Slashdot Mirror

Most database products (even open source ones) compete very well against PostgreSQL because the support of parallel query execution has been supported for years. PostgreSQL has only recently added some features in the direction. And we are not talking about N clients to 1 server, but 1 server with 1 query where only 1 CPU is used in the PostgreSQL case, and others automatically spread the workload.

Some other tricks from SQLserver are obviously the integration services (Extract-Transform-Load) and Analysis Services. Again you could use open source third party product such as Talend et al. Now I am aware these kind of smart clients should be seen as tooling on top of a database itself. But as long as those tools work very closely with one database... it basically prevents you from using anything else unless you want to integrate your own solution.

PostgreSQL has such amount of features, that as data integration toolkit it is great. But for performance I would not use a vanilla instance... since even tuning and automatically indexing is not something that the database does for you. And BTW... for performance look in this list. http://www.tpc.org/tpch/result...

Oracle is popular with mission-critical enterprises, who have LARGE checkbooks, HUGE transaction volumes, and cannot afford ONE MINUTE of downtime.

This is not, and has never been, Microsoft's target market.

No one has attempted in quite some time to seriously challenge Oracle in massive transaction volumes.

You will notice that Oracle's top score is on SPARC, and is from nearly three years ago. There is no significant challenge to them on TPC-C.

How about look at TPC-H, where in the 10TB test the T5-4 gets beat in performance by the DL580 G8 and the x64 system is half the cost per transaction? The number of shops that need more than 120 threads and 3TB of ram is vanishingly small.

If you examine the top two best performing database platforms (as benchmarked by TPC-C score) you will discover that they are both sold by Oracle, and that the SPARC version has both higher performance and a lower cost per transaction than the x86-64 version.

You might find this quote to be particularly interesting:

"I am going to make a promise to you," [Larry] Ellison said. "By this time next year, that Sparc microprocessor will run the Oracle database faster than anything on the planet."

SPARC: "I am going to make a promise to you: By this time next year, that Sparc microprocessor will run the Oracle database faster than anything on the planet.". SPARC now beats x86.

RDB: Enough said.

Columnstore databases such as MonetDB and their commercial spinoff Vectorwise (now Actian) already showed this can be achieved with open source and proprietary code.

There are plenty of benchmarks that show MSSQL and Oracle with better price/performance ration than MySQL. Here is an example of a benchmark and here is a random person who did the math and found out that the licensing cost of MSSQL is more than balanced by the lower support cost in a large installation.

I'm not saying that MySQL isn't a good choice, just that the licensing cost is only a small part of the cost of a DBMS installation, so MySQL being open source really isn't that big of a price advantage unless the installation is very small. And for small customers, both Microsoft and Oracle have a free option.

No its not going to be me...
http://www.tpc.org/

So if Oracle is so happy with MySQL. Give us a TPC-H benchmark. Scalefactor: 100.

You're conflating throughput with speed. Speed is latency (how fast). Thoughput is how many.

You need four 5GHz POWER6 to match two 2.93GHz Intel Nehalem on official TPC-C benchmarks

What are you talking about? TPC-C benchmarks entire systems not cpus. Mostly I/O bound... RAM, network and disk.

http://www.tpc.org/tpcc/results/tpcc_perf_results.asp

The IBM box is one machine with 32CPUs 64 cores. The Sun, a cluster of what? 12 systems? with 4 CPUs each, 384 cores. See what I mean by IBM getting it?

I think you're right though. x86 will eventually eat power too, simply because the market for "ultimate performance" is much smaller than "good enough".

http://www.tpc.org/tpcc/results/tpcc_perf_results.asp - I wouldn't say "clobber," but they're roughly at par on performance and Itanium has an edge on price/performance.

http://www.tpc.org/tpch/results/tpch_price_perf_results.asp

So basically Superdome is not in the top spot (even not the top not clustered spot) for anything until you get to 30TB, where it is the ONLY entry. Now true I will give you that it was a unisys ms sql box that topped at 3TB that no one would really buy (and it does not do much in the realm of QphH), but that was a Xeon box, as were the top few in every other category always Xeon or Opteron, no Itanium. In fact pretty much everywhere below 3TB there were even Xeon and Opteron boxes with better Price/QphH and QphH than HP Superdome. (You do get some other oddballs that no one would really buy though too.)

