TPC-H benchmark & Linux/Microsoft results.

[Sweede]

First, lets figure out what the TPC-H benchmarks are.

The TPC Benchmark™H (TPC-H) is a decision support benchmark. It consists of a suite of business oriented ad-hoc queries and concurrent data modifications. The queries and the data populating the database have been chosen to have broad industry-wide relevance. This benchmark illustrates decision support systems that examine large volumes of data, execute queries with a high degree of complexity, and give answers to critical business questions.

The performance metric reported by TPC-H is called the TPC-H Composite Query-per-Hour Performance Metric (QphH@Size), and reflects multiple aspects of the capability of the system to process queries. These aspects include the selected database size against which the queries are executed, the query processing power when queries are submitted by a single stream, and the query throughput when queries are submitted by multiple concurrent users. The TPC-H Price/Performance metric is expressed as $/QphH@Size.


seems reasonable, the number of database queries that can be performed per hour.

lets loot at one line.

These aspects include the selected database size against which the queries are executed, the query processing power when queries are submitted by a single stream, and the query throughput when queries are submitted by multiple concurrent users


lets look at the results..


Rank #1.

Rank Company System QphH Price/QphH System Availability Database Operating System Date Submitted
1 SGI 1450 Server with DB2 UDB EEE v7.2 2733 347 US $ 10/31/01 IBM DB2 UDB EEE 7.2 Linux 2.4.3 05/11/01

Total System Cost 948,966.00 US $
Metric 2733.80 QphH@100GB
Price/Performance 347.00 US $ per QphH@100GB
Availability Date 10/31/01
Database Manager IBM DB2 UDB EEE 7.2
Operating System Linux 2.4.3


Server Information CPU: Intel Pentium III Xeon 700MHz
# of CPUs: 16
Cluster: Y
Load Time (hours): 2.23
Total Storage/Database Size Ratio: 44.880000000000003



347$ per Query per Hour @ 100 gigs with a clustered 16 CPU system.

Total system cost is almost 1 million.


Rank #2.

Rank Company System QphH Price/QphH System Availability Database Operating System Date Submitted
2 ProLiant 8000-X700-8P 1699 161 US $ 08/01/00 Microsoft SQL 2000 Microsoft Windows 2000 07/21/00

System Information Total System Cost 273,689.00 US $
Metric 1699.80 QphH@100GB
Price/Performance 161.00 US $ per QphH@100GB
Availability Date 08/01/00
Database Manager Microsoft SQL 2000
Operating System Microsoft Windows 2000


Server Information CPU: Intel Pentium III Xeon 700MHz
# of CPUs: 8
Cluster: N
Load Time (hours): 4.59
Total Storage/Database Size Ratio: 13.4



161$ per Query per Hour @100 gb.
a NON clustered Single 8 CPU Intel system,
1/3rd the price at $270,000, 1/2 the per query cost, but half the speed.

well, get two of these Compaq systems, cluster them and you have a Linux killer thats still cheaper.

I love the TPC-? benchmarks, here's another data-set.
a little info..

TPC benchmarks also differ from other benchmarks in that TPC benchmarks are modeled after actual production applications and environments rather than stand-alone computer tests which may not evaluate key performance factors like user interface, communications, disk I/Os, data storage, and backup and recovery. The difficulty in designing TPC benchmarks lies in reducing the diversity of operations found in a production application, while retaining its essential performance characteristics, namely, the level of system utilization and the complexity of its operations

the TPC-C benchmarks are REAL LIFE production (say, amazon.com book database ) functions/systems, etc.

who's on number 1 there?

top 10 in total performance.

