Slashdot Mirror

Windows 2000 with SQL Server not only tops the performance charts, but it also owns every single spot on the performance/price charts.

free!=better

lets see INTEL go up aganst a SUN on a large oracle DB then I will take notice

Actually, Intel systems do pretty well, indeed, better than Sun running Oracle with a 3000G test database. And they do a good job on transaction throughput too.

lets see INTEL go up aganst a SUN on a large oracle DB then I will take notice

Actually, Intel systems do pretty well, indeed, better than Sun running Oracle with a 3000G test database. And they do a good job on transaction throughput too.

it kicked sql servers ass on eweek

I quote from the eweek article:

"Due to its significant JDBC (Java Database Connectivity) driver problems, SQL Server was limited to about 200 pages per second for the entire test..."

Uhhh yeah, sounds like a fair benchmark... using JDBC drivers that are known to be buggy with SQL Server 2000? I, and most people using SQL Server, don't use JDBC drivers for most things.

But I guess if you only read the front pages of mysql.com for information on benchmarks, you might like to believe MySQL won...

Try the TPC page, where Microsoft SQL Server 2000 owns the top 3 spots for performance. This is comparing SQL Server 2000 vs. Oracle on all kinds of high end machines (presumably to remove the hardware as the bottleneck). I think most people would agree Oracle is considered to be the top-end RDBMS, and here MS SQL 2000 beats it.

IBM is introducing low-end servers based on the Power4. They compete head-on with the V880 from Sun Microsystems (NASDAQ: SUNW). The V880 is the only profit generator for SUNW right now; the high-end SunFire 15K is almost unsellable. The new IBM servers will dent the sales of the V880 and bring its days as a profit stream for SUNW to an end.

With the new p670, IBM now has the entire range of servers from the low-end to the high-end covered by Power4-based systems. Each system outperforms the equivalent (in terms of the number of processors) system from SUNW. Worse, at the end of May, the Oracle DBMS will run on the Power4. Finally, the TPC-C numbers will be out. SUNW's systems will be crushed.

Just read the performance numbers at SPEC and TPC .

If you look at the related articles at CNET, you will notice something that is terribly wrong with SUNW. The spokesman for SUNW is Shahin Khan. The spokesman for IBM is Ravi Arimilli. Shahin Khan is a marketing drone. Ravi Arimilli is an IBM Fellow, firmly grounded in engineering. Khan is telling Arimilli why SUNW's systems are technically superior! (Read "IBM, Sun to release dueling servers".)

Finally, on the day that IBM unveils its arsenal of Power4-based servers, SUNW rushes to cut prices. (Read "Prices Lowered on Popular Sun Fire Server Family".) SUNW appears to be running scared. You should dump SUNW stock because SUNW will not achieve profitability by 2002 June. How do you achieve profitability by slashing prices and, hence, margins?

You mean like the top three database performance records on TPC? Or all of the top ten performance to price ratio records? Unisys' ES7000 DataCenter server, which is a single 32-CPU Pentium Xeon machine is taking on machines with twice as many CPUs (the top performer in the non-clustered arena is a Sun StarFire with 128 CPUs.) This machine is also fully load-balancing and failover clusterable, although Unisys has not posted any benchmarks of that kind of setup. The system is powerful, and proves itself. In order to even participate in the program Microsoft requires Unisys to guarantee to their clients 99.99% system availability.

BSoDs? Kernel panics.

You mean like the top three database performance records on TPC? Or all of the top ten performance to price ratio records? Unisys' ES7000 DataCenter server, which is a single 32-CPU Pentium Xeon machine is taking on machines with twice as many CPUs (the top performer in the non-clustered arena is a Sun StarFire with 128 CPUs.) This machine is also fully load-balancing and failover clusterable, although Unisys has not posted any benchmarks of that kind of setup. The system is powerful, and proves itself. In order to even participate in the program Microsoft requires Unisys to guarantee to their clients 99.99% system availability.

BSoDs? Kernel panics.