I think it does show that HP Itanium is very expensive unless you need to push upwards of 200K QphH on insane data sizes in a single box. There are other systems that can do more QphH far cheaper as long as you limit yourself to 1TB. When you look at simply QphH then again the only time that Itanium is on top is the one where it is the sole entry. A power6 beats it in the 3TB category, can push more than 50% more QphH but only costs about 20% more and is running software that people would really use. It is a cluster though.

So have you found any measurements that I couldn't find that you could point out that demonstrate lingering categories in which a mainframe might consistently outperform commodity hardware (ie, any measurement that is or can be compared to another at least somewhat related measurement on commodity hardware which demonstrates an advantage for the mainframe)?

Have you tried http://tpc.org/ ? Transaction processing is a good measure, and the tpcc top 10 has a long and competitive history. You'll get a good idea of who's currently fastest processing a standard fairly complex (and reasonably representative) OLTP transaction. It's also a good measure of who's top dog in the database world. It's also worth some thought about database residency -- because of the needs of lock management, most databases (and I specifically exclude BigTables from this) tend to work better with scale-up rather than scale-out, thus the popularity of DB2 on big iron.

Hmm... in light of AKAImBatman's insight, I may have to re-think that...

I have noticed over the years that the top ten have seen a constant game of leapfrog between Oracle, SQL Server (yes, that's not a typo) and DB2.

Whats so special/magical about a mainframe?

Mainframes have followed Moore's Law just like the rest of the chip vendors. You buy a new mainframe, you get new chips.

But the main difference is essentially their slightly different design philosophy. Reliability is built into the price, for one thing -- part of the reason it costs more is that conservative design - not the most cost effective in terms of power -- as you often lose power per component from the "underclocking" attitude that a focus on reliability will engender (and they're tested to buggery before delivery, too). You also get a much higher standard of module connectivity and far more robust power supplies and inbuilt hardware redundancy.

They also tend to support and address much more memory than you'll see on the smaller servers.

The other main point in favour of mainframes is their orientation toward massive IO. Really massive IO. With the scale out design of i86 processors a lot of IO happens between network cards; on mainframes a lot of that interprocessor data flow happens on the backplane, and significant investment in optimising data channels means you're paying for that IO more than raw computation. The network interfaces on mainframes are pretty massive too, and can support fairly impressive tube bandwidth.

Mainframes using the IBM architecture for a long time have been represented in the TPCC transaction processing top ten, although the trend lately at the very high end is to run AIX on top of P5 architecture. Have a look, it's illuminating, and Red Hat gets a look in too. You can see the numbers at: http://www.tpc.org/tpcc/ .

but "Enterprise" software is normally never sold at the list price...

Pretty much anything "Enterprise" is never sold at list price.

Case in point: TPC benchmark style systems

http://tpc.org/results/individual_results/IBM/IBM_595_20080610_ES.pdf

Look at "Total" and "Total IBM Discounts*" toward the end of page 3. Total list price is $35,263,161. Total "IBM Discounts" are $20,273,753.

IBM is willing to give *anyone* purchasing this configuration nearly a 60% discount right off the bat. Obviously, any company giving an immediate $20 million discount has their shit overpriced by $20 million.

And this is before negotiations even begin, which would likely net the buyer another $7 million or so in savings.

DB2 just doesn't scale down as well as some of the others so it doesn't get as much exposure to the masses, if you check out things like the TCP-H results you'll notice at the 10TB level DB2 is #1 and #3, it's typically used for very large databases running on IBM big iron. It's yet another IBM technology that kind of sits in the corner running some of the largest financial systems in the world without getting a lot of exposure.

If you mean that x86 won't handle I/O load, you may be right. But Windows runs on Itanium, so an Intel box running Windows (like a Superdome) can certainly handle the OLTP load. The particular test I linked to handled about 1M users, and that was a couple years ago. If you're going to spend millions of dollars on a server, it could just as easily run Windows, handle the same load, and be just as reliable.

dom

Yes, the number 8 overall and number 3,4,5,6,7,8 and 10 by price/performance database systems on the TPC transactional database benchmark is DEFINITELY not a real DB /duh

Yes, the number 8 overall and number 3,4,5,6,7,8 and 10 by price/performance database systems on the TPC transactional database benchmark is DEFINITELY not a real DB /duh

The results in terms of a car are miles/kilometres travelled. In terms of computers, MIPS or MFLOPS are however not results, they are performance measurements. Using them would be like describing car efficiency in RPM per gallon. Not the results you're after.