Rank Company System tpmC Price/tpmC System Availability Database Operating System TP Monitor Date Submitted Cluster
1 IBM/eServer xSeries 370 688,220 28.89 US $ 05/31/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/06/01 Y
2 ProLiant 8500-700-192P 505,302 19.80 US $ 11/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 10/06/00 Y
3 Netfinity 8500R c/s 440,879 32.28 US $ 12/07/00 IBM DB2 UDB 7.1 Microsoft Windows 2000 Microsoft COM+ 07/03/00 Y
4 IBM eServer xSeries 370 363,129 28.10 US $ 05/31/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/06/01 Y
5 ProLiant 8500-X700-96P 262,243 20.24 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00 Y
6 ProLiant 8500-X550-96P 229,913 23.08 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00 Y
7 PRIMEPOWER 2000 c/s w/ 32 Front-Ends 222,772 51.40 US $ 06/30/01 SymfoWARE Server Enterp. Ed. VLM 3.0 Sun Solaris 8 BEA Tuxedo 6.5 CFS 03/16/01 N
8 IBM eServer pSeries 680 Model 7017-S85 220,807 43.30 US $ 04/13/01 Oracle Oracle 8i Enterprise Edition v. 8.1.7 IBM AIX 4.3.3 Webshpere App. Server Ent. Edition V.3.0 10/13/00 N
*** Escala EPC2450 220,807 43.31 US $ 04/15/01 Oracle 8i Enterprise Edition IBM AIX 4.3.3 Webshpere App. Server Ent. Edition V.3.0 11/07/00 N
9 HP 9000 Superdome Enterprise Server 197,024 53.77 US $ 05/01/01 Oracle8 Enterprise Edition v8.1.7.1 HP UX 11.i 64-bit BEA Tuxedo 6.4 03/05/01 N
10 PRIMEPOWER 2000 c/s w/32 Front-Ends 183,770 56.16 US $ 06/30/01 SymfoWARE Server Enterp. Ed. VLM 3.0 Sun Solaris 7 BEA Tuxedo 6.4 CFS 01/12/01 N


clustered only.

1 IBM/eServer xSeries 370 688,220 28.89 US $ 05/31/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/06/01
2 ProLiant 8500-700-192P 505,302 19.80 US $ 11/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 10/06/00
3 Netfinity 8500R c/s 440,879 32.28 US $ 12/07/00 IBM DB2 UDB 7.1 Microsoft Windows 2000 Microsoft COM+ 07/03/00
4 IBM eServer xSeries 370 363,129 28.10 US $ 05/31/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/06/01
5 ProLiant 8500-X700-96P 262,243 20.24 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
6 ProLiant 8500-X550-96P 229,913 23.08 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
7 ProLiant 8500-X700-64P 179,658 19.75 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
8 ProLiant 8500-X550-64P 161,719 21.86 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
9 IBM e-server xSeries 370 136,766 23.09 US $ 09/20/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/20/01
10 Enterprise 6500 Cluster 135,461 97.10 US $ 01/31/00 Oracle8i Ent. Edition 8.1.6.0 Sun Solaris 2.6 BEA Tuxedo 6.3 09/24/99




Home Results Benchmarks Technical Articles Related Links What's New About the TPC Who We Are
©2001 TPC. All Rights Reserved.





