Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?
Oracle Businesses Databases IBM Sun Microsystems Hardware

Explaining Oracle's Sun Takeover — "For the Hardware" 154

blackbearnh writes "Brian Aker, former Sun MySQL guy, and current proponent of the Drizzle MySQL fork, gave O'Reilly Radar an update on where MySQL is at the moment. During the interview, he was asked to speculate on Oracle's original motives for acquiring Sun. 'IBM has been moving their pSeries systems into datacenter after datacenter, replacing Sun-based hardware. I believe that Oracle saw this and asked themselves, "What is the next thing that IBM is going to do?" That's easy. IBM is going to start pushing DB2 and the rest of their software stack into those environments. Now whether or not they'll be successful, I don't know. I suspect once Oracle reflected on their own need for hardware to scale up on, they saw a need to dive into the hardware business. I'm betting that they looked at Apple's margins on hardware, and saw potential in doing the same with Sun's hardware business. I'm sure everything else Sun owned looked nice and scrumptious, but Oracle bought Sun for the hardware.'"
This discussion has been archived. No new comments can be posted.

Explaining Oracle's Sun Takeover — "For the Hardware"

Comments Filter:
  • Error 503 everwhere I go!
    • Re: (Score:2, Funny)

      by CoffeeDog ( 1774202 )
      Maybe /. needs some new Sun hardware to run an Oracle back-end on!
      • Sun hardware is pretty much ideal for a site like Slashdot, but I suspect that the cause of the problem is a product that Sun bought and now Oracle owns: MySQL.
  • Hmmmm..... (Score:5, Funny)

    by feepness ( 543479 ) on Friday April 09, 2010 @08:46PM (#31796948) Homepage
    Not for the lulz as I originally supposed.
  • Customers, potential customers, and to stick it to IBM.

    The OS is just a vehicle for a database.

    • For IBM software seems to be a way to sell expensive consulting services. Don't buy it, it just encourages them.
      • "I'm betting that they looked at Apple's margins on hardware, and saw potential in doing the same with Sun's hardware business."

        Yeah, because Sun's target market is Apple's target market. NOT.

        HP, Dell, Lenovo, Acer all look at Apple's margin's and say to themselves "I wish I could get those margins". And the shareholders in those companies should line up all the executives, then each shareholder will kick each executive in the nutsack, and say "Each kick may not be that powerful, but we'll make it up wish

        • "I'm betting that they looked at Apple's margins on hardware, and saw potential in doing the same with Sun's hardware business."

          Yeah, because Sun's target market is Apple's target market. NOT.

          How do you interpret the quote to mean that Oracle want to sell to fashionistas and creative wannabes? It just means they'd like the same margins. The quote would be equally valid if they made ice-cream or frying pans.

          • Except that Sun and Oracle don't look at Apple's margins and want the same, they look at Apple's margins and shudder. Apple's margins are high for the consumer market, but they're not even close to the same league as Sun or Oracle.
    • Re: (Score:2, Insightful)

      They're proving both quotes that 'real men build hardware" and that "real software lovers build hardware" from IBM and Apple.

      Both IBM and Apple design Software and Hardware to complement each other. Compare an iSeries or iPad to the typical Oracle setup where they are at the mercy of Intel, AMD, Microsoft, IBM, etc to get their Database to work. Defining a basic Schema is full of so many tips and tricks compared to any other database. Sure, it's nice to choose the "optimum" setting for every single block of

      • by bertok ( 226922 ) on Saturday April 10, 2010 @12:19AM (#31797936)

        How did this get modded up? I know that it... sounds like it makes sense, but it's the exact opposite of what actually goes on.

        They're proving both quotes that 'real men build hardware" and that "real software lovers build hardware" from IBM and Apple.

        Both IBM and Apple design Software and Hardware to complement each other. Compare an iSeries or iPad to the typical Oracle setup where they are at the mercy of Intel, AMD, Microsoft, IBM, etc to get their Database to work. Defining a basic Schema is full of so many tips and tricks compared to any other database

        WTF? It's hard to define an Oracle schema because of a client's choice of instruction level compatible CPUs? Are you kidding me? I've never heard of anyone actually altering their database schema design to target it for either "Intel" or "AMD". That's insane.