You mean like the top three database performance records on TPC? Or all of the top ten performance to price ratio records? Unisys' ES7000 DataCenter server, which is a single 32-CPU Pentium Xeon machine is taking on machines with twice as many CPUs (the top performer in the non-clustered arena is a Sun StarFire with 128 CPUs.) This machine is also fully load-balancing and failover clusterable, although Unisys has not posted any benchmarks of that kind of setup. The system is powerful, and proves itself. In order to even participate in the program Microsoft requires Unisys to guarantee to their clients 99.99% system availability.

BSoDs? Kernel panics.

This was my final year project thesis. Just remember the golden rule unstructured 2 structured == convert 2 XML I wrote a [very bad] program in C++/Perl/tcsh IPC=pipes to add XML tags to English, and then index them into a search engine which would use the lingual data stored in the XML tags to help the search.

NIST does a MASSIVE competition on this annually. I don't want to be an XML-buzzword whore <Arnold Schwarzenegger accent> (XML commando eats Green berets, C++, Java, Perl, COBOL for breakfast)</Arnold Schwarzenegger accent> but you can't beat XML for easily converting anything that you can make sense out of into computer readable format. Real h3cKoRs use SGML, but us underlings have to stick with things we can understand like XML. As for expandability, if we want to encode something else into the document, then just tag-it-and-go

It took me 200 hours to fish out all these links (before the Google days), I don't want anyone to have to waste as much time as I did feeding the search engines exotic foods. It's a year old so pardon me for the odd broken link, armed with these you could probably turn jello into XML ;-)

My favourite bookmarx
PROJect[21 links]
Beginners' Guide[13 links]
Berkeley Linguistics Dept. Course Summaries, general stuffzzzzzzzzzzzzzzCryptic IR Vocabulary defined
Explanations of weird words like hypernym zzzzzzzzzzzzzzHow do we produce and understand speech
How Inverted Files are Created - Univeristy of Berkeley zzzzzzzzzzzzzzNLP Univ. of Indiana, very good basics e.g. word sense d
Simple langauge - useful.... zzzzzzzzzzzzzzWhat is Natural Language Processing, links
What is POS tagging........ zzzzzzzzzzzzzzWord Sense Disambiguation defined
Word Sense Disambiguation in detail, scroll down far zzzzzzzzzzzzzzWord Sense Disambiguator - LOLITA (tested at MUC-7 and SENSEVAL competition as best)
XML for the absolute beginner

HTML, XML stuff + parsers[19 links]
Apache plug-in that uhhh does stuff with XML zzzzzzzzzzzzzzConvert COM to XML
convert XML, HTML to Unix pipeable formats zzzzzzzzzzzzzzconverters to and from HTML
expat XML parser zzzzzzzzzzzzzzHTML Tidy - converts HTML 2 XML + source code!!
Parse DB (RDBMS, whatever) to XML zzzzzzzzzzzzzzPerl-XML Module List
PHP Manual XML parser functions - what the hell are they talking about, PHP Virtual M... zzzzzzzzzzzzzzPublic SGML-XML Software
Pyxie - XML Processor for Python, Perl, etc. zzzzzzzzzzzzzzSGML+XML tools.org
The XML Resource Centre - massive number of links zzzzzzzzzzzzzzW4F wrapper - wrapper converts XML to HTML
XFlat - convert flat file into XML zzzzzzzzzzzzzzXML Parsers and other XML stuff
XML.com - Parsers, etc. zzzzzzzzzzzzzzXML-Data Catalog System - uhhhh looks close
XTAL's general converter - convert anything 2 XML

other Background[8 links]
Is Linux ready for the Enterprise, scalable... zzzzzzzzzzzzzzLinux reliability
Linux Versus Windows NT, Mark(sysinternals bloke) zzzzzzzzzzzzzzPC reliability (pcworld)
SPEC - Standard Performance Evaluation Corp. zzzzzzzzzzzzzzSystems benchmarks
TPC - Transaction Processing Performance Council zzzzzzzzzzzzzzUnix Beats Back NT In EDA Workstation Arena
Proper TREC(-8) QA systems[2 links]

pg. 387 LIMSI-CNRS pretty deep parsing[2 links]
More links....
NLP, IR links - lots to corpii, etc.