So the first thing to do is define what your results are. The results computers produce are the "bits of information you want".

SPEC and TPC both have benchmarks which already attempt to describe the results that customers are after.

http://www.spec.org/
http://www.tpc.org/

"You cannot justly prefer what you do not know so well."

I defintely prefer a Ferrari to a Sienna mini-van, but I know the mini-van so well...

You talk about feel, but you don't talk about comparing Oracle and MSSQL and (DB2) with serious benchmarks (competitor reviewed even). When they're benchmarked and submitted to TPC, MSSQL doesn't win on most tpmC per system. Oracle does. With 4 million to 1.2 million from MSSQL. However on cost per tpmC MSSQL's most cost effective configuration is twice as efficient as Oracle's most cost efficient configuration. So, if you're setting up an $11 million system for your corporation, you're probably better off with Oracle.

http://www.tpc.org/

So, MSSQL might seem expensive compared to MySQL, but it's certainly inexpensive compared to Oracle, and for at least some corporate purposes, it will provide enough performance to do the job.

My point, that developers/DBAs who want to make a product perform do better with that product than developers/DBAs who feel saddled with the product is still a valid point, even inside Corporate IT shops. Even if it doesn't apply to you personally.

Postgres isn't making the list. Neither is MySQL.

This is at the expense of all that extra storage area.

The people for whom these high end disks are intended aren't concerned with the "storage area" of individual devices. They care about the ratio of storage to spindles and arms. They buy things like this.

Why is this front page news?

Because it's a site about stuff geeks want to read. It's actually rather nice to hit the page and find some news about the latest incremental change in storage, as opposed to more move-slash, dot-on politics.

One objective benchmark to consider when comparing clusters vs. a server or mainframe is the TPC-C online transaction processing benchmark. http://tpc.org/tpcc/results/tpcc_perf_results.asp? resulttype=all Clusters get beat here in both absolute performance and performance/price, with the first cluster in the top ten being a cluster of 4-way HP Itanium servers. Granted this benchmark may or may not be relevant to what you care about, but it is definitely a very clear objective test for online transaction processing.

All of the commercial database software packages (that I know of) have a specific part of the license which forbids publishing benchmarks. You'll never see head-to-head benchmarks of MS SQL, Oracle, DB2, or any others, because they don't feel that any benchmarks would reflect the results of their product fully optimized.

Actually, I wouldn't have been surprised if they had come out with a blanket "no benchmarking" clause (except that it's difficult to benchmark an entire O/S). That's pretty common with high-end software (databases in particular), so MS would just be applying it to something more common.

I don't remember exactly what the restriction is, but I think that nobody can publish TPC (the Transaction Processing Council) benchmarks, they can only submit them to the TPC, which will then make them available. So if your new system breaks the old record, you can't tell anyone until the TPC releases the information, which I'm sure just twists the guts of many a marketroid in the RDBMS and high-end server segments....

you sir are a moron. until MySql can get here http://www.tpc.org/tpcc/results/tpcc_results.asp?p rint=false&orderby=dbms&sortby=desc then you can't ask stupid questions like that. MySql isn't an enterprise database, Oracle is. They sit in different markets. There is no debate about this.

I thought that the >4 CPU Windows systems were, in essence, specially tweaked systems to make it all worthwhile and that standard setups couldn't really make effective use of more than four processors. If so, I stand corrected. *looks around* Err, sit corrected, sorry.

Multi-core restrictions on Windows versions are mostly artificial. For example, 8-CPU systems running just fine on Windows 2003 Advanced Server without any special tweaking. The system the grandparent referred to must have been running Windows 2003 Data Center Edition to support more than 8 processors, but should still require no special tweaking.

That said, I'm sure the systems that make it to the top of TPC benchmarks are highly tweaked.

AMD offering with comparable RAM size: 2xOpteron 280, 114K thoughput, $2.99/unit cost, 32-bit OS, available May5 2006.

I looked for, but did not find, any 64-bit OS vs 64-bit OS comparisons (they only have 64-bit on 4-way Opteron, which isn't fair), this is unfortunate as the larger address space should make a real difference in TPC benchmarks. Also, the AMD ones are on 280s, AMD has 285s with larger caches now.

Difference in raw score is impressive though. It's a race ... can DDR2 + RevF improve Opteron enough in the six months before "availability" of Woodcrest?