        . Sure, it's nice to choose the "optimum" setting for every single block of data... but wouldn't it be BETTER to simply format the hard drive the way you want it in the first place and to build the most critical functions directly into firmware?

        First of all, it's quite possible to "format the disk" natively with Oracle's database files, bypassing the OS filesystem. Even Microsoft SQL Server can do that [microsoft.com], it's just not advertised as a big feature. Yes, there are performance gains (I've heard up to 20% in some corner cases), but it's almost never worth it, because the downsides are enormous. Managing a LUN is much harder than managing a file. Either way, this can be done now. There's no reason for some sort of magic hardware support.

        Second, somehow 'burning' Oracle in the firmware is neither going to make it faster, nor improve anything else. It'll just make it harder to patch and manage, and it'll mean that a future service pack may not fit into the limited flash space. I can't imagine too many deployments where the speed of the program storage is the limit. Even if it is, it's not like you can't boot-from-SAN or just buy an SSD for any old server now!

        IBM stuff can do really neat things like split database writes in the disk controller and keep track of multiple copies at once on redundant systems.

        Err.. you mean scatter-gather IO [wikipedia.org] and synchronous mirroring [microsoft.com]? Ooo... fancy stuff, I bet nobody's ever managed to do that in software!

        You just can't do that level of stuff with the tools Oracle or Microsoft has now.

        Yes, you can. The differences between the major vendors at the "low level" have been tiny for years and years now. The real differences are at the high-level, pure-software layer. Features like RAC differentiate DB2, Oracle, and SQL Server from each other, not the RAID controllers.

        Microsoft's sole existence is based on separation of hardware and software... so everybody squabbles between Intel/AMD, ATI/Nvidia, Oracle/MySQL, etc... and Microsoft gets rich playing "middleman" being the only party the others can legally talk to.

        Are you kidding me? Since when is Intel some poor pauper holding out a begging bowl to Microsoft? Last time I looked, both Intel and Oracle had market capitalisations over USD 100 billion, and were 'legally allowed' to talk to each other.

        There is already a company that makes a Sparc based blade for IBM BladeCenter chassis, drop it in an IBM Blade and share your SAN and have backplane-level network between the other hardware and OSes....this is what Oracle is after. Rather than keep playing games with other vendors, simply sell "Oracle" like IBM sells System i (iSeries). You would by an Oracle blade and simply connect that to your network. There's no point in loading multiple apps on hardware...
        it's so cheap now versu

        • Except that any company big enough to have a CIO is thinking of "software as a service"... not some buzzword like borrowing something over the internet, but setting up the entire data center under a service level agreement. If they want Oracle... they want Oracle and don't CARE about the HARDWARE OR SOFTWARE to get it.... They want to pay for solutions not "products". My company is pushing everything to IBM at the corporate level. Our equipment will sit in IBM's data centers and the factories will just conn

        • Microsoft dropped the "database on raw partition" in 2005 as I recall, because the performance advantage was just a few percent even if your workload was totally disk-bound. Also, not having an actual file to move around, resize, etc was so inflexible that it just wasn't worth it. Finally, I'm sure a lot of idiot admins set this up, then had some other admin accidentally reformat the partition containing valuable data later on.
  • by Anonymous Coward on Friday April 09, 2010 @08:53PM (#31796994)

    Oracle has been saying that they won't support Solaris on non-Sun/Oracle branded gear. This essentially means that even if 70% of your gear is Sun hardware running Solaris they won't support the 30%, even if that 30% was bought because there wasn't a good fit with Sun gear.

    I've heard the same thing about Java support.

    To add insult to injury, Project Caiman in OpenSolaris is going to force everyone to rebuild a lot of infrastructure and process (for reasons that all seem to point to ego and a complete misunderstanding of how sysadmins actually do their jobs).

    As a result, many companies (including the one I work for) are looking at making the jump to Linux on cheaper hardware. Given some of the other posts (including fanboi's like BenR), we're clearly not the only ones thinking this.