pg. 575 U. of Ottawa and NRL (shit system, got 0%)[1 links]
LAKE Lab
pg. 607! University of Sheffield (crap system, but OPEN SOURCE!)[2 links]
GATE - FREE IE app w`source code
LaSIE - ER, coreference, template (cv)

pg. 617 Univ of Surrey (inconclusive matches)[2 links]
System Quirk - Or is this their search system..... Hmmmmmm
Univ of Surrey - pointers (hopefully this is their WILDER search system...)

SMU - Pg. 65[1 links]
Natural Language Processing Laboratory at SMU

Textract[2 links]
Cymfony - Technology
Textract - State of the Art Information Extraction

Xerox uhhhhh maybe[1 links]
Xerox Palo Alto Research Center
(OVERVIEW) 1999 TREC-8 Q&A Track Home Page
NLP bloke, Univ Sussex


Tcl-Tk[4 links] Tcl tutorial
Tcl-Tk Contributed Programs Index
Tcl-Tk Resources, sources
TclXML - manipulating XML using Tcl-Tk
Artificial Natural Language - Is this what I'm trying to parse into...
Comparison of Indexers - Prise vs. Inquery vs. MG, etc.
Eagles - Language Engineering Standards
Language Technology Group - lots of modules!
LDC - Linguistic Data Consortium, lots of corpora
Lexical Resources
Links 2 resources, indexers.....
Lots of IR stuff, University of uhhh
Managing Gigabytes Indexer
Managing Gigabytes Manuals and stuff
Htdig search system
NLP & IR (NLPIR, NIST) Group
OVERVIEW OF MUC-7-MET-2
Perl XML Indexing - XML search engine type thing
Phrasys Language Processing Software Components (money)
QA HCI bullshit
SIGIR - TREC-type thing, resources
SMART indexer system documentation
Text REtrieval Conference (TREC) Home Page
The Natural Language Software Registry
Thunderstone IE and IR products
WordNet - FREE DOWNLOADABLE lexical English database

Page created with URL+, nice utility for working with internet shortcuts

Indeed, both the Power4 and the Pentium 4 (P4) significantly outperform an UltraSPARC III. Just visit the SPEC web site and the web site of the Transaction Processing Council (TPC) . The performance results at those web sites show that the UltraSPARC III significantly underperforms against the competition. For this very reason, Sun is refusing to run Linux on the UltraSPARC III. Running Linux on UltraSPARC III would allow an even more direct, head-on comparison between the UltraSPARC III and the Power4 or P4. Same OS (Linux). Yet, Power4 and P4 outperform UltraSPARC III. There would be no way for Sun to say, "Well, Solaris causes the UltraSPARC III to run slower than the competition because Solaris is using the extra CPU cycles to give you that much more reliability."

As the official line, Sun Microsystems derides with the SPEC benchmarks and the TPC benchmarks as being unfair and unrepresentative of the "real world". How can any company utter such asinine comments? Both SPEC and TPC are fair, reputable organizations that have set forth to provide a fair and unbiased means to compare a range of computer systems from various companies. You might say that both SPEC and TPC are the "Consumer Reports" of the server market.

To look at something that is really asinine, I highly recommend Big Blue Smoke , which is a childish web site that Sun established to ridicule IBM. Sun must be getting really desperate.

One of these clusters handles up to two hundred millions transactions / day.

You cannot do that with x86, sorry.

"I am sorry", but Compaq and tpc.org disagree with you, read this. Mainframe processors aren't that powerful, the Xeon series beats all of them, and you can also find reliable and fault-tolerant hardware on the x86 market.

BTW, you should tell Google.com that they can't have such a busy and critical service (as their x86 boxes represent the whole bussiness of Google.com, they're the most critical part of their company), they would be glad to listen to your opinion.