non-clustered

Rank Company System tpmC Price/tpmC System Availability Database Operating System TP Monitor Date Submitted
1 PRIMEPOWER 2000 c/s w/ 32 Front-Ends 222,772 51.40 US $ 06/30/01 SymfoWARE Server Enterp. Ed. VLM 3.0 Sun Solaris 8 BEA Tuxedo 6.5 CFS 03/16/01
2 IBM eServer pSeries 680 Model 7017-S85 220,807 43.30 US $ 04/13/01 Oracle Oracle 8i Enterprise Edition v. 8.1.7 IBM AIX 4.3.3 Webshpere App. Server Ent. Edition V.3.0 10/13/00
*** Escala EPC2450 220,807 43.31 US $ 04/15/01 Oracle 8i Enterprise Edition IBM AIX 4.3.3 Webshpere App. Server Ent. Edition V.3.0 11/07/00
3 HP 9000 Superdome Enterprise Server 197,024 53.77 US $ 05/01/01 Oracle8 Enterprise Edition v8.1.7.1 HP UX 11.i 64-bit BEA Tuxedo 6.4 03/05/01
4 PRIMEPOWER 2000 c/s w/32 Front-Ends 183,770 56.16 US $ 06/30/01 SymfoWARE Server Enterp. Ed. VLM 3.0 Sun Solaris 7 BEA Tuxedo 6.4 CFS 01/12/01
5 IBM eServer iSeries 840-2420-001 163,775 58.88 US $ 12/15/00 IBM DB2 for AS/400 V4 R5 IBM OS/400 V4 R5 BEA Tuxedo 6.4 09/27/00
6 Starfire Enterprise 10000 156,873 48.81 US $ 02/28/01 Sybase ASE 12.0.0.2 Sun Solaris 7 BEA Tuxedo 6.3 08/29/00
7 AlphaSeerver GS320 Model 6/731 155,179 55.68 US $ 02/02/01 Oracle 8i Enterprise Edition v. 8.1.7 Compaq Tru64 UNIX V5.1 Compaq DB Web Connector 11/10/00
8 IBM eServer iSeries 840-2420 152,346 51.50 US $ 09/15/00 IBM DB2 for AS/400 V4 R5 IBM OS/400 V4 R5 BEA Tuxedo 6.4 07/12/00
9 AlphaServer GS320 Model 6/731 144,331 57.25 US $ 09/22/00 Oracle 8i Enterprise Edition v. 8.1.7 Compaq Tru64 UNIX V5.1 Compaq DB Web Connector 09/07/00
10 GP7000F Model 2000 c/w 138,735 74.32 US $ 01/14/01 Oracle8i Ent. Edition 8.1.5 Sun Solaris 7 BEA Tuxedo 6.4 CFS 07/21/00

wow, first and ONLY result set with no windows. non clustered and look at the outragous prices !
the #1 system in this result set has a total cost of 11,449,392 US $ and is slower than a compaq cluster of 96 Xeon 550's that only costs 5,305,571 US , and is more than half the Price/Performance at 23$.


top ten in price performance

Rank Company System tpmC Price/tpmC System Availability Database Operating System TP Monitor Date Submitted Cluster
1 PowerEdge 6400 3P Std 20,331 8.99 US $ 02/02/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 02/02/01 N
2 PowerEdge 6450 3P Std 20,320 9.17 US $ 02/02/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 02/02/01 N
3 ProLiant ML-570-6/700-3P 20,207 9.51 US $ 09/26/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 08/21/00 N
4 PowerEdge 6450/3P 24,925 9.90 US $ 12/20/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 12/20/00 N
*** PowerEdge 6400/3P 24,925 9.91 US $ 12/19/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 12/19/00 N
*** PowerEdge 6450 31,187 10.41 US $ 10/26/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 10/26/00 N
5 PowerEdge 6400 31,187 10.42 US $ 10/26/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 10/26/00 N
6 Proliant ML570 6/700 2 34,324 10.93 US $ 04/19/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Advanced Server Microsoft COM+ 04/19/01 N
7 NetServer LH 6000 33,136 11.25 US $ 03/06/00 Microsoft SQL Server Enterprise Edition 7.0 Microsoft Windows NT Enterprise Edition 4.0 Microsoft COM+ 03/06/00 N
8 ProLiant ML-570-6/700 32,328 12.49 US $ 08/01/00 Microsoft SQL 2000 Microsoft Windows 2000 Microsoft COM+ 07/21/00 N
9 Proliant DL-580 6/900 38,639 12.67 US $ 04/20/01 Microsoft SQL Server 2000 Ente Microsoft Windows 2000 Advanced Server Microsoft COM+ 04/20/01 N
10 HP NetServer LXr 8500 43,046 12.76 US $ 01/09/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 01/09/01 N