    • by ducomputergeek ( 595742 ) on Friday April 09, 2010 @10:17PM (#31797442)

      FreeBSD is a more likely replacement for some of the Solaris market. Especially since it's had DTrace for a while and now ZFS is now production ready on FreeBSD 8

      • Re: (Score:2, Insightful)

        by hoggoth ( 414195 )

        > now ZFS is now production ready on FreeBSD 8

        I do not think that means what you think it means.

        • by dnaumov ( 453672 ) on Saturday April 10, 2010 @06:04AM (#31798822)

          > now ZFS is now production ready on FreeBSD 8

          I do not think that means what you think it means.

          If ZFS isn't production-ready in FreeBSD 8, it isn't production-ready in Solaris either.

        • Care to elaborate? The ZFS developers on FreeBSD have been running it on their own systems for a couple of years and now trust it enough to recommend it for deployment. It was mostly fine with 7.x, but there were a couple of obscure corner cases where there it had problems, so they didn't recommend it. They don't suggest using it on anything other than x86-64 with over 1GB of RAM, but aside from that caveat it's considered production ready.
          • by hoggoth ( 414195 )

            Sure, I'll elaborate.

            As I understand it (and I could be wrong), on BSD ZFS is several versions behind. That means lots of bug fixes are missing. It is also missing the in-kernel CIFS server, which is key to serving files to Windows systems. It is also missing iSCSI, and de-duplication.

            • FreeBSD 8 contains ZFS 13, which is not particularly old. It's missing deduplication, which would be nice to have but is not a requirement for a lot of use cases. There's no in-kernel CIFS server or iSCSI target, but it can still run Samba and userspace iSCSI targets, so you get the same features just without the performance benefit of having them in kernel.

              I'm not convinced by the argument with regard to bug fixes. Some bugs have been fixed since ZFS 13, but new features have been added so other bugs

      • by wmac ( 1107843 ) on Friday April 09, 2010 @11:34PM (#31797752) Homepage
        I have been an IT manager in a bank and I do not agree with you. OS selection is not just a technical matter.

        Those who have selected proprietary OS will replace it with another proprietary one. If they intend to use open source they will more likely to choose the most popular one.

        We were using HP-UX and Solaris. When we decided to use an open source OS for a particular server farm we selected Linux because that selection is less dangerous politically (it matches the consensus) and Linux market is more diverse and more supported.
        • And notice I said some of the Solaris/OpenSolaris market. Not all. In the past two years we've been working on a project and took a look at Solaris and OpenSolaris for the enterprise version of our software products. We could deploy smaller/medium sized companies who were growing with OpenSolaris on x86 and then if/when they needed the support, they could always move to bid daddy Solaris.

          Then Oracle announced they were buying Sun and having been an IT manager, I absolutely hate and refuse to deal with Or

    • WTF? Oracle have said no such thing. They've changed the licensing agreement so that you would need to buy a support contract if you plan on running Solaris on non-Sun gear for more than 90 days, and that is all. What are you - a FUD machine?
    • don't know why people post about OpenSolaris when speaking of major Oracle project directions, it is a minuscule niche project that isn't used by business. It hardly matters to Oracle's (nor 99.999999% of the world's) future whether OpenSolaris flops or succeeds.

    • Project caiman the installer? I don't follow.

  • Hogwash (Score:4, Interesting)

    by Third Position ( 1725934 ) on Friday April 09, 2010 @08:58PM (#31797020)

    Oracle originally only made an offer for Sun's hardware assets. They only bought the entire company after IBM made a bid for it. That doesn't sound much like Oracle had much enthusiasm for Sun's hardware. Apparently they bought it only because it came with the dinner.

    • Re:Hogwash (Score:5, Interesting)

      by Third Position ( 1725934 ) on Friday April 09, 2010 @09:00PM (#31797028)

      Ugh - make that "Oracle originally only made an offer for Sun's software assets.

      • Re: (Score:3, Interesting)

        by Anonymous Coward

        Ugh - make that "Oracle originally only made an offer for Sun's software assets.

        According to Larry Ellison (in September 2009), they want to sells "systems":

        We are not going into the hardware business. We have no interest in the hardware business. We have a deep interest in the systems business.


