On Monday 02 December 2002 08:21 pm, you wrote:
> While reading the debate, the MySQL benchmarks were brought up. It is
> apparent that we need to contact MySQL (even thought hey claim to have
> contacted us first) and run through a proper benchmarking. We should do a
> variety of tests that mimic what MySQL people get (i.e.. forget the foreign
> key constraints) and them against MySQL in transaction safe and
> non-transaction safe modes, along with PostgreSQL in all her glory.
Is it possible to get something like TPC-C on postgresql upto a limited scale?
May be TPC won't list it but we would ahve a repeatable benchmark to start
with..
Other way around is post numerous benchmarks on live system. We would set some
criteria as in what all should a entry consist of e.g.
* database size
* number of records
* kind of queries
* best and worst performances
* hardware design
* database design
* database tuning
etc. Now if we have say hundreds of these, that would give a pretty good idea
of postgresql abilities. After all, benchmarks are supposed to do the same
thing, isn't it?
Shridhar