clustered systems

Rank Company System tpmC Price/tpmC System Availability Database Operating System TP Monitor Date Submitted
1 ProLiant 8500-X700-64P 179,658 19.75 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
2 ProLiant 8500-700-192P 505,302 19.80 US $ 11/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 10/06/00
3 ProLiant 8500-X700-96P 262,243 20.24 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
4 ProLiant 8500-X550-64P 161,719 21.86 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
5 ProLiant 8500-X550-96P 229,913 23.08 US $ 09/30/00 Microsoft SQL Server 2000 Microsoft Windows 2000 Microsoft COM+ 07/25/00
6 IBM e-server xSeries 370 136,766 23.09 US $ 09/20/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/20/01
7 IBM eServer xSeries 370 121,319 25.47 US $ 05/31/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 03/23/01
8 IBM eServer xSeries 370 363,129 28.10 US $ 05/31/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/06/01
9 IBM/eServer xSeries 370 688,220 28.89 US $ 05/31/01 Microsoft SQL Server 2000 Microsoft Windows 2000 Datacenter Server Microsoft COM+ 04/06/01
10 Netfinity 8500R c/s 440,879 32.28 US $ 12/07/00 IBM DB2 UDB 7.1 Microsoft Windows 2000 Microsoft COM+ 07/03/00


It should also be noted that there are no Linux systems on the TPC-C benchmarks.

[Tyr-7BE]

So what's with the article then? :confused:

By GOD you're an M$ advocate Sweede!

[Unruly]

sweede does enjoy trying to prove us wrong at ever turn we go, with mini-facts. I think he should go someplace else and complain about linux... but hey, free country... :(

[Urko]

Once again I have to reply to Sweede the Fudster and paid Microsoft rep.

We ,unlike you with your paycheck from Microsoft Sweede, advocate linux because we like it.

Now on to your disinformation
1. TPC is completely NON-realworld which is exactly why nobody who is informed pays much atttention.
2. In cost per transaction how can any High cost product like SQL SERVER compete with linux and postgres or mysql which a have cost per transaction of zero. Even you Sweede the person who thinks Quake frame rates is the ultimate test of a good business computer should be able to figure that out.
3. Don't forget all the NT or 2000 Client Access licenses you have to buy to run that crap ,Sweede ,did you factor that into the cost. Oops non-microsoft employees have to pay for MS software Sweede.
4. quote 'just cluster' and how do you do that reliablely and for free with SqlServer Sweede . Oh that's right one of your professors at Bob's college of Knowledge says your the best cluster setter upper and software writer in the whole world and you are going to do yourself personally on your win2003 box where you have been running beta win 2003 for seven years with no reboots Honest it's true

[Sweede]

Urko: that was sad.

if you read my first post, TPC benchmarks are real world results using real world programs in REAL ENVORIMENTS. not a lab setting.

the cost per transaction is the initial investment of the server.
if i spend 10 million on licensing and hardware, but only get 200,000 transactions per minute (the Solaris Box), my cost per transaction is higher than the Windows 2k box that costs 15 million (INCLUDES LICENSES) that does 600,000 transactions per minute (which, btw, is 20$ per transaction).

The systems tested on TPC.org are NOT some techno geek who has no life, but VENDOR systems that they SELL PRECONFIGURED READY TO RUN to companies.


as i said above, the TOTAL cost includes any and ALL licensing fee's.

if TPC says you'll pay $14,312,345.32 for the server, thats how much you'll pay including SunOS, Oracle, DB2, SQL2000, W2k licensing.

a personal freind of mine does his own benchmarking between.


before you try to flame me, read first..

[JasonC]

I think one of the main points in the article has been overlooked, Linux is a player. No matter which side your on, there's not much arguing this point. Linux has gotten better every week, month year and it will continue grow into new markets. It would be interesting to see apples to apples hardware configurations in a comparison. The price/performance ratio would take swing to other side. Unix systems can be $$$$$.

[Sweede]

it's going to be interesting to see what linux does when vendors make Linux systems avilable in the TPC-C and the new TPC-W (Web performance) benchmarks.

the new 2.4.x kernels really are a huge step forward in performance.

Uraldix: The Linux machine really does have about 2 to 2.5 times more computing power than the Number 2 Windows machine, if you installed Linux on the Windows configuration, there'd be a good chance its performance wouldnt be as good.

Another huge factor in these tests is the database system itself, DB2 simply 0wnz in speed.