        Of course software is a significant component of any system:

        Systems is about eighty-five percent software--if you take out the microprocessor design. Microprocessor design is a complicated deal, it's a very complex component; Sun has a significant team, IBM has a significant team, Intel has a significant team, designing microprocessors. But when you get by the microprossor component, and you're assembling these systems, Cisco--I'm guessing--85-90 percent software.


      • by Sycraft-fu ( 314770 ) on Friday April 09, 2010 @09:45PM (#31797276)

        They are doing a crap job. Why? Well if you buy expensive SPARC hardware, you are going to run Solaris on it. It is the only thing really well made for that architecture. So what is Oracle now doing? Charging for Solaris. Not just charging, but being total dicks about it. You have to have their agreement, if you at any time lapse in the agreement, not only do you not get security updates, you are required to uninstall all the ones you've already installed.

        Hmmmmm... How do I feel about that for critical systems.... Oh ya: Fuck you.

        Seriously, this kind of shit could well kill SPARC. It is a very limited use platform anyhow. If you start screwing people over they may well abandon you for IBM's offerings, or just commodity x86 stuff (which is getting more and more high end offerings all the time).

        To me, it seems like Oracle WANTS to kill off the hardware. They can't just say "Nope, it is all discontinued, go away," as Sun has preexisting contracts with people and the contracts come with everything else. However if they are big enough dicks, everyone will switch of their own accord.

        It's that, or they really don't know how to try and run a competitive hardware business.

        • Re: (Score:3, Insightful)

          It's that, or they really don't know how to try and run a competitive hardware business.

          That's okay -- neither did Sun.

        • Your logic makes sense, until you realize that IBM offers a LOT more then Oracle does. IBM can be a total solution provider (including that problem of what to do with your cash) and will be more then happy to replace Oracle for you.

          So I think it is the last. Oracle just doesn't have a clue. They are used to be seeing as the only professional database, so they don't really think in terms of competition. You don't really compare quotes on databases like you do with say webservers or NAS storage. Your IT guys

        • I can't believe no-one has mentioned this yet. Oracle's Exadata2 solution uses Sun x4175 and x4275 servers, and runs on NO, not Solaris, but Oracle Enterprise Linux. (Which I believe is just a RedHat variant.)

          Its my impression that Oracle bought Sun for the hardware, in order to deliver a one-stop-shop solution for Oracle clusters. The one-throat-to-choke model, if you will.

          http://www.oracle.com/technology/products/bi/db/exadata/pdf/exadata-storage-technical-overview.pdf [oracle.com]

          slides 16, 17, 22, and 57. And th

      • Re:Hogwash (Score:5, Informative)

        by fm6 ( 162816 ) on Friday April 09, 2010 @11:27PM (#31797724) Homepage Journal

        Where did you hear that? I was working for Sun at the time, and there was nothing official about Oracle until after talks with IBM broke down. And then it was for the whole company. It's true that Sun restructured itself so that all the software businesses (minus Solaris, which was moved into the hardware division) could be sold. But there were no offers. The sad truth is that Sun's software initiatives generated tons of press (even people who don't know what "high level language" or "virtual machine" mean have heard of Java) but not much in the way of revenue.

        This acquisition was never about software. People assumed it was, because software is all they know about Sun. But most of the revenue came from selling hardware. Buying Sun for the software is as silly as buying Oracle for Larry Ellison's yacht.

        • Re: (Score:1, Informative)

          by Anonymous Coward

          "No offers told to the engineers" does not mean "no offers". Unless you were the engineer asked why the VP's and the lawyers took more than 30 seconds to arrive, and called on the carpet about it, most of us who do real work were unlikely to get reliable information until after the managers with stock options managed to sell them or reinvest them quietly, before any actual offer is put on the table and blocks them from trading.

          If you think I'm kidding, keep a very close eye on the sales leadership and lawye

        • Re: (Score:3, Insightful)

          by williamhb ( 758070 )

          This acquisition was never about software. People assumed it was, because software is all they know about Sun. But most of the revenue came from selling hardware

          It's not Sun's revenues that are relevant -- if Sun's revenues were good enough it would have been able to stand on its own two feet. Oracle's revenues are all about software and in it's expansion from databases into other middleware, it had bet the house on Java. The words "Our biggest competitor is talking about buying the company that directs the stuff we are totally reliant on [IBM bidding for Sun]" would rightly have been ringing alarm bells in Larry Ellison's ears whether he was on the yacht or not.