<flame>Just because your admin is lazy enough to not want a distributed environment, and wants systems where all the hardware work is done by IBM support staff, it doesn't mean that x86 aren't capable of performing the same task as mainframes</flame>

AMD will enjoy a short period of incredible success with Hammer, its 64-bit 0x86 architecture. Intel will see the success and immediately release its own 64-bit extension to the basic IA32 architecture. Both chips will be the foundation of commodity, ultra-low cost, servers.

These servers will annihilate Sun in the low-end to mid-range portion of the server market. These servers will gradually creep into the high-end of the server market, where machines having 32 or more processors dominate.

Sun has seen the writing on the wall. As a last desperate measure, Sun has announced that it too will sell Intel/AMD-powered servers running Linux more than 1 year after IBM has been successfully doing the same.

Just look at the performance data at SPEC and TPC . The x86 processors crush UltraSPARC III across a broad range of benchmarks.

An answer already lies on TPC, their top three performing machines are Windows clusters, because Windows 2000 and SQL Server 2000 cluster very well together.

Top Ten TPC-C by Performance

One reason cited for why Reporters (those who find flaws) go public:
...malicious intent to damage the reputations of specific vendors

If you find a glitch with a competitors product, why is it automatically evil to publicly disclose? One valid tactic of advertising (in the US) is to denigrate competing products. When Microsoft announced that Windows NT beat Linux in performace tests, did they give Linus private notice and 30 days to respond before issuing the press release? Does Compaq let IBM know beforehand that it has better TPC numbers? After Dateline NBC staged exploding gas tanks, did the Wall Street Journal give them a month to come clean before revealing the scam?

If you worked at Be in 1998 knew of a fundamental, nearly unfixable flaw in Windows, how much time would you grant Microsoft to address it?

By the way - this record-setting TPC-H benchmark was set with a Sun E15K.

THe speed of SQL Server 2k comes from its ability to horizontally partition tables across a cluster. Its clustered and its cheaper. You need to compare SQL server is limited to clusters because it runs on Intel hardware. Since when isnt' cost important? and now clusters are shitty?. an oracle *nix solution costing 2x as much and not as fast is better? i am sure Oracle or DB2 might be the best choice right now because of maturity but claiming something is crap when its plain faster and cheaper is silly. If Oracle was at the top you'd be quoting the results. If you look at who sponsors tpc you will notice some pretty familiar names. larry ellison is just as big of a greedy bastard as bill gates is. Oracle has been over charging for its products for years.

Why not look at the "All" performance results, rather than just "Non-Clustered":

Top Ten TPC-C by Performance

Paints a whole different picture, doesn't it?

Okie. Let's look at the sheer performance then, using the same TPC benchmarks.

Top Ten TPC-C by Performance

Let's see... The top 3 spots are SQL 2000 running on Windows 2000 clusters.

Of course, if that's too "cheap" for you, you can spend an additional $2M for a "high-end" runner-up Fujitsu machine that will give you approx 60% performace (455K tpmC vs. 709K tpmC) of the Windows/SQL 2000 solution.

Okie. Let's look at the sheer performance then, using the same TPC benchmarks.

Top Ten TPC-C by Performance

Let's see... The top 3 spots are SQL 2000 running on Windows 2000 clusters.

Of course, if that's too "cheap" for you, you can spend an additional $2M for a "high-end" runner-up Fujitsu machine that will give you approx 60% performace (455K tpmC vs. 709K tpmC) of the Windows/SQL 2000 solution.

Yawn... Microsoft makes an inexpensive database. According to the price/performance table, it seems to be really great. On the other hand, if you look at JUST performance numbers, the difference becomes clear. SQL server doesn't even make the list Oh well, I guess you get what you pay for.

Yeah, SQL Server really sucks (NOTE: SARCASM INTENDED).

The Cray T3E weighs in at up-to-3-TFLOPS; depending on number of processors. Of course, this machine costs over $10,000,000.

For something a little more practical and realistic, the extremely-fast yet value priced Compaq AlphaServer rings in at 47 GFLOPS.