[Urko]

Sweede TPC is not, I repeat , not a realworld
performance measure . Larry Ellison Ceo of Oracle personally challenged Microsoft to show one realworld application that ran faster on Microsoft cluster than Oracle on Sun. MS answer was our licensing prohibits you from benchmarking your system against ours so we are going to sue you

See this http://slashdot.org/articles/00/10/27/1457215.shtml

Secondly Sweede Microsoft been caught before
doctoring TPC .

Third Microsoft has been to known price things based on the assumption you get there absolute lowest pricing which is never the case in the real world.

Sweede please don't launch a major FUD missile and then say,.. why is everybody always picking on me.

[JasonC]

Originally posted by Sweede:
<STRONG>
Uraldix: The Linux machine really does have about 2 to 2.5 times more computing power than the Number 2 Windows machine, if you installed Linux on the Windows configuration, there'd be a good chance its performance wouldnt be as good.
</STRONG>

I'm not refuting the power gap one bit, I personally would just be interested in a test where hardware was comparable.

[Sweede]

thats great, using a slashdot article to defend your post.


TPC-C simulates a complete computing environment where a population of users executes transactions against a database. The benchmark is centered around the principal activities (transactions) of an order-entry environment. These transactions include entering and delivering orders, recording payments, checking the status of orders, and monitoring the level of stock at the warehouses. While the benchmark portrays the activity of a wholesale supplier, TPC-C is not limited to the activity of any particular business segment, but, rather represents any industry that must manage, sell, or distribute a product or service.

TPC-C involves a mix of five concurrent transactions of different types and complexity either executed on-line or queued for deferred execution. It does so by exercising a breadth of system components associated with such environments, which are characterized by:

The simultaneous execution of multiple transaction types that span a breadth of complexity
On-line and deferred transaction execution modes
Multiple on-line terminal sessions
Moderate system and application execution time
Significant disk input/output
Transaction integrity (ACID properties)
Non-uniform distribution of data access through primary and secondary keys
Databases consisting of many tables with a wide variety of sizes, attributes, and relationships
Contention on data access and update
TPC-C performance is measured in new-order transactions per minute. The primary metrics are the transaction rate (tpmC), the associated price per transaction ($/tpmC), and the availability date of the priced configuration.



TPC-H

The TPC Benchmark™H (TPC-H) is a decision support benchmark. It consists of a suite of business oriented ad-hoc queries and concurrent data modifications. The queries and the data populating the database have been chosen to have broad industry-wide relevance. This benchmark illustrates decision support systems that examine large volumes of data, execute queries with a high degree of complexity, and give answers to critical business questions.

The performance metric reported by TPC-H is called the TPC-H Composite Query-per-Hour Performance Metric (QphH@Size), and reflects multiple aspects of the capability of the system to process queries. These aspects include the selected database size against which the queries are executed, the query processing power when queries are submitted by a single stream, and the query throughput when queries are submitted by multiple concurrent users. The TPC-H Price/Performance metric is expressed as $/QphH@Size.


TPC-R

The TPC Benchmark™R (TPC-R) is a decision support benchmark similar to TPC-H, but which allows additional optimizations based on advance knowledge of the queries. It consists of a suite of business oriented queries and concurrent data modifications.

The performance metric reported by TPC-R is called the TPC-R Composite Query-per-Hour Performance Metric (QphR@Size), and reflects multiple aspects of the capability of the system to process queries. These aspects include the selected database size against which the queries are executed, the query processing power when queries are submitted by a single stream, and the query throughput when queries are submitted by multiple concurrent users. The TPC-R Price/Performance metric is expressed as $/QphR@Size.



TPC-W

TPC Benchmark™ W (TPC-W) is a transactional web benchmark. The workload is performed in a controlled internet commerce environment that simulates the activities of a business oriented transactional web server. The workload exercises a breadth of system components associated with such environments, which are characterized by:
Multiple on-line browser sessions
Dynamic page generation with database access and update
Consistent web objects
The simultaneous execution of multiple transaction types that span a breadth of complexity
On-line transaction execution modes
Databases consisting of many tables with a wide variety of sizes, attributes, and relationships
Transaction integrity (ACID properties)
Contention on data access and update
The performance metric reported by TPC-W is the number of web interactions processed per second. Multiple web interactions are used to simulate the activity of a retail store, and each interaction is subject to a response time constraint.