          • by fm6 ( 162816 )

            Oracle's revenues are all about software and in it's expansion from databases into other middleware, it had bet the house on Java.

            What do you mean "bet the house"? Last time I looked, most Oracle software was native code. Yes, they rely on Java a lot, but no more than a lot of other companies, including IBM.

            And even if Java technology is somehow crucial to Oracle's survival, how does owning Sun help them? You don't need Sun's permission to use Java. At best you need permission to use the trademark.

            Recall that IBM walked away from Sun rather than meet the asking price. If Sun is worth owning just for Java, that would be insane. But IBM

    • by vbraga ( 228124 )

      Do you have a source for this? I'm curious.

    • Oracle originally only made an offer for Sun's hardware assets. They only bought the entire company after IBM made a bid for it.

      My theory all along was that Oracle bought Sun to stop anyone else getting it. What would they do with it? And what would anyone else do with it? I have no idea, and neither does Ellison. It's just how his mind works - he's like a dog with a bone.

  • And scaling (Score:4, Interesting)

    by bill_mcgonigle ( 4333 ) * on Friday April 09, 2010 @09:00PM (#31797034) Homepage Journal

    Cringely was on about this a year ago - Oracle needs Sun hardware to scale [cringely.com].

    Go go ahead and GPL ZFS, guys.

    • Go go ahead and GPL ZFS, guys.

      I would LOVE for this to happen. I have some systems I've kept on Solaris purely for ZFS. ZFS on Linux would really be the best of both worlds.

      Oh - and DTrace please while you're at it.

      • Re:And scaling (Score:5, Insightful)

        by ducomputergeek ( 595742 ) on Friday April 09, 2010 @10:20PM (#31797452)

        Then what you want is FreeBSD, not Linux. FreeBSD has had DTrace for a few years now and ZFS support for a couple years in experimental mode. As of FreeBSD 8-Release, ZFS is now considered "Production Ready". We've been slowly moving the last of our Solaris stuff over to FreeBSD the past year even before ZFS was officially supported in the FBSD 7.x series.

  • If Sun hardware can't compete with the P-series, why would Oracle want to buy it? (If current Sun customers don't even want to buy Sun hardware, why would anyone?)

    • I think the idea is that IBM are more competent at selling a complete stack than either Sun or Oracle alone, so neither could compete individually. When Oracle buys Sun, they can offer hardware, OS, database, and enterprise app stack all from the same vendor, supported by the same support contract. This makes them competitive with IBM.

      I'm not sure who is replacing SPARC hardware with pSeries stuff. Maybe if you've got old UltraSPARC II stuff it makes sense. The POWER chips are designed for similar work

  • Isn't Sun's ridiculously overpriced and underpowered hardware the reason they went bankrupt?
    • Isn't Sun's ridiculously overpriced and underpowered hardware the reason they went bankrupt?

      Um, one, they never went bankrupt. They had billions in cash just sitting in the bank, in fact. Next, hardware wasn't why they declined. Hardware sales were keeping them afloat. There are three reasons they were declining:

      1 - Software is one reason they declined... specifically, Linux software, as it did much of what Solaris did at no or lower cost. Windows was also cheaper when you considered the cost of the hardware it ran on.

      2 - Leadership was non-existant, and the sales strategy was all over the place like an ADHD kid bouncing off the walls. "We'll push Java! It'll make us rich! No, we'll push network computers, it's the wave of the future! No, we'll compete at the low end by GPL'ing and giving away our software! No, we'll spend a billion dollars on a free database system, and then give THAT away! Riches will follow!"

      3 - With this lack of focus, IBM attacked them from the top, and Microsoft from the bottom, squeezing them out of former markets

      Larry Ellison has made what I think is a prudent decision; stick to the expensive, profitable high end, and quit giving your software away. Pump money into your hardware, as your latest CPU offerings compete very well on the high end with lots of servers, especially on performance per watt costs.

