus and leadership at Sun and the people behind these programs.

  14. Matt,
    I agree with Mark’s advice to you. The Sun horse is good
    enough if you are a good enough Jockey!


  15. I was administering a few sun/solaris/sparc sites in the dot-com times, and felt like they were pretty good- not responsive- we hosted our databases on them (cold fusion apps/apache web servers). The few PCs with MS/SQL were more problematic and difficult. I felt like that’s kind of decadent now- that we even had competing systems. The Solaris decision – it was more expensive getting developers, but the geek & free community were far more involved, than the MSFT one. At that point in time MSFT was really not scalable.

    I had more tangles with Oracle, as customer support, licensing, weird installations, stuff like that- really – than sun.

    This isn’t from a hardware perspective, though my boss managed that aspect and he did grumble quite a bit. I knew a few other startups that were buying Sparcs, or had them lying around, but they were ex-employees.

  16. The problems with Sun marketing vs technology don’t end with startups. Over the years we have pumped a lot of money into Sun and for several reasons will keep doing so on a small scale, but Sun needs a major reality check. ZFS and zones are wonderful but they’re just starting to get ready for production now while Sun has been pushing it for well over a year. Most companies want proven technology and LAMP has proven its worth. Solaris 10, not yet. The techies will admit this, but marketing is pushing it hard.

    It just happens too often that Sun comes up with a solution looking for a problem. Sorry, it’s a beautiful solution but when it doesn’t solve _my_ problem I _don’t care_!

  17. Mark: Sometimes it would be nice if Sun would deign to sell things to small customers though. The incredible amount of hassle I had just trying to buy some servers from them was immense. It took weeks for me to find someone who’d fill the order.

  18. Matt, of course every situation is unique and it seems that you didn’t get the particular alignment/support you were looking for…however I can say that our start-up experience with Sun over the last 18 months has been nothing short of fantastic. Best of luck with your new venture.

  19. What has caught my eye recently with Sun is project blackbox. How anyone can say building a datacenter into a cargo container that ranks in the worlds top 200 supercomputers and can be deployed anywhere in the world quickly and easily has missed the point.

    This may not be for every company, but the idea that they can have this on your doorstep possibly within 24 hours, and get your datacenter back online after a complete failure shows they are still in the game.


  20. I run a startup that *WAS* accepted into the SE program. I would say we are going to purchase a Thumper within the next 3 months. There is a 60% plus markdown off the device compared to MSRP. WTF is that? In this day and age, Sun – a company billing itself as the iron behind Web 2.0, still plays the “get as much as you can for it” sales game??? I mean seriously, as if people don’t talk and share pricing or use search engines. Instead we are forced to play this ridiculous shell game. It’s insulting to customers and keeps me from buying more. Put one price up. Make it fair, make some money on it – but for crying out loud, cut this 1960’s salesman deal crap.

  21. Bureaucracy in all its forms can be difficult, and getting nothing but PR hot air for months and an email misspelling your name isn’t the best way to build relationships. But I can say that Sun does have a commitment to assisting startups, I can say that because I work for Sun, and am now responsible for our startup customers in San Francisco.

    The underlying need for any IT technology- be it a server or a LAMP based web architecture- is to create a dependable platform that can provide a service. The combination of technologies is really going to depend on the needs of company, and the users. A large big iron server isn’t going to be the answer to every technical need just as much as a small whitebox server isn’t going to be the answer to every workload demand. That being said, Sun provides some innovative products that truly can assist startups in scaling their business, and making it more dependable. Want to deploy a LAMP architecture? Awesome- Sun has products that can host that at a more dependable and affordable level than almost any other option out there. Interested in deploying apps in a secure, flexible and easily virtualizable environment that’s open sourced? Download OpenSolaris. At the end of the day, technology for technology’s sake won’t help a startup be successful. Technology has to address not just the technical needs of an organization; it has to address the business needs of the organization.