TPC-W simulates three different profiles by varying the ratio of browse to buy: primarily shopping (WIPS), browsing (WIPSb) and web-based ordering (WIPSo). The primary metrics are the WIPS rate, the associated price per WIPS ($/WIPS), and the availability date of the priced configuration.



Microsoft has no control or influcence or anything over any results.

compaq, dell, IBM all submit both Unix and Win32 machines for testing, since they are the vendors that are competing against each other.


but, this is more microsoft fud im sure.

[Urko]

This has become tedious and as I am not getting payed ,unlike you Sweede, I will leave fellow LNOer's with this.
The #1 and #2 DB vendors IBM, Oracle both support linux as one of their top three plaforms.

If you have further doubts generated by Sweede 'Elmer' Fudd just go to google and search on TPC and realworld

[Sweede]

If i was getting paid by MS for marketing, i really wouldnt be posting to this message board causs i'd be making like, 80-180k a year out livin it up.

i also have contributed a lot of usefull info/help to people that i can.

[ssadams]

Originally posted by Sweede:
<STRONG>If i was getting paid by MS for marketing, i really wouldnt be posting to this message board causs i'd be making like, 80-180k a year out livin it up.

i also have contributed a lot of usefull info/help to people that i can.</STRONG>

yes you do help people in the http,php,perl and other areas but you too often quote useless benchmarks and generally say "there is a Win option..." if someone cant get something to work.All opinions are valid but you have to remember this is a LINUX bbs and for what ever reasin people are trying different solutions.Have a good day :)

[Sweede]

you lie !
i do not help people with perl... thats absurd.. perl that is ;)

[JasonC]

Sweede, since you seem to checking this thread an not the "If you love MS" one I had a question for you. Here is the link:

If you love Microsoft (http://www.linuxnewbie.org/cgi-bin/ubbcgi/ultimatebb.cgi?ubb=get_topic&f=6&t=003845)

[stiles]

Sweede did you read the fact that this entry wasn't about cost / transaction, but was about showing the viability of a new platform (aka a proof of concept test per say). Yea SGI has like ever sold resonably priced equipment LMAO. Out in the real world shops that run MS run 1.5 to 4 time as many boxes (and have 4 to 10 times as many issues with their backends, is that figured into the TCO for MS products, I think not) as unix shops. MS shops are haveing a hell of a time upgrading from MS SQL 6.5 cause 7.0 and 2000 is a dog. Even though the marketing people claims that the new versions parallelize queries on smp hardware, yet I know of 3 instaliations that have gotten 3 to 10% less performance on SQL 2000 when compaired to SQL 6.5, (all on quad xeons), even the great tech support from MS land hasn't solved these "issues". Why don't "issues" like these show up on TCP test if they are so realworld? :rolleyes:

And has tcp total disclosure documents been so skimpy, I tried to find out what filesystem was used by searching throught the pdf file and got no returns for ext2 or XFS, lame.

What's really funny is that the kernel hackers know that the kernel isn't really ready for primetime when it comes to RDMS (of course you don't know what's in the W2k exectutive so you'll have to take MS' word). The block device layer needs a total rewrite, a better light weight thread implimation, ability to bypass the elavator among other improvments (maybe even a transaction API in the kernel, that idea has been thrown around on the kernel list). The next kernel series will really be the one targeted at running transactions. I bet MS isn't felling too good about that seeing as how they are haveing scalling issues with their last two releases of SQL server.

[ 20 May 2001: Message edited by: stiles ]

[bdg1983]

Originally posted by Sweede:
<STRONG>Urko: that was sad.

if you read my first post, TPC benchmarks are real world results using real world programs in REAL ENVORIMENTS. not a lab setting.

the cost per transaction is the initial investment of the server.
if i spend 10 million on licensing and hardware, but only get 200,000 transactions per minute (the Solaris Box), my cost per transaction is higher than the Windows 2k box that costs 15 million (INCLUDES LICENSES) that does 600,000 transactions per minute (which, btw, is 20$ per transaction).

