The Observation Deck

Search
Close this search box.

The sudden death and eternal life of Solaris

September 4, 2017

As had been rumored for a while, Oracle effectively killed Solaris on Friday. When I first saw this, I had assumed that this was merely a deep cut, but in talking to Solaris engineers still at Oracle, it is clearly much more than that. It is a cut so deep as to be fatal: the core Solaris engineering organization lost on the order of 90% of its people, including essentially all management.

Of note, among the engineers I have spoken with, I heard two things repeatedly: “this is the end” and (from those who managed to survive Friday) “I wish I had been laid off.” Gone is any of the optimism (however tepid) that I have heard over the years — and embarrassed apologies for Oracle’s behavior have been replaced with dismay about the clumsiness, ineptitude and callousness with which this final cut was handled. In particular, that employees who had given their careers to the company were told of their termination via a pre-recorded call — “robo-RIF’d” in the words of one employee — is both despicable and cowardly. To their credit, the engineers affected saw themselves as Sun to the end: they stayed to solve hard, interesting problems and out of allegiance to one another — not out of any loyalty to the broader Oracle. Oracle didn’t deserve them and now it doesn’t have them — they have been liberated, if in a depraved act of corporate violence.

Assuming that this is indeed the end of Solaris (and it certainly looks that way), it offers a time for reflection. Certainly, the demise of Solaris is at one level not surprising, but on the other hand, its very suddenness highlights the degree to which proprietary software can suffer by the vicissitudes of corporate capriciousness. Vulnerable to executive whims, shareholder demands, and a fickle public, organizations can simply change direction by fiat. And because — in the words of the late, great Roger Faulkner — “it is easier to destroy than to create,” these changes in direction can have lasting effect when they mean stopping (or even suspending!) work on a project. Indeed, any engineer in any domain with sufficient longevity will have one (or many!) stories of exciting projects being cancelled by foolhardy and myopic management. For software, though, these cancellations can be particularly gutting because (in the proprietary world, anyway) so many of the details of software are carefully hidden from the users of the product — and much of the innovation of a cancelled software project will likely die with the project, living only in the oral tradition of the engineers who knew it. Worse, in the long run — to paraphrase Keynes — proprietary software projects are all dead. However ubiquitous at their height, this lonely fate awaits all proprietary software.

There is, of course, another way — and befitting its idiosyncratic life and death, Solaris shows us this path too: software can be open source. In stark contrast to proprietary software, open source does not — cannot, even — die. Yes, it can be disused or rusty or fusty, but as long as anyone is interested in it at all, it lives and breathes. Even should the interest wane to nothing, open source software survives still: its life as machine may be suspended, but it becomes as literature, waiting to be discovered by a future generation. That is, while proprietary software can die in an instant, open source software perpetually endures by its nature — and thrives by the strength of its communities. Just as the existence of proprietary software can be surprisingly brittle, open source communities can be crazily robust: they can survive neglect, derision, dissent — even sabotage.

In this regard, I speak from experience: from when Solaris was open sourced in 2005, the OpenSolaris community survived all of these things. By the time Oracle bought Sun five years later in 2010, the community had decided that it needed true independence — illumos was born. And, it turns out, illumos was born at exactly the right moment: shortly after illumos was announced, Oracle — in what remains to me a singularly loathsome and cowardly act — silently re-proprietarized Solaris on August 13, 2010. We in illumos were indisputably on our own, and while many outsiders gave us no chance of survival, we ourselves had reason for confidence: after all, open source communities are robust because they are often united not only by circumstance, but by values, and in our case, we as a community never lost our belief in ZFS, Zones, DTrace and myriad other technologies like MDB, FMA and Crossbow.

Indeed, since 2010, illumos has thrived; illumos is not only the repository of record for technologies that have become cross-platform like OpenZFS, but we have also advanced our core technologies considerably, while still maintaining highest standards of quality. Learning some of the mistakes of OpenSolaris, we have a model that allows for downstream innovation, experimentation and differentiation. For example, Joyent’s SmartOS has always been focused on our need for a cloud hypervisor (causing us to develop big features like hardware virtualization and Linux binary compatibility), and it is now at the heart of a massive buildout for Samsung (who acquired Joyent a little over a year ago). For us at Joyent, the Solaris/illumos/SmartOS saga has been formative in that we have seen both the ill effects of proprietary software and the amazing resilience of open source software — and it very much informed our decision to open source our entire stack in 2014.