    I feel our products value and price truly do shine- and startup customers using Sun demonstrate this. Joyent does run their Rails architecture on Solaris and have done an incredible job using it as a platform to deploy their services. Those apps aren’t running on Dell servers anymore, they’re Sun:

    I can’t defend the way your case was handled, but I can say Sun is committed to our startup customers- and I can demonstrate its not all just PR fluff

  22. Matt, thanks posting your experiences and frustrations with Sun and our Startup Essentials program. As someone who works for Sun in the OpenSolaris kernel group, it was a difficult post to read, as were most of the ensuing comments. But as sobering as it was to read, I’m glad you took the time to write it and I do hope you’ll take a second look at what we’re doing because I know from a technology standpoint there’s an incredible amount of activity which addresses the needs that startups have today and in their future. However, we’re also aware of how far we need to go. Thanks for the honesty.

  23. Onno,

    You don’t have to use ZFS and Zones to use Solaris 10. If you want proven filesystem technology, UFS is a very, very stable filesystem. And Solaris 10 running with a single, Global Zone (i.e., the default) is exactly like every other UNIX OS namespace structure. The point is, you can use Solaris 10 the same way as Solaris 9, or Linux, without Solaris 10’s advanced features, and adopt the advanced features when needed.

    Solaris 10 was released in February 2005, the same month as RHEL 4. Public early access to Solaris 10, via Solaris Express, occurred a few weeks after the official release of the Linux 2.6 kernel. Solaris 10 is just as proven as RHEL 4, and Solaris 10 kernel is just as proven as the Linux 2.6 kernel.

    Solaris’ development approach drives compatibility and stability first, at the cost of speed of feature add.

    Perhaps Sun marketing is pushing the advanced features of Solaris 10, but certainly Red Hat does the same with those features which differentiate RHEL, Novell does the same with SLES, etc.

    There are good reasons for not using Solaris 10. Compatibility with new hardware and speed of feature add are two of them. Feature overshoot may be another (do I really need a 128-bit filesystem?). Certainly the lack of tools like apt-get and Red Carpet is a big factor. But lack of being proven is not one of them.

    Oh, and all of the complaints about Sun’s support of small customers is fair. The company is just not structured to do that. And because of that, pursuing all startups simultaneously is not a good marketing strategy for Sun.

  24. To digress on a parallel path, I’d think that it would take time for Jonathan Schwartz to steer the sunny ship in the right direction, but guys, let’s give a round of applause to a big old skool gorilla (Sun) for trying to breakdance to hip hop or rock to . I mean, I’d like to, by comparison and for illustration’s sake, see Steve Ballmer & co. blog openly like Schwartz, release the Windows kernel under an OSI license and GPL2 C#2.0. Try getting the same response from HP or IBM. Open source HPUX or AIX or OS/2? Fergettaboutit! Give them time guys. Sun is fuelling innovation – both technically and in the corporate politics space.

  25. I think ‘if you build it they will come’ still holds true to web 2.0 or classic .com start ups. I think Sun is taking a model of a razor they use to determine what in open source gets their money and trying to apply it here wondering why it won’t work.

    Sun has great products, but start up .coms are looking only to avoid being victims of their own success. The biggest reason this happens is due to your hosting platform not being able to scale and adapt cheaply.

    Since you aren’t going to really start monetizing until the end of the first year, cashflow is king and Linux is cheap. Sun can’t help this effort beyond Open Solaris without shooting themselves in the foot.

    The first ingredient for making things work is a good plan. Looks like Sun forgot that entirely and started with the cart before the horse.

  26. I just found out yesterday that the Storagetek (now Sun) disk array I purchased in the not so distant past is not supported any more .. I can’t get drive trays, I can’t get drives…forget it. Sun’s only solution is to replace the equipment we paid in the 5 digits for with gear that will cost even more. I just about lost it…talk about how NOT to look out for the little guy.