The systems tested on TPC.org are NOT some techno geek who has no life, but VENDOR systems that they SELL PRECONFIGURED READY TO RUN to companies.


as i said above, the TOTAL cost includes any and ALL licensing fee's.

if TPC says you'll pay $14,312,345.32 for the server, thats how much you'll pay including SunOS, Oracle, DB2, SQL2000, W2k licensing.

a personal freind of mine does his own benchmarking between.


before you try to flame me, read first..</STRONG>


You're a ****ing lame troll. What??? Mummy not giving you enough lovin' and you feel the need to attention seek here on LNO all the time to make up for this lack in your real life?

You know what I love even more than this type of thing? I love it when you get involved in an argument involving OSS as a business model, and then you start quoting all your pseudo-economics and all that other vast knowledge you learn off &lt;insert favorite business show here&gt; and you seem to think that this ARMCHAIR knowledge can replace REAL WORLD education. well quite frankly it cant. Stop trying to be an expert in all areas of ya life, i just doesnt work, and losely quoting "facts" completely out of contect has two effects. The use of "facts" means you dont need to be smart, and secondly, because they are out of context, we cant see for ourselves. i notice you never seem to simply post the link to the relevant stuff, you always post up only the bull**** that supports your ****ed up troll opinion.

Now run along, mummy wants to give you a hug M$ troll *****.

[Sweede]

http://www.tpc.org/information/benchmarks.asp http://www.tpc.org/tpcc/faq.asp http://www.tpc.org/tpch/default.asp#spec


btw, my 'airmchair' economics where taught to me at DeVry by a professor who has a PH.D. in economics, and to pass the class, i wrote a paper about the Open Source idea in the business world and its affects into the Business model, the Invisable Hand, and the Economic Machine.

She found it very very well written and very accurate. to bad i never actually kept a copy of it, but im sure i can get one from her.

btw, a paper like that does require me to research it and provide real examples of open source businesses.

[bdg1983]

My primary school english had a ph.d and she taught me to write really cool creative writing assignments. i particularly like the one i wrote about the little green martians coming down to earth and being our friends. This does not make me a publisher's dream now does it. my dad has got a ph.d for fscks sake.

I have heard in quite a few posts how u reference the fact that you were taught by a ph.d. Mate, i dont think there is a lecturer at my University that doesnt have a ph.d, its a pretty common thing in academic circles.

I just think that maybe you have a BIAS issue. your arguments are usually very polarised and skewed, and it just comes across as trolling and FUD. And you also never seem to be bold enough to draw conclusions from your own statements and arguments.

You may have written some good paper (and if it was that good then why didnt you keep a copy of it????). But in the real world bias like that will just destroy your credibility. People wont/dont take you seriously when you just seem so obviously skewed. The stramge thing is that i cant work out you motivation for being this way? Is it that you dont like linux? if so then why are you here? do you feel the need to cause conflict in the linux community and make us look petty and argumentative when we respond emotionally to your inflamatory remarks?

Quite frankly, i think the person who said you get paid by M$ to be a fud fountain had it closest. If you dont get paid, then that makes you stupid, and basically Microsoft's *****, to be bend over and butt fscked at their will, because you are doing their work for them.

[ 21 May 2001: Message edited by: SlCKB0Y ]

[JasonC]

Sweede, I am having difficulty buying into you being even halfway accurate with economic data. Your numbers in the other thread MS operating systems up 6 billion or 40% are inaccurate. First of all an increase of 6 billion would be a 67% increase and MS overall is only forecast to have 2.5 billion (at best) gain in revenue. If you are this economics guru that you claim to be, how's come you can't read an annual report?

[MkIII_Supra]

I have nothing useful to add other than let the smack don commence! Damn this is pretty heated!

[JasonC]

Well at least he shut up :D