On Mon, 2008-12-01 at 22:51 -0500, justin wrote:
> On idiotic benchmark comparisons "Try to carry 500 people from Los
> Angeles to Tokyo in an F-15. No? Try to win a dogfight in a 747. No?
> But they both fly, so it must be useful to compare them... especially
> on the basis of the most simplistic test case you can think of. For