  • looked nice and scrumptious

    not a fan of this guys tech vocabulary

  • by swordgeek ( 112599 ) on Friday April 09, 2010 @11:24PM (#31797716) Journal

    So I've been working with Unix vendors for wow--decades now--and have worked very closely with some of them, as a big customer and also as a 'strategic partner.' I've never been close enough to see the email in the company, but maybe that gives me a bit of neutrality to my knowledge. Anyways, here's what I see:

    1) IBM? Nobody buys P-series. Oil/Gas doesn't buy them, telecom doesn't buy them, entertainment doesn't buy them, and that leaves financials. Maybe the banks are buying P-series, but to replace Sun gear? I doubt it. More likely, they're replacing VAX and S/390 gear. (Yeah, still.)

    2) Sun's hardware (i.e. SPARC gear) has some very nice features, but is not in the same category for _general_ computing power. Massively multithreaded jobs belong on SPARC, small-thread number crunching belongs on the GHz-of-the-day winner, and that's x86-derived. Sun has also thrown away most of their competitive advantage in the x86 market by embracing Windows. If it weren't for Windows compatability, they could have had Open Boot Prom on every single box they sell, but instead we're stuck with a third-rate BIOS and ILOM (LOM designed by committee of middle managers).

    3) Software ls really the most valuable asset that Sun had at the end, but the problem has always been monetizing software. Sun's model actually worked well (it was the follow-through they eventually fell apart on)! Sell hardware, give away software, include training credits with hardware purchases, and soak you for enterprise support. There aren't a lot of big companies unwilling to pay Sun's prices for great support on rock-solid products, but there are a lot who don't want to pay for CRAP support on flakey products, which is what Sun has been offering for two years now.

    Oracle could make out like a bandit if they rationalised the SPARC lineup, maintained the model, and fixed the support issues. Instead, they're destroying the business model, breaking support EVEN MORE, and ignoring all Sun products. I'm afraid that Larry Ellison thinks he just bought a hardware monopoly to support his software monopoly, and is going to be in for a rude surprise when customers leave him in droves for Linux or Microsoft.

    I don't like it, but I don't see much of an alternative. The egos are too big to keep good products alive and relevant, so they're all going to fall apart.

    • In the telecommunications industry, at least at the company that I work at, of the tens of thousands of Unix servers, over half of the servers are Sun, roughly a quarter are HP-UX, most of the last quarter are Linux and a very tiny sliver are IBM running AIX.

      No one is buying new IBM servers. There has been a slight rise in Linux over the last few years, but the continuing growth is on Solaris/Sun.

      • Yep. That's about what I've been seeing - except that in my chunk of telecom, Linux is gaining ground rapidly. Probably one server out of three that we deploy is Linux now, and considering that only 5-10% of our data centre runs it, that's a big increase.

        In the Oil/Gas (and mines and resources) sector, I see more and more Windows showing up - mostly due to reduced time to deploy, regardless of how good the deployed product is. No matter how many times they get bitten, some people would still rather have a f

    • Re: (Score:3, Insightful)

      by Anonymous Coward

      Sun hardware was always underpowered for the price, and they've always played strange and silly games trying to tie the hardware to the software. Remember how SunOS "wouldn't run on sun4m hardware", until Tatung released a modified SunOS that ran just fine, and Sun played catchup with their own OS? Their own engineers refused to run Solaris and preferred the more BSD like and open source compatible SunOS: I'm aware of at least 3 who ran illicit copies of modified SunOS, at Sun, for their own day to day work

      • by davecb ( 6526 ) *

        Sun's recent problem is that it's been underpowered for the price. When they started out, the rule of thumb was an engineering workstation (anyone's, not just Sun's) was twice the price and ten times the performance. Smart people bought workstations, pulled the heads and put them into racks as servers.

        This advantage has been degrading over time, as hardware performance got into diminishing returns, and has only started coming back with the T5000 series. Alas, the T is for small/medium business, so they'

    • One can only hope, in many ways Oracle is worse then Microsoft.

  • by Kymermosst ( 33885 ) on Friday April 09, 2010 @11:27PM (#31797726) Journal

    We've deployed a few P-series systems in place of where we would have deployed big Sun boxes.

