Обсуждение: MySQL disputes benchmarks..

Поиск
Список
Период
Сортировка

MySQL disputes benchmarks..

От
Chris Bitmead
Дата:

Re: MySQL disputes benchmarks..

От
Thomas Lockhart
Дата:
> http://slashdot.org/article.pl?sid=00/08/16/0010230&mode=thread
> http://www.devshed.com/BrainDump/MySQL_Benchmarks/

The thing that's especially funny to me is that Monty is the first one
to cry, but has been making qualitative, unsubstantiated claims about
Postgres' supposed shortcomings for years (OK, he thinks they are
substantiated by *his* single-user benchmarks :). And contrary to his
claims in the past, imho Postgres has been quite silent on the issue of
testing fairness and MySQL attributes, trying to let the product do the
talking instead.

Great Bridge (not "the Postgres people" as he claims) actually went to
the trouble to do specific multi-user stress tests using published
industry-standard benchmarks, which is the closest thing to a fair test
we've ever seen. Everyone was suprised by the results. It will be nice
to see a few more of these kinds of fair tests in the future.

All imho of course...
                     - Thomas


Re: MySQL disputes benchmarks..

От
Kaare Rasmussen
Дата:
> The thing that's especially funny to me is that Monty is the first one

> to cry, but has been making qualitative, unsubstantiated claims about
> Postgres' supposed shortcomings for years (OK, he thinks they are

Remembering "crash-me" ...

> substantiated by *his* single-user benchmarks :). And contrary to his
> claims in the past, imho Postgres has been quite silent on the issue
of
> testing fairness and MySQL attributes, trying to let the product do
the
> talking instead.

Only when somebody comes to a pgsql-* list and starts talking about it,
they get some (well, sometimes heated) argumentation.

I would strongly urge people to let this argumentation go - don't get
involved in a pissing contest. Instead, put the test and any other
relevant documentation on the home page.

> Great Bridge (not "the Postgres people" as he claims) actually went to

> the trouble to do specific multi-user stress tests using published
> industry-standard benchmarks, which is the closest thing to a fair
test
> we've ever seen. Everyone was suprised by the results. It will be nice

> to see a few more of these kinds of fair tests in the future.

At least the test shows that PostgreSQL is fastest under _some_
circumstances. I think it would be far fetched to claim PostgreSQL to be
fastest for all purposes, but that's not what I've been reading.

It would be nice to see PostgreSQL go through a real TPC test, so it
could claim a place on the "most transactions for the buck" list. But I
believe it's very expensive ?

--
Kaare Rasmussen            --Linux, spil,--        Tlf:        3816 2582

Kaki Data                tshirts, merchandize      Fax:        3816 2501

Howitzvej 75               Åben 14.00-18.00        Email: kar@webline.dk

2000 Frederiksberg        Lørdag 11.00-17.00       Web:      www.suse.dk