Re: Oracle TPC-A Benchmarks

In article <248859$9jr_at_infochi.com> rfinkel_at_infochi.com (Richard Finkelstein) writes:
>It is very disappointing to see Oracle revert back to its old marketing
>and sales practices. Oracle's use of the TPC-A benchmark and its discrete
>transaction in its advertisements goes to the heart of the Standish Report
>and Oracle's subsequent lawsuit against Standish. The issue is whether
>Oracle is being deceptive in its presentation of the facts.
>
>In my own experiences, and based upon interviews with hundreds of Oracle
>users and users who have benchmarked Oracle, Oracle7 would be hard pressed
>to achieved 1/100th of these stated figures (10,000 concurrent users,
>1,000 TPS) on two low-end Unix computers running in a production
>environment. It is too bad that Oracle finds itself in the position again
>that it has to start running these types of ads again.
>
>Thanks for your reply.
>
>-- Rich
>--

Rich - I am disgusted but not surprised to see your post on this. First you bait the crowd with a nice innocent "What do you think of this new Oracle ad?" posting. Then you wait for the first response and post what amounts to a series of propaganda pieces putting forth your same old tired opinion you hashed out last month about Oracle and TPC-A's and discrete transactions.

Enough already. Why is bashing Oracle your favorite pastime ? Of course you're smart enough to realize that NO real applications can get the performance of ANY vendor's results in the TPC-A benchmark. That's why there doing the TPC-C,D,E, etc. Of course this was already fully discussed on the net last month. This problem is not unique to Oracle, it is the nature of the TPC-A. Of course, now we can go through the whole discrete transaction argument again, but I don't think anyone is interested.

Of course you neglect "The good stuff" about this benchmark. How much do you think the 5 year cost of ownership of 3090's that could perform the equivalent TPC-A rates would be - $50 Million ? More ?

The point is 1) The new Pyramid and Sequent clusters really work, and give large shops an option besides MVS. 2) Oracle runs on them, and runs pretty damn well. 3) Lots of smart unix vendors have noticed this, and are developing their own clusters to take advantage of these capabilities of Oracle. 4) Your own pet database, Sybase, cannot do this. (By the way, I think Sybase is GREAT for lots of stuff, and very innovative, but not on the high end -IMHO) 5) Oracle grew by something like $400 Million in revenue last year; doesn't sound to me like they're in a desperate financial/technological position that you imply above.

Rich - You're a smart guy. Please add value with your posts, and don't post a question simply to generate propaganda and flames and another round of TPC-A/discrete transaction wars. Your posts on the net are becoming more and more inflamatory, and contain less and less valuable new information.

• Dan

Daniel Druker
Anderson Graduate School of Management at UCLA                    

| Dan Druker                                                               |
| agsm mail 	: ddruker                                                  |
| internet 	: ddruker_at_agsm.ucla.edu                                    |
| oracle*mail	: unix:ddruker_at_agsm.ucla.edu                               |
----------------------------------------------------------------------------

Disclaimer: None. I'm a student now and I don't care what you think. Received on Wed Aug 11 1993 - 17:55:05 CEST