Sun Microsystems' Solaris/SPARC systems have an undeserved, and out of date, reputation for being higher priced than Linux-on-Intel servers, says Unix book author and IT expert Paul Murphy. He believes that IT shops should stick with Sun when upgrading, rather than switching to the flavor of the day, Linux. But that argument doesn't float with consultant and author Bernard Golden. In this face-off debate, Murphy makes his case for Sun, and then Golden puts that case under his magnifying glass.
Moore's Law and upgrade pressures: Why you should stick with SPARC/Solaris
By Paul Murphy
If you're running Solaris, particularly if it's on older SPARC gear, you may be under pressure to switch to Linux on "industry standard" hardware. Here's why you should resist that pressure and what you should do instead.
Listen carefully to what you're being told, and you'll notice that public acceptance of Linux on Intel as cheap, effective and gaining in popularity often has the implicit corollary that Solaris on SPARC is expensive and out of date. That reasoning comes from the fact that people update their certainties about systems pricing only when someone draws a related issue to their attention. Recently, the price reductions and advantages in the world of Linux on Intel is front and center all the time, while corresponding changes in RISC/Unix are not.
Suppose, for example, that back in mid-2000, you installed a pair of Sun 450s running Oracle, each with two 327 GB T3 workgroup arrays, 4 CPUs at 450 MHz, and 8 GB of RAM. Taken together that might have cost you around $700,000 initially, and something like $2,900 a month for hardware support after the first year. Now, five years later, users are complaining about those servers' performance. Also, some of your colleagues are saying you'd get a six-month payout just on support costs by replacing this stuff with a couple of Dell boxes running Linux. They add that you'd get an enormous performance boost into the bargain.
Your colleagues are about half right on the money, absolutely right about the performance opportunity and dead wrong about the implicit exclusion of SPARC/Solaris from the operation of Moore's Law.
That dual 450/T3 combination gave you a good bang for the buck in its day. Do a detailed review of the options available to replace it today, and you'll find that vendor pricing relationships haven't changed much on, say, Dell's 6600 quad Xeons, Sun's V40z quad Opteron, IBM's newest PowerRISC 720, or Sun's 440 UltraSPARC.
Back in 2000, both Compaq and IBM had gear that could compete on performance, but not on price, with the Sun 450/T3 combination. Today that range is narrower with Sun, Dell and IBM all clustered around $40,000 for a single system capable of handling the entire load. Even so, the cheapest and fastest options are still from Sun.
The fastest option, built around Sun's four-way Opteron V40z, is also the cheapest from a hardware perspective. At about $38,000 inclusive, it should handle the load formerly shared by the two 450s while cutting response time in half. The most expensive option is probably the Dell 6600 (I could not get fully configured pricing on the IBM 720 from IBM's Web site) at about $45,300 inclusive.
The most reasonable option, the Sun 440, is neither the cheapest nor the fastest at about $41,000 and including a 3310 external disk array. It has, however, the benefit that you can put it in place without procedural, technical or software change.
It's the overall magnitude of the change in Sun's pricing, however, that's stunning. The SPARC/Solaris solution that cost $700,000 in 2000 comes in at about $41,000 now; a 17 to one change in cost over less than five years with an approximate doubling of performance thrown in.
That, of course, is the problem: Everyone knows that Sun's servers have been getting cheaper, but they haven't put two and two together. The people pushing you to go to Intel-based solutions just haven't internalized the reality that a comparable bang now costs about 6% of the bucks it used to, whether you go with Linux on Intel or Solaris on SPARC.
So, the factors that should influence your upgrade decision have little to do with the perception that Unix/RISC has somehow fallen by the wayside and everything to do with the familiar litany of paying heed to vendors' business strategy, software licensing, retraining and procedural changes.
Although Oracle licensing, to continue with the example above, has achieved roughly the transparency of a thick coal seam, it's probably safe to say that most organizations would face no cost to transition from two 450s to one 440, but would face at least media and service costs for the alternatives.