Judging merely by its tombstone, the life of Solaris can be viewed as tragic: born out of wedlock between Sun and AT&T and dying at the hands of a remorseless corporate sociopath a quarter century later. And even that may be overstating its longevity: Solaris may not have been truly born until it was made open source, and — certainly to me, anyway — it died the moment it was again made proprietary. But in that shorter life, Solaris achieved the singular: immortality for its revolutionary technologies. So while we can mourn the loss of the proprietary embodiment of Solaris (and we can certainly lament the coarse way in which its technologists were treated!), we can rejoice in the eternal life of its technologies — in illumos and beyond!

38 Responses

  1. I’ve been down this road before. Today feels much like the day that I realized that my beloved Tru64 was no more. There is pain. There is sadness. But like you point out, there is still joy.

    Software does not die. It may no longer be a commercial product, but it lives on in our hearts. It lives on in our memories. It lives on in the lessons it taught us.

    However it may have been shaped by humans, we must not forget that the best software turns around and shapes those who use it. Would I be who I am today without SunOS and Solaris? No. They were my teachers, but more importantly, they were my mentors. They taught me more than I ever realized.

    When I used Solaris, I wasn’t just using a piece of software. I was using the combined wisdom of hundreds of the most talented individuals around.

    During these trying times I am reminded of the words of a fine poet:

    Pay my respects to grace and virtue.
    Send my condolences to good.
    Give my regards to soul and romance,
    They always did the best they could!
    And so long to devotion.
    You taught me everything I know.
    Wave goodbye, wish me well,
    You’ve got to let me go.

    Thank you to all who have worked on Solaris. You have made me a better person.

    Your friend in computing,
    Timmy

  2. RIP Solaris.
    It was my favourite OS so far. ZFS, DTrace, SMF, FMA and Zones. What an amazing bunch of terrific technologies included in an OS! And every of them was easy to learn. I will miss you, Solaris. Only time will tell us what we lost today. For me, and for many of us, this is a sad day.

    I want to take this opportunity to thank you all, Bryan Cantrill, Adam Leventhal, Brendan Gregg, Mike Shapiro, Garrett D’Amore, Darren Moffat, Jeff Bonwick and every other talented engineer who was able to create this masterpiece called Solaris. SUN Solaris.

  3. I left the Solaris team a year and half ago since I knew this day would come but I really wished I was wrong. However, I feel honored that I was part of the team that created the best operating system that has been built so far. It’s just a shame more people did not realize that.

    1. We missed you Bill! I’m not worried about the people I worked with, they are all amazing and talented. They will find much more satisfaction, respect and money at a different company. I’m sad that Oracle was so short-sighted to kill the best OS currently available. It was the one thing they could have used to get an edge on Amazon in the cloud wars. Now it will be a race to the bottom, and I can’t see how O can win that one.

  4. More than 20 years ago, I decided to build my career on Solaris. Linux wasn’t really on the map, Windows was boring mainstream, but Solaris was different and exciting. Same with SPARC. Big Irons, robust software, where the really important stuff runs. That’s where I wanted to be, and that’s where I went. After my current excursion into Linux territory, I hoped to return fully to the Solaris world soon. Seems like Oracle pulled the rug out beneath my feet now.
    RIP Solaris, and thanks for the great ride.

  5. It’s another sad day, almost as much as Sun extinction day.
    Though, thanks to Sun and thanks to you guys, we have illumos.

    Still, I’m concerned.
    How long will we have support for Solaris kernels (illumos only from now on) from third parties?
    The world is moving fast towards virtualization, and if you have solutions based on illumos, you need to be prepared to install on VMware or anything we may have tomorrow.
    How long will they support the “Oracle Solaris 11 64 bit” VM? We already have the same issue with Hyper-V, which I don’t care, until a customer wants me to put my solution there or just leave.
    Just few days ago, MongoDB dropped support for Solaris.
    Not very important to me, but signals.

    What’s your opinion on this?