    My observations are thus:

    1. I like Solaris way better than AIX.
    2. If you consider Linux and Solaris to be cousins from an administrative standpoint, then AIX is a 3rd- or 4th-cousin. Lots of things are different.
    3. smit is my friend and helps deal with #2.
    4. Virtualization on the IBM gear is powerful. And WAY complicated.
    5. I keep hoping we'll change our mind and go back to Sun gear, but it's rather unlikely.

    • by the linux geek ( 799780 ) on Saturday April 10, 2010 @12:23AM (#31797958)
      A lot of the appeal of AIX and i is the support for virtualization (#4 on your list.) Seriously, the LPAR system and other virtualization bits are the absolute best available, and they blow away everything else on the market. AIX may be a weird-ass UNIX with a lot of strange and occasionally unpleasant quirks, but there are perfectly good reasons why IBM is #1 on UNIX hardware, and the speed of POWER processors isn't the only one.
      • by Kymermosst ( 33885 ) on Saturday April 10, 2010 @12:59AM (#31798054) Journal

        Overall I like the hardware, though there are a few things that I find annoying. People say "fast" with the P6 CPUs...but they don't execute instructions out-of-order, so a high clock rate isn't what it appears. Another complaint is that I can't add/remove/swap CPUs and memory while the system is running, even on the 595s. Sun had this figured out ages ago. Lack of simple integrated systems management forcing the use of the HMC on the bigger boxes is also kind of annoying. IBM also requires that the HMC be placed within a certain distance of the systems, which forced me to get creative with a particular data center.

        There's a bit of complexity with support plans and cost as well. Even though AIX only runs on IBM hardware (as far as I am aware), you have to buy separate support for it. I suspect that there may be a few customers who ditch AIX and run Linux instead, I'd rather see the AIX support and Right to Use "in the box", so to speak.

        I have to say though that you're right about the virtualization and partitioning capabilities being some of the best out there. It sure comes with a steep learning curve (and don't get me started on LHEAs!).

        I suppose my dream world might be Solaris on either SPARC or POWER, with IBM partitioning/virtualization capabilities in a Sun frame with RIO connections...

        • Re: (Score:3, Informative)

          by MistrBlank ( 1183469 )

          The price to run AIX and get software support where I am is about roughly the support of Linux. We saw that and decided to stick with the OS that's fully supported by the company rather than one that needed to be hacked to run on Power.

          BTW, that's BS about the add/remove on CPUs and memory, especially on the 595s. The only problem is getting a CE that feels comfortable doing it. But really the future is in being able to migrate your partitions off of the hardware that you want to change and then move the

          • BTW, that's BS about the add/remove on CPUs and memory, especially on the 595s. The only problem is getting a CE that feels comfortable doing it.

            Well, extensive googling has yielded marketing material that makes the claims that the feature to hot-add/swap/remove CPU nodes would be added to the 570 and 595 in Q4 2008. Our systems were purchased prior to that so maybe the new system firmware update our FE is recommending has that capability.

            I do know that when we were negotiating the purchase and benchmarki

      • Virtualization is also pretty nice on the UltraSPARC T2. Like the IBM hardware, there's a simple hypervisor in the firmware and you can partition it into logical domains easily. It's very well supported on Solaris and also on OpenBSD (you can run a firewall in an OpenBSD VM to protect your Solaris instances). Not sure how well it works with Linux, but I presume the support is there too.
        • I have also deployed LDOMs on the CoolThreads stuff, but it's not quite like the IBM offering. The capability delivered with the IBM systems (provided you have purchased the appropriate license!) is more like a combination of domains (eXXXX[X]-style) and LDOMs rolled into one.

          However, LDOMs a are definitely far simpler to set up. Also, being able to set up a system and do dynamic reconfigs without an HMC is a good advantage for the Sun boxes.

      • I'll echo the parent's sentiments. I love working with AIX, in some ways it feels like it's behind Linux and in others it feels like it blows the doors off of it. I only wish I could run it at home so I could actually learn how to administer it without spending an arm and a leg on hardware and dropping a new power circuit to my home.
  • by gig ( 78408 ) on Friday April 09, 2010 @11:53PM (#31797826)

    If you create a complete solution, you can tune it for best performance, you can make it easier and cheaper to deploy, you can guarantee a certain level of quality, you can include a warranty, you can harden it in ways that software alone can't.