Note that TPC numbers are usually price/performance; already, AMD is in a dead heat (if not ahead) here, and then factor in the six month release date difference. Hmm... looking at the numbers, the best 2xdual-core Xeon system reached $3.28/unit cost, Woodcrest made it to $2.99/unit, AMD 2xdual cores reach $2.73 (on a 128GB mem config) and 4xdual reached $2.13. Well... Intel is finally in the right ballpark for cost/performance! Now if only they had a pre-production 4xdual to show off?

Thanks for the link. Most interesting.

AMD offering with comparable RAM size: 2xOpteron 280, 114K thoughput, $2.99/unit cost, 32-bit OS, available May5 2006.

I looked for, but did not find, any 64-bit OS vs 64-bit OS comparisons (they only have 64-bit on 4-way Opteron, which isn't fair), this is unfortunate as the larger address space should make a real difference in TPC benchmarks. Also, the AMD ones are on 280s, AMD has 285s with larger caches now.

Difference in raw score is impressive though. It's a race ... can DDR2 + RevF improve Opteron enough in the six months before "availability" of Woodcrest?

Note that TPC numbers are usually price/performance; already, AMD is in a dead heat (if not ahead) here, and then factor in the six month release date difference. Hmm... looking at the numbers, the best 2xdual-core Xeon system reached $3.28/unit cost, Woodcrest made it to $2.99/unit, AMD 2xdual cores reach $2.73 (on a 128GB mem config) and 4xdual reached $2.13. Well... Intel is finally in the right ballpark for cost/performance! Now if only they had a pre-production 4xdual to show off?

Thanks for the link. Most interesting.

TPC numbers would be ideal

2P Xeon 5160

Responding to your comment: Microsoft is the one prohibiting their licensees from benchmarking their database against others. They must have something to hide.

All commercial software vendors generally prohibit benchmarking to some extent these days. Yes, Oracle and IBM too. The idea is to prevent "bad" benchmarks by improperly configured setups. Most vendors' service organizations will assist journalists and even end-users with performing standard benchmarks (sometimes for a price).

As for SQL benchmarks, they do exist. And Microsoft SQL Server fares very well (espicially on a price/performance basis). See the non-clustered TPC-C benchmarks for four-socket servers for example. I've never seen a TPC result for an open source database listed. Why? Probably because MySQL and even PostGres can't scale like the commercial DBs, and nobody wants to invest time and money it takes to making a sucessful benchmark submission for a product they can't actually sell.

Note that Oracle has largely stopped submitting TPC results, as they've been getting their ass handed to them by MSFT and IBM so regularly.

Itanium2 systems are among the top in transaction processing
http://www.tpc.org/tpcc/results/tpcc_perf_results. asp?resulttype=all
and THE top one for clusters.

Oh please. The Itanic cannot compete with even old versions of POWER. IBM's dated result from 14 months ago is still 2.6 times ahead of Itanic's best showing.

So when is Itanic going to get dual core? What will be the next delay or cancellation in the roadmap.

Itanic's future: less like a roadmap and more like a funeral procession

Itanium2 systems are among the top in transaction processing
http://www.tpc.org/tpcc/results/tpcc_perf_results. asp?resulttype=all
and THE top one for clusters.

Oh please. The Itanic cannot compete with even old versions of POWER. IBM's dated result from 14 months ago is still 2.6 times ahead of Itanic's best showing.

So when is Itanic going to get dual core? What will be the next delay or cancellation in the roadmap.

Itanic's future: less like a roadmap and more like a funeral procession

There is definitely a place for high-end server processors, and for the last few years IBM and Intel have been leapfrogging each other with the POWER4/5 and IA64. Unfortunately for Intel, when the POWER5 appeared, IA64 was left behind. The TPC-C benchmark http://tpc.org/tpcc/results/tpcc_perf_results.asp is a big deal, and HP lost the lead when the POWER5 servers appeared. My guess is that HP+Intel (along with NEC, Unisys, Bull, and others who also make >= 32-way IA64+Windows Server machines) want the IA64 to be competetive with the POWER5, hence the $10B investment. I have no doubt they can pull it off. The x86-64 processors are great for some applications, but the large transaction processing server is not one of them. If support for the IA64 is withdrawn, then IBM will be alone at the top - and they certainly are NOT abandoning big iron any time soon. There is too much money to be made in TPC.