    1. I think that there are two separate questions here: (1) what of illumos itself? and (2) what of the application binding to illumos-based systems? To me, these are separate because I have long come to accept that the Linux binary interface has become the de facto binary interface — and I have believed (and continue to believe) that the OS must support this interface natively. (To me, this is like BSD vs. System V interfaces back in the day — and Solaris pioneered systems that provided both.) So I believe that the presence of LX branded zones allow us to refocus on the system itself — and focus on the system as a container hypervisor, a domain which I believe is a great fit for illumos.

      1. But why in this case use Illumos (SmartOS), if you can use native Linux? Containerisation / Hypervisor projects in Linux are developing rapidly, making a bet on Illimus today you can stay behind tomorrow.
        I really bitter to say this, but de facto all Unix like OS are dead, Linux killed everyone.

        1. Because in Linux you can’t get full isolation, since it doesn’t exist. Therefore, to get full isolation, people run containers inside of virtual machines on Linux. Apart from being really clumsy, it incurs a performance penalty for no reason other than wanting to use Linux.

          1. >> Because in Linux you can’t get full isolation, since it doesn’t exist.
            Yes it is.
            But Linux and the entire ecosystem around it is developing very fast. Previously, Linux did not have a counterpart to dtrace, but now it has eBPF. @brendan already said about it in 2016 (http://www.brendangregg.com/blog/2016-10-27/dtrace-for-linux-2016.html)

            Linux has a unquestionable advantages:
            * The largest community.
            * The largest budget, it is supported and developed by hundreds of companies (Red Hat, Sles, IBM, Oracle, HPE ….)
            * No problems with drivers / hardware, now any manufacturer of HW claims that it is supported by Linux
            * At this time, many engineers have heard about Solaris, BSD, AIX, HP-UX only from stories, the share of these systems is getting smaller every year, but the share of the Linux market is growing every day
            * Linux binary interface will be fully supported in Linux, but I doubt that there will be no problems in Illumos implementation.

            It is a pity that the source codes of the solaris were opened so late and that the community could not gain a critical mass.

        2. a) you get to use ZFS, DTrace, FMA, SMF, …, which Linux doesn’t have or doesn’t do as well,

          b) you get a better host kernel too.

          There are downsides. Mainly that Linux is still, and will continue to be, a moving target, and that Linux is full of stupid things you have to end up copying in Illumos (or whatever host OS you want to use).

          If it were me running a cloud, I’d try hard to use Illumos. It sure is appealing on an aesthetic level too. But I know it would take a commitment to developing Illumos, and at most companies (though certainly not Joyent/Samsung!) that would be an insurmountable negative. Of course, using Linux sometimes means the same thing, but since it’s more mainstream that’s an easier pill to swallow.

  6. “To me, these are separate because I have long come to accept that the Linux binary interface has become the de facto binary interface”

    De facto as it might be, but both you and I know SmartOS can run all those applications natively (thank you Joyent for ~15000 SmartOS native software packages). Then comes the logical question: if I can ran the same freeware, open source application natively on SmartOS in the joyent branded zone, other than a transitional measure, why do I need lx_branded zones again?

    On an unrelated note: I want to become an expert in illumos kernel debugging with mdb. Do you take on apprentices?

  7. Any sense of how support will be managed during the death throes? Security patches, slowly-rising ‘extended’ support costs, increasingly insistent nudges to migrate off the platform, etc?

    Certainly new feature development is toast.

    Thanks for this post, and this work.

  8. I have to agree with dhelios. Linux won. The best course would be to try to incorporate the best features of Solaris into Linux itself. The world would be better off.

    Sun died way before the Oracle acquisition. It was never the same after the dot-com bust and the end of the workstation business. For those of us who were there, we will remember our comrades and our victories and try to forget the politics and stupidity that killed the best company in the Valley.

  9. Which is mostly why I have at various times over the years have suggested that Microsoft, HP and IBM open-source the source trees of VAX VMS, OS/2, and Windows NT (plus of course the inevitable MS Windows 3 and 9x source trees.). There are a lot of dedicated hard-core Windows hobbyists out there …

    I was pleased when Sun open-sourced Solaris. I’d so wanted a SunOS box when I started computing, but of course, couldn’t afford the price … I liked OpenSolaris, but couldn’t get used to the differences between it and Linux … I’ve been thinking of getting used to those differences since I realised the OpenSolaris source tree was still in development.

