Re: ** FOR IMMEDIATE RELEASE - RECORD-BREAKING TPC-C RECORD!! **

In article <Pine.SOL.3.91.951101110402.4162A-100000@swindon>, Akmal B Chaudhri <akmal@sarc.city.ac.uk> wrote: > >As I've said before, you should now run a _real_ OLTP benchmark. > >Further details from: > > http://www.cs.city.ac.uk/homes/akmal/citybm.html >or > http://www.city.ac.uk/~ad533/citybm.html Yes, you've said this before. However, I've looked at the web pages you reference and the same bald assertions are there with no substantiation. By claiming the TPC benchmarks do not represent customer workloads you are making a very broad, sweeping statement. Are you saying that 1, The TPC benchmarks do not do the same sort of SQL as customer workloads? 2, They do the same sort of SQL but in different proportions? 3, All benchmarks, including TPC-C, TPC-D, and TPC-E are useless predictors? Or just one or two of these? Pat