  • by PhunkySchtuff ( 208108 ) <kai@@@automatica...com...au> on Saturday April 10, 2010 @04:02AM (#31798526) Homepage

    "I'm betting that they looked at Apple's margins on hardware, and saw potential in doing the same with Sun's hardware business."

    Are you freaking crazy? Sun's margins on hardware make Apple's margins look like small change. Having sold both in my career, there are retail margins of 8% on Apple hardware and anywhere up to 20-30% on Sun hardware. That's just the margins that the resellers make. Then there are the margins that Apple or Sun make themselves. Apple's are generally worked out to be around 30%, and I'd shit a brick of Sun's margins on hardware were anywhere less than this...

    • To build on a point that you are making, though, Oracle would like to be Apple for the business world. Steve Job is known to be a close friend of Larry Ellison. Larry Ellison also takes business very personally--to him, if he has a grudge against you, it's not enough to just crush you. He wants to see you suffer.

      I'm not surprised Oracle wants to provide an entire database stack like Apple systems and Apple's software, in the manner of how they are sold. I don't think it will be too long before we will see O

  • I don't know if its me (I'm getting jaded and cynical in my old age) but I do keep wondering how Oracle's takeover will affect Sun's OS efforts.

    The only reason I mention this is that there has been a noticeable (at least IMHO) change in VirtualBox development. Since the Oracle takeover, VirtualBox development seems to have changed direction or slowed down... I can't really put my finger on it but something noticeable has happened. I don't if the core devs have been affected/left or what.... but certain Virt

    • i think netbeans is still coming along - at least for 6.9. Not sure if there'll be a 7 any time soon though! i.e. Larry may have too much of a vested interested in JDev's tooling for Fusion Middleware. Still, keeping Java vibrant through the status quo is in Oracle's interests since their whole platform is built on Java EE running on Weblogic. While NetBeans is small fry, it contributes to the overall ecosystem.

      From what I hear, any additional resources for Java will be going into the core - merging jrockit

  • Sun doesn't have any hardware anymore. They killed off most of the value in SPARC years ago. Solaris isn't going to win back the data center in any major way any time soon(if ever), and it's the only OS which really runs particularly well on SPARC. Intel has pretty much taken the general purpose CPU crown at this point and may very well stand alone in that arena by the time the economy comes back to normal.

    The rest of SUN's supposed hardware is the same stuff everyone else distributes upmarket models of the

  • Maybe they should reduce the eff'ing price model on their Oracle line and maybe, just maybe focus on the AIX version instead of leaving it last for support.
  • One only needs to pick up any copy of the Wall Street Journal from the past couple of months to know that Oracle bought Sun for the hardware. On the front page of every edition there is an advertisement comparing Sun and IBM hardware. "Sun 7 times the performance, IBM 6 times the power consumption."

  • Oracle's main product functions in a cluster quite well for normal business use. With 12 to 48 core chips in a single x86-64 processor, the advantage for scalability of the Sun kind (32 to 256 cores per machine) will disappear in the next year. The price/performance will favor 8 chip SMP x86-64 systems or clusters of them to run Oracle. UltraSparc is dead meat.

    • Unless you have seen just how efficient the T5XXX series is with Java, I would not count the SPARC line out. Bad code is suddenly workable

      Oracle now has a mirror to what IBM has had: An application tier (Weblogic) designed to run on a specific chip for maximum performance.

      IBM has always had a DB to run on specific hardware, the Mainframe.

      Now, if the next generation of SPARC (VIII) optimizes Oracle DB proper, you have a killer application stack: Optimized JAVA engine and Optimized DBA stack.

      • One problem, the next generation of sparc, VIIIfx, is designed by Fujitsu who might have totally non-Oracle goals in mind.

        Now if you mean Niagra-3, haven't heard any recent news from Oracle about that.

Karl's version of Parkinson's Law: Work expands to exceed the time alloted it.