Itanium2 systems are among the top in transaction processing
http://www.tpc.org/tpcc/results/tpcc_perf_results. asp?resulttype=all
and THE top one for clusters.

It makes sense for such an inventmen to go to
  a) improving the fabrication facilities - achieving lower defect rates
        and reducing price;
  b) improving the fabrication process - aiming at higher clock rates

Remember also the recent announcement that an Itanuim CPU will no longer contain essentially a whole IA-32 CPU.

~velco

http://tpc.org/tpcc/results/tpcc_perf_results.aspT PC-C benchmark

I am so damn sick and tired of biased headlines and (often) even more biased comments. I know Slashdot is part of OSTG and all, but come on - it is beneficial to all concerned to be FAIR and unbiased. To compare MySQL to Oracle, DB2, or SQL Server is a joke. Anyone with half a hint of what they are talking about would know this. MySQL is only better for SMALL websites where speed is more of a deciding factor.

In any case, even ignoring the fact that these are crippled versions of the real deal, this isn't even a proper test! Let's see what a REAL DB comparison looks like:
Complete TPC-H Results List

I know that MySQL and PostgreSQL aren't included in that result list but that is how a test SHOULD be performed, not with the ridiculously hand-wavy methods the authors use to 'score' each DB software.

That being said, no one has any business saying MySQL >> DB2 or Oracle. That's a joke. MySQL would SUFFER in the 10TB test. Also, where is Teradata? Furthermore, the way that the article treats SQL Server is even more ridiculous, because their 'free' version is likely the least functional of the lot since it is SPECIFICALLY aimed at learning on one's own desktop. Nothing to see here, just a random useless article trying to say something to push its writers' ideas without much basis.

On the one hand I am intrigued enough to revisit PHP and MySQL. I used these 3 years ago and found both easy to use. I then moved on to .NET, Java and Oracle. Partially because of circumstance, partially because .NET and Java had superior IDEs and extensible object-oriented approach. I felt that the code I wrote was "more OO". I would use .NET for UI and Web apps and Java for anything I could run from command-line (or non-windows apps).

So part of me is still skeptical about the "findings" in this article. For one thing book sales are maybe _correllated_ to Java's popularity but certainly that isn't the cause for the alleged decline in popularity. In the last 2 years of consulting I have seen a good share of new Java and .NET web applications built and I am not convinced that either is going away any time soon. Eclipse Java IDE is still going strong (check out their web tools plugin) and the next .NET release has been generating a lot of buzz.

If you are rolling out a commercial web app then you should probably check out http://www.tpc.org/ before making a commitment to MySQL. Again, I haven't used MySQL for a couple of years, but there is no way they have caught up to Oracle in that short of a timeframe.

Depending on the application you use a different tool. When I'll need a quick dynamic site up I will try LAMP, but I am still largely unconvinced how exactly LAMP will replace a commercial J2EE web app. What IDE can I use with LAMP? Is there a concept similar to Java Struts?

That is the most uninformed and dumbass thing I've seen written on Slashdot in a while, and that's saying something.

http://www.tpc.org/tpcc/results/tpcc_perf_results. asp
Check the www.tpc.org top 10 list. At #8, from way back in 2003, is a 64 way HP Itanium system running SQL Server. Everything else in the top 10 is more recent.

I really wish slashdot had a -1 (Idiot) mod.

Where did you get this wallop of information? It is not true, MS SQL Server performs very well in multiprocessor environments (not using Hyperthreading). Checkout the TPC benchmarks if you don't believe me: http://www.tpc.org/

Wow, this post sure attracted a lot of flame bait from M$ 'n FUD crew.

Read the original post, "and likely largely tested in a single processor system".

I don't think Microsoft gave it's developers a $5.8M USD machine in #4 www.tcp.org spot that you can't even buy yet to develop MS SQL. It was more likely a PC, single processor and subsequently and later tested on the bigger iron.

Instead of looking at the www.tcp.org site where vendors post systems you can buy, why not look at what organizations are really buying?

http://www.top500.org/lists/2005/11/

There must be some reason that Microsoft consistantly is excluded completely from the top 10 by *real* world purchases. I didn't check to see how far down the list you have to go to see a Microsoft product. I guess those Dells run Linux nicely.

Go ahead M$ pundits, mod this down too. After all it is the M$ way. You don't like the facts so you FUD it and mod it down.