No matter what your actual application, however, a phone call to each vendor involved is called for and will produce results that should influence your decision. Beyond that, there may be good reasons to go with Linux: If your company has a Linux skills development strategy in IT or a company-wide desire to align your business with IBM, going to Linux on either x86 or Power5 makes sense.
Then again, your company should check out the costs of this change. Going from the mid-1990s Solaris 2.5.1 on UltraSPARC I and II to Solaris 10 on UltraSPARC III or IV means changing the box today and eventually learning to use some of the new capabilities. But, that's all it means: Your software will run unchanged.
In contrast, going from Solaris to Linux requires extensive training, systems and application testing, and procedural debugging.
If you see a clear gain, do it. Oracle, or whatever you're running, will work just fine on Linux. Your existing Sun gear will keep on ticking until you're ready to go live with the new gear, new people and new procedures.
But if you don't see a clear gain, why not just ask those pushing you to check their assumptions?
The SPARC of Moore's Law
By Bernard Golden
In Paul Murphy's column, he points out that one shouldn't imagine that Moore's Law only applies to Intel-based systems, but that the dramatic price/performance implications of the law also apply to other chips like Sun's SPARC servers.
Consequently, he suggests that when considering a replacement for an aging SPARC-based system, new SPARC systems should be considered along with Linux/Intel (Lintel) systems. In the specific scenario he discusses, a pair of old Sun 450 boxes need to be replaced. The candidate machines he discusses are:
- Dell 6600 quad Xeon
- Sun V40z quad Opteron
- IBM PowerRISC 720
- Sun 440 UltraSPARC
The application he describes running on the Sun 450s includes Oracle, but he makes no mention of any other software. His assessment is that pricing for all four options runs around $40,000, with the Intel-based machines costing just as much as the SPARC machine.
Of course, Murphy is right in his observation that moving to a new operating system, even one that looks like Solaris, as Linux does, still involves some training, extra costs as employees familiarize themselves with the new systems and so on. He says that one should strongly consider sticking with SPARC-based equipment, rightly pointing out that the most important issues that should be looked at are business strategy, software licensing, retraining and procedural change. Essentially, Murphy is making the case that, with all costs considered, the marginal cost of upgrading from one SPARC system to another is less than the switching cost of moving to a system based on new chip architecture.
Murphy would be quite right in his assessment if, indeed, SPARC-based systems are similarly priced to Intel-based systems. One must admit, too, that it's impressive that there's been a 30-fold price/performance improvement in machines since the Sun 450s first saw the light of day.
However, I am not sure that SPARC and Intel (actually, x86, since one of the boxes is based on AMD Opteron chips) prices are really that similar. In just a bit of poking around, I found four-way Opteron boxes significantly less expensive than the prices he quotes.
More to the point, the general industry evidence seems to indicate that SPARC-based systems are not price competitive. I attended a presentation by the chief information officer of Charles Schwab in which he discussed how the company was replacing Sun boxes with Linux-on-Intel boxes as fast as the Sun servers' leases expired. Since Schwab would be subject to all the switching costs Murphy discusses, it seems to me that the Linux-on-Intel alternative must be significantly cheaper for Schwab.
Furthermore, Sun itself is pushing x86 machines. If Sun could really make the case that SPARC boxes were just as cheap as x86 ones, why would it be adding them to its product mix?
I can only conclude that, in the real world, x86 boxes are less expensive. On the other hand, as Murphy points out, there may be other costs like retraining that could outweigh the cost savings of x86 machines. That's why I'm surprised he didn't include a fifth option: Solaris on x86. Many of the switching costs, like retraining and employee familiarization, could be avoided, while taking advantage of the real-world price advantage of x86-based machines.
While sticking with SPARC machines may seem like the safest, most conservative option, it may actually be the riskiest. The chip industry has proven that cost advantage lies with volume manufacturing; one day in the future sticking with SPARC may feel more like being stuck with SPARC.
Send your feedback to the SearchEnterpriseLinux.com news team.