  10. It’s a shame Solaris has gone this way. When I think what could have been?
    BSDs were arguing in the 90s. GNU-Linux hadn’t really come of age (2000).
    OpenSolaris was just too late in 2005.
    GNU-Linux does well because it is GPL, as opposed to BSD (foundation for propriatry products).
    Sun could have GPL licensed in 1997, riden the dot-come wave/bubble!
    What could have been eh?
    Look at how Microsoft has reinvigorated itself from the old days of “embrace and extend” and “Ballmer: Linux is a cancer”. They now use Github, can run Ubuntu binaries, make millions from GNU-Linux licencing and Outlook and Office on Android.

  11. Linux won, OK, however it didn’t win in base of quality or features. There is still no native linux filesystem that rivals ZFS, and DTrace is still awesome.

    There is probably still some place left for a propietary UNIX, but my bet would be starting from – and building upon – FreeBSD, pretty much like Apple and SONY did for the desktop and gaming respectively.

    All in all, Solaris did last for a while and it was nice to see it opensourced before disappearing.

  12. I feel as if the last 15 years (yes, going back to our days together at Sun) have been a “death of a thousand cuts” for Solaris. As clumsy and brutal as it was, perhaps this is the kindest cut – I hope this brings more of the extraordinarily capable Solaris team’s considerable skills to bear on worthy projects and products.

    But, of course, I can’t help but be sad, reflecting on what might have been. And, in all candor, I wish I had found a way to do more, back when it might have counted.

    Thank you for a fine and fitting obituary, Brian. All the best to you.

  13. Solaris was, in my opinion, the finest piece of software engineering the world has ever seen. In 2010 (at least in my mind) that distinction changed to illumos.

    I actually think this development is probably a good thing. The sooner Solaris is no longer viable, the sooner the collective conscious will hopefully stop thinking of illumos as a “kind of Solaris” and begin to recognize it for the distinct and amazing project it really is.

  14. It is an incredibly sad day.

    Even as the on again, off again maintainer of a linux distro for the past 13 years I still when I had a choice predominately deployed Solaris for all workloads that mattered.

    As the mindshare dropped after Oracle took over (and support for (open)solaris in opensource projects fell, to the point of quite a few even rejecting patches) the effort of porting across required software increased. But the rewards at the end in terms of system reliability and application uptimes mostly outweighed this cost.

    Now we finally got support for a decent c++ runtime in both the forte compilers and supplied by the OS, and the OS goes over the cliff.

    So we are left now with taking all the bits solaris did so well ( and which were so polished) and trying to cobble them together (or emulate them) into the mish-mash known as GNU/Linux.

    Vale the good ship solaris, and all who sailed on her, you will be missed.

  15. Sun made so many mistakes… Canceling Sun PS in 2003, briefly canceling Solaris on x86 in 2002, not making a deal with Google in 2001, the MySQL acquisition, and so many other poor, poor decisions.

    Yet Sun had the best engineering culture, one that allowed Sun to produce revolutionary technologies — most as yet unmatched in over 15 years since! — such as ZFS, SMF, DTrace, FMA, and others, along with a rock-solid kernel, and unsung-but-amazing pieces like the run-time linker.

    Working at Sun was a dream come true. Working in a sea of brilliant people was very fun indeed, and an invaluable education.

    RIP Sun and RIP Solaris, but long live Illumos!

  16. The remorseless corporate sociopath also derided cloud technologies as marketing ‘vapor’, as it severed Sun’s cloud BU in 2010. Today, the devil claims to define the ‘next generation cloud’, where it could have already been standing toe-to-toe with the likes of Amazon AWS and trumping efforts like Google Cloud Platform’s TPU infrastructure. Paraphrasing LBJ and Bill Joy, it’s certainly fun to be outside the tent, or ‘elsewhere’ …. Scott, thanks for creating the culture!

  17. As the last Senior Product Line Manager of Solaris (my position was eliminated exactly 10 years ago, at least that is what they told me then) I still feel sad for all the people who kept the faith and decided to stick around and concerned about how Oracle has dealt and will (not) deal with the installed base. The writing was on the wall years ago when I made the mistake of checking my voicemail while on vacation and having to deal with a firedrill in operations because a junior director of support services wanted to call a stop ship of the new release (stopships are for major defects) because he did not want to have his department pay to have media kits shipped to customers with a support contract. The a..h… did not realize that this was an obligation to honor what is in the contracts that funded his paycheck and that of his entire staff.