"MS SQL was designed and likely largely tested in a single processor system and multiprocessor or HT support is somewhat less than optimal. So MS SQL is likely best tuned to single processor."

Where did you get this wallop of information? It is not true, MS SQL Server performs very well in multiprocessor environments (not using Hyperthreading). Checkout the TPC benchmarks if you don't believe me: http://www.tpc.org/

Try the TPC benchmarks.

I have no account, so I guess I'm an Anonymous Coward. You can call me Dave Hagedorn, though.

I'm using Linux right now, am not a fanboy, and am not picking sides. I feel it is only fair to point out:

http://www.tpc.org/tpcc/results/tpcc_perf_results. asp

W2k3 Datacenter Edition on Itanium, 64-way SMP, fastest non-clustered Itanium system on the TCP-C benchmark. Yes, Superdome gets thrashed by the p595. Either Power5 is way the hell faster than Itanium or AIX is way the hell more scalalable than W2K3, or both.

I do not believe W2K3 supports 128-way SMP, but there is an SKU available for people who want to run two instances on a 128-way box.

Dave

Food for thought, but of course not an end to all discussion:
http://www.tpc.org/tpcc/results/tpcc_price_perf_re sults.asp

This is a price/performance benchmark. It is mostly supported by hardware manufacturers, so it's less likely to be influenced by Microsoft's or Oracle's bank accounts. Notice that in a price/performance benchmark, free DBMS don't even show up. The most obvious conclusion to be drawn is that free choices require more expensive hardware to get to the same performance level. I'm quite certain that if HP could leapfrog over Dell by using MySQL or PosgreSQL, they would. Also note that RedHat and SUSE made it on the list. They aren't anti-open source.

this is where you usualy go for perfomrance benchmarks on databases

http://tpc.org/

You will notice that neither mysql or postgres are in there. Not sure why. Oracle RAC on Linux is showing up prominately...

I believe the http://www.tpc.org/tpcc/results/tpcc_perf_results. asp?resulttype=all
explains why people are using power servers for things, throw in the LPARs that run, AIX, LINUX, and AS400. If you have more then 2 servers, and want to share those unused dev/test CPU's and memory, pSeries boxes are the way to go.

Yes we love linux on little boxes, but with a few P570s loaded with multiple versions of Linux on Lpars you could go a lot farther then with the HP itanium which is quite a bit more expensive for the output numbers an oracle server will need.

Throw in that almost all nextversion gaming consoles will be sporting a power Cell chip system, dumping the Intel business to a producer that can lowball the market to death doesn't seem such a bad play.

Allow me to refute your alternative possible explanation with *facts*. :)

[snip netcraft]


Ha ha. Very funny. You know, it has been nearly a decade since you needed a really good OS to serve websites, right?

No, if you want to be a Linux contender, you have to do this, this, this, this, this, this, this, this, etc.

I don't see the word FreeBSD anywhere, do you? Did you think maybe SGI is "obscuring" FreeBSD because it runs on their 512 CPU SSI Servers so much better than Linux? Or the alternative explanation is that FreeBSD (even the current 6 branch) barely scales to 4 CPUs.

Oh, and don't get me started on clueless idiots.

Allow me to refute your alternative possible explanation with *facts*. :)

[snip netcraft]


Ha ha. Very funny. You know, it has been nearly a decade since you needed a really good OS to serve websites, right?

No, if you want to be a Linux contender, you have to do this, this, this, this, this, this, this, this, etc.

I don't see the word FreeBSD anywhere, do you? Did you think maybe SGI is "obscuring" FreeBSD because it runs on their 512 CPU SSI Servers so much better than Linux? Or the alternative explanation is that FreeBSD (even the current 6 branch) barely scales to 4 CPUs.

Oh, and don't get me started on clueless idiots.

Allow me to refute your alternative possible explanation with *facts*. :)

[snip netcraft]


Ha ha. Very funny. You know, it has been nearly a decade since you needed a really good OS to serve websites, right?

No, if you want to be a Linux contender, you have to do this, this, this, this, this, this, this, this, etc.

I don't see the word FreeBSD anywhere, do you? Did you think maybe SGI is "obscuring" FreeBSD because it runs on their 512 CPU SSI Servers so much better than Linux? Or the alternative explanation is that FreeBSD (even the current 6 branch) barely scales to 4 CPUs.

Oh, and don't get me started on clueless idiots.