Granted, FLOPS aren't a very good judge of speed for this application, but they are easy stats to find. If you really want a standardized test, take a look at the TPC-C stats for the fastest cluster machines in the world. These more accurately reflect the kind of performance stats you're looking for in relation to this article.

As others have already pointed out, Itanium does fairly badly on SpecINT and moderatly well (though not spectacularly) on SpecFP, but there are other benchmarks to consider: Stream and TPC are the two that leap to mind, but I am sure that there are some others that measure more than just raw CPU performance. (you won't be running any of your code on a bare CPU anyway, so you should consider some measures of full system performance) I don't see any numbers for Itanium on the Stream or TPC web sites, but maybe I missed something.

The "proven JVM" - is that as in "proven to be slower than a one-legged caterpillar"? Why is there not a single benchmark on the TPC performance site that includes Java technology? However, if you want to see an "enterprise scaled application using COM, MTS, or COM+" there are a legion of examples on the site...

ADO, OLE DB, ODBC and DB-Library all expose the functionality of cursors through recordsets. From my experience recordset operations have to be one of the most common operations of database applicaions. How often do you open a recordset and loop through the records? Sound like the functionality of a cursor? Is it a coincidence that there are arguments to OpenRecordset methods (such as adOpenForwardOnly, adOpenKeyset) that map almost directly to arguments that are passed when creating T-SQL cursors. These API's create cursors when you use OpenRecordset. So basically from your statement you can conclude that one of the most common data operations performed against an MSSQL database from a client application is absolutely horrible in performance. I'm not buying it. Neither are the millions of applications and developers using ADO and ODBC recordsets with a SQL server database - I think somebody might have noticed.

Now since we know it's not cursors that are slow, perhaps it's just the database product itself? Nope, it's the fastest tested database in the world in several areas.

Looking over the "paper" I noted some interesting things on just a quick viewing: 1)under "Performance Date" item 2 "GNU/Linux was the May 2001 performance leader in the TPC-H decision support (database) benchmark (``100Gb'' category)"
Um yes, they did, but they did it on a machine that costs $948966.00. System description It was one of the most expensive machines in the running. The number 2 machine is an Win2k / SQL Server 2000 machine for a third the price. The Top Ten price / performance list is dominated by Windows 2000 / SQL Server 2000. TPC.org
2)The count of web servers in operation is a bit misleading as the source of the information states that "...host addresses of the .edu domain were used..." and if you look at the report there are NO .com domains represented. Well, gee. I wonder why there are so many linux boxes in the report? Just pointing out that statistics represent those that present them.

Looking over the "paper" I noted some interesting things on just a quick viewing: 1)under "Performance Date" item 2 "GNU/Linux was the May 2001 performance leader in the TPC-H decision support (database) benchmark (``100Gb'' category)"
Um yes, they did, but they did it on a machine that costs $948966.00. System description It was one of the most expensive machines in the running. The number 2 machine is an Win2k / SQL Server 2000 machine for a third the price. The Top Ten price / performance list is dominated by Windows 2000 / SQL Server 2000. TPC.org
2)The count of web servers in operation is a bit misleading as the source of the information states that "...host addresses of the .edu domain were used..." and if you look at the report there are NO .com domains represented. Well, gee. I wonder why there are so many linux boxes in the report? Just pointing out that statistics represent those that present them.

I'm shooting in the dark here, but I'd wager that it's not incredibly likely that Casinos (or anyone who actually needs a RDBMS) are running it on 486SX', so as a metric that is rather worthless (sort of like saying a Acura NSX is slow when you hook a truck trailer to it). On a real machine SQL Server will rock any system's world, including the big big boys.

Oh yeah, COM+ didn't work and didn't scale.

I call bullshit.

Apparently you haven't been to the transaction processing council's benchmark page recently?

COM+ and SQL Server 2000 currently hold 9 of the top 10 benchmark results for the TCP-C Clustered results.

Don't make comments you can't back up.