  18. I ca understand the emotion and disappointment of the end of the road for something like Solaris. However I have to take exception to some of the statements about Oracle here (Disclaimer I work there .. still) The decision was commercially sound – the market does not value Solaris as it once did and as we move to CLOUD the OS offers nothing for us to differentiate – in the CLOUD no one cares about the OS … which is not quite the same as for the hardware as ironically that part of SUN may well turn out to be a competitive factor against the likes of AWS – however the truth is Oracle in the CLOUD will not sink or swim on our ability to win the IAAS wars – its the SAAS war that will define us – specifically how successful we are against SAP with ERP CLOUD and SF.COM with the CX CLOUD.

    1. In an effort to shy away from the way this decision makes me feel, I’d instead like to respond to what Mr. Winter has posted here.

      Over the past 5 years Sparc/Solaris SuperCluster has seen quite a bit of success, having been used to migrate many customers from other vendors and bring them into the Oracle fold. The optimizations made to SPARC and the Solaris OS over the years, and particularly in the area of database performance enhancement built in to the Silicon specifically for Oracle RDBMS, have proven to out perform any other Oracle database machine on the market, and with the advent of the latest set of cloud deployment tools for SuperCluster, and it’s entry level counterpart MiniCluster, provide Oracle with systems that are ready to deploy Cloud based customer RAC/RDBMS and HA application environments with a few clicks of a mouse, tear them down and return resources to the general pool, indeed even to halt them and relocate them to other server pools and dynamically add/remove CPU, memory and IO resources on the fly.

      Performance and feature wise, these systems beat anything else Oracle has to offer. Therefor the logic behind this decision is not based on “no one cares about the OS”. Logic dictates that it must be based on other factors that I do will not speculate on in public, and the same goes for my views on the methods used to implement the RIF.

      That Oracle’s future is Cloud oriented there can be no doubt; but my question would be why waste all of the years of development that culminate in this set of products just because someone in the Executive Circle prefers Linux on x86 over Solaris on SPARC?

      In my view it has to be a decision based on the KISS principle: How can Oracle do this as cheaply and as fast as possible vs. provide truly excellent performance and manageability in the Cloud? The bottom line is Oracle cares more about “Their Bottom Line” and winning in The Cloud then they do about how they get there and the customer experience.

      Who can blame them? Their primary motivation is profit and return on share holder value. Such is our market driven, corporate run society these days, like it or not.

      Namaste

  19. As I sat here reading this having been an Operating Systems Ambassador at Sun for many many years all I can think of is….

    “And all because he lost a boat race and wanted to take his frustration out on somebody…”

    But never forget “Innovation happens elsewhere” — Bill Joy.

    And the Network has become the computer.

  20. If there`s one thing that stays with me forever, then it is:

    “Remember, behind every cloud, there`s Sun”

    Thanks to the engineering and support teams, it was a pleasure working with you guys and wherever the future is going to take us, those days (and nights) will always be a fond memory.

Leave a Reply

Recent Posts

November 18, 2023
November 27, 2022
October 11, 2020
July 31, 2019
December 16, 2018
September 18, 2018
December 21, 2016
September 30, 2016
September 26, 2016
September 13, 2016
July 29, 2016
December 17, 2015
September 16, 2015
January 6, 2015
November 10, 2013
September 3, 2013
June 7, 2012
September 15, 2011
August 15, 2011
March 9, 2011
September 24, 2010
August 11, 2010
July 30, 2010
July 25, 2010
March 10, 2010
November 26, 2009
February 19, 2009
February 2, 2009
November 10, 2008
November 3, 2008
September 3, 2008
July 18, 2008
June 30, 2008
May 31, 2008
March 16, 2008
December 18, 2007
December 5, 2007
November 11, 2007
November 8, 2007
September 6, 2007
August 21, 2007
August 2, 2007
July 11, 2007
May 20, 2007
March 19, 2007
October 12, 2006
August 17, 2006
August 7, 2006
May 1, 2006
December 13, 2005
November 16, 2005
September 13, 2005
September 9, 2005
August 21, 2005
August 16, 2005

Archives