[This tagline was intentionally left unblank]

TPC-C is the reference benchmark. Linux has no entry there. http://www.tpc.org/tpcc/results/tpcc_perf_results. asp Also, this is a measure of the Database software speed. DB2 and SQL2000 are quite fast and on any 32-bit x86 OS they would beat the pants off of mySQL or some of those other freebies. Look at the TPC-C benchmark to see the real battle between Microsoft and Sun.

Yeah. Looks like they're lagging way behind.
Also, ODBC? I hear it scales _really_ well. Doesn't CPAN have any Perl modules to natively connect/execute stored procs? I thought it had everything?

A clip from this story, posted today actually:

Using the SAP Standard Application Sales and Distribution (SD) Benchmark, an industry-standard measure of server performance, a Unisys e-@ction Enterprise Server ES7000 equipped with 32 Intel Pentium III Xeon 32-bit processors supported 18,500 mySAP.com SD Standard Application benchmark users. This result is the third highest result ever recorded on any platform tested with the SAP SD benchmark methodology, regardless of the number of processors per server tested.

Oh, and don't mention the Transaction Processing Performance Councel, by Performance or by price/tpmC (a hint: MS has 10 of the... top ten), or heck, just overall!

Yes, MS has made some mistakes in the past, but they are learning from them and are making a quiet comeback. Nothing comes close to touching thier data mining/warehousing product.

A clip from this story, posted today actually:

Using the SAP Standard Application Sales and Distribution (SD) Benchmark, an industry-standard measure of server performance, a Unisys e-@ction Enterprise Server ES7000 equipped with 32 Intel Pentium III Xeon 32-bit processors supported 18,500 mySAP.com SD Standard Application benchmark users. This result is the third highest result ever recorded on any platform tested with the SAP SD benchmark methodology, regardless of the number of processors per server tested.

Oh, and don't mention the Transaction Processing Performance Councel, by Performance or by price/tpmC (a hint: MS has 10 of the... top ten), or heck, just overall!

Yes, MS has made some mistakes in the past, but they are learning from them and are making a quiet comeback. Nothing comes close to touching thier data mining/warehousing product.

A clip from this story, posted today actually:

Using the SAP Standard Application Sales and Distribution (SD) Benchmark, an industry-standard measure of server performance, a Unisys e-@ction Enterprise Server ES7000 equipped with 32 Intel Pentium III Xeon 32-bit processors supported 18,500 mySAP.com SD Standard Application benchmark users. This result is the third highest result ever recorded on any platform tested with the SAP SD benchmark methodology, regardless of the number of processors per server tested.

Oh, and don't mention the Transaction Processing Performance Councel, by Performance or by price/tpmC (a hint: MS has 10 of the... top ten), or heck, just overall!

Yes, MS has made some mistakes in the past, but they are learning from them and are making a quiet comeback. Nothing comes close to touching thier data mining/warehousing product.

A clip from this story, posted today actually:

Using the SAP Standard Application Sales and Distribution (SD) Benchmark, an industry-standard measure of server performance, a Unisys e-@ction Enterprise Server ES7000 equipped with 32 Intel Pentium III Xeon 32-bit processors supported 18,500 mySAP.com SD Standard Application benchmark users. This result is the third highest result ever recorded on any platform tested with the SAP SD benchmark methodology, regardless of the number of processors per server tested.

Oh, and don't mention the Transaction Processing Performance Councel, by Performance or by price/tpmC (a hint: MS has 10 of the... top ten), or heck, just overall!

Yes, MS has made some mistakes in the past, but they are learning from them and are making a quiet comeback. Nothing comes close to touching thier data mining/warehousing product.

Here's a nice example of the bloat for y'all. Try those benchmarks on a properly configured system.

A perfectly reasonable response. No one is going to force you to buy a major db, with the possible exception of a boss. You have every right to not buy any product you wish, just like they have the right to restrict what can be done with their product.

As much as I don't like to admit it, I can see the DB vendors point here. The major DB systems take a tremendous amount of knoweldge to tune. They don't want a published benchmark to show their product is inferior simply because someone didn't know how to set it up.

On the other hand, I think there are results that can be published if they are properly audited.

Let's just say they're slightly better than MySQL on Linux.

Something is definitely fishy with their hardware if Win2K is twice as slow as NT4. I've run both servers with SQL7 intensively. You couldn't pay me to move back to NT4. 2000 isn't all that much faster, but it is much more stable and its a lot easier to use and administrate.

Want some real benchmarks? Try here. Notice a pattern? SQL Server is the fastest database server in the world. Not only that, but Win2K is in the top four slots. 2nd place is a DB2 server on Win2K. Here are real, industry standard tests performed by an independent organization, not a company with an agenda to promote or magazines to sell.

I'm not sure what the point of this article is, other than to stir up more mindless MS-bashing. Well, Timothy, maybe you should try SQL Server or another real database. Pretty much every day around noon we get the same problem because Slashdot can't handle displaying stories while lots of people are posting. A real database would do wonders to fix that.

http://www.tpc.org/tpcc/results/tpcc_perf_results. asp offers some answers. Of course few folks who read slashdot actually care about which database might actually be best. Instead they blindly claim JimmyBob's Free DB with Source is best followed closely by the Mr. Larry Ellison's Oracle. Fine. Sign a deal with the devil.

While all you blowhards are trumpeting Oracle, you might consider the cost and performance leaders. See http://www.tpc.org for a constantly updated list.

Check out the TPC benchmarks (a standardized set of very strict performance tests) for your answer. Pick either by raw performance or by price/performance. You don't have to be a genius to understand the numbers: http://www.tpc.org/

9. Performance - simple databases are fast with simple queries, but sophisticated databases are fast with complex queries (and, maybe, slower with simple queries...). M$ uses simple benchmarks to tell their customers that they are fast (TCP-C), but most other vendors have stopped three years ago to post TCP-C benchmarks, so M$ is able to beat three year old databases on three year old hardware... TCP-D is more complex, and M$ couldn't show that their database wasn't 100 (one hundred) times slower on a specific TCP-D part than Oracle - though they tried hard. But benchmarking is a complex matter...

Are you trolling or astroturfing for Oracle?

Go visit the TPC website. Funny how they don't mention the "all important" TPC-D benchmark on their front page. The TPC-D benchmark is listed under the obsolete section under "benchmarks".

Now go visit the TPC-C results page.. The last posted results by Oracle are from January the 18th 2001. Doesn't quite sound like something they don't care about to me. (Although the results are listed as being posted by the hardware manufacturer, anyone who has worked with benchmarks knows it's impossible to post strong results without strong support from the database vendor).

9. Performance - simple databases are fast with simple queries, but sophisticated databases are fast with complex queries (and, maybe, slower with simple queries...). M$ uses simple benchmarks to tell their customers that they are fast (TCP-C), but most other vendors have stopped three years ago to post TCP-C benchmarks, so M$ is able to beat three year old databases on three year old hardware... TCP-D is more complex, and M$ couldn't show that their database wasn't 100 (one hundred) times slower on a specific TCP-D part than Oracle - though they tried hard. But benchmarking is a complex matter...

Are you trolling or astroturfing for Oracle?

Go visit the TPC website. Funny how they don't mention the "all important" TPC-D benchmark on their front page. The TPC-D benchmark is listed under the obsolete section under "benchmarks".

Now go visit the TPC-C results page.. The last posted results by Oracle are from January the 18th 2001. Doesn't quite sound like something they don't care about to me. (Although the results are listed as being posted by the hardware manufacturer, anyone who has worked with benchmarks knows it's impossible to post strong results without strong support from the database vendor).

High-end RDBMs (Oracle, MS-SQL, DB2) are beasts of a whole different nature from your regular, run-of-the-mill MySQL or Posgress installation.

First they have a price tag attached. A very high-end price tag. The figure you give for Oracle is the barely minimum a company would expect to expend at the lowest entry-level.

Second, they all must be baby-sited. You do not install Oracle from a CD, make it run and leave it alone. No, you musrt first pain-stakingly install it, sweat and curse for a couple of hours to make it run and them administer it forever. Leave it in the hands of the users for a couple of weeks and the meaning of the word entropy will soon be known at the upper organization levels.

Allright, these above are a bit dramatic but true. In a sense, RDBMS makers and specialists are fighting hard to defend the last outpost of what Allan Kay once called "the high priesthoood of a low cult" (talking about the computer people of the 60's).

Now, high-end software is for high-end jobs. First, they have a bunch of weird performance marks to show at Transaction Processing Performace Council results for TPC-C, for instance. Carrefully note the kind of hardware they are talking about. They are measuring RDBMS performance in machines costing millions of dollars!!

And then these beasts are capable of performance peaks unheard of in the free packages. Not that I do not like MySQL. I like it a lot and have used it in many enterprise systems I developed. But those were small enterprise systems. They did not had to sustain high transaction periods, nor high connection counts, nor gigabyte sized database tables. And certanly not all three together.

I am currently working in the development of a system for a large laboratory. They process thousands of clinical exams per day, mostly in the morning. The organic material for the exams must usually be processed within two to four hours after it is collected from a patient. We will pay whatever it takes to have peace of mind. And I tell you, those will be some well expend millions of dollars.

Incidentally, SQLServer is currently the holder of the fastest database setup in the world, as measured in transactions per second by independant council TPC, on an expensive custom hardware/software setup, designed by Microsoft.

Tartley

There's a reason they call me that. We won't go into it here.

Incidentally, SQLServer is currently the holder of the fastest database setup in the world, as measured in transactions per second by independant council TPC, on an expensive custom hardware/software setup, designed by Microsoft.

Tartley

There's a reason they call me that. We won't go into it here.

Incidentally, SQLServer is currently the holder of the fastest database setup in the world, as measured in transactions per second by independant council TPC, on an expensive custom hardware/software setup, designed by Microsoft.

Tartley

There's a reason they call me that. We won't go into it here.

The system from the benchmark report has 24 RS64-IV 64-bit processors running at 600 Mhz with 96GB (yes, GB) of system DRAM. Each processor has 128kB L1 data cache, 128kB L1 instruction cache, and a 16MB L2 cache. The chips also support course-grain multithreading (simpler, but similar to SMT).

(600 Mhz sounds slow until you realize that it uses a simple, very efficient 5-stage pipeline. Intel and others achieve high clock rates through deep piplines and rely on branch prediction and other techniques to keep the pipe full. Branch mispredictions and cache misses can kill the actual performance of these chips on real server code.)

This system with 24 processors outperforms HP's 48 processor "SuperDome" and Sun's 64 processor EU10k (though the UE10k is an old system by now, it is the fastest server Sun is shipping.)

The above system is not using the Power3 chip from the posted story. You can bet IBM will port Linux to this beast next. We won't see a 24 processor systems with Linux right away, but an s80-like system would make a sweet 4-processor Linux server.

One last note: these systems are not vapor-ware. A 12-processor system with an earlier version of the same processor has been shipping since the summer of '98.

The system from the benchmark report has 24 RS64-IV 64-bit processors running at 600 Mhz with 96GB (yes, GB) of system DRAM. Each processor has 128kB L1 data cache, 128kB L1 instruction cache, and a 16MB L2 cache. The chips also support course-grain multithreading (simpler, but similar to SMT).

(600 Mhz sounds slow until you realize that it uses a simple, very efficient 5-stage pipeline. Intel and others achieve high clock rates through deep piplines and rely on branch prediction and other techniques to keep the pipe full. Branch mispredictions and cache misses can kill the actual performance of these chips on real server code.)

This system with 24 processors outperforms HP's 48 processor "SuperDome" and Sun's 64 processor EU10k (though the UE10k is an old system by now, it is the fastest server Sun is shipping.)

The above system is not using the Power3 chip from the posted story. You can bet IBM will port Linux to this beast next. We won't see a 24 processor systems with Linux right away, but an s80-like system would make a sweet 4-processor Linux server.

One last note: these systems are not vapor-ware. A 12-processor system with an earlier version of the same processor has been shipping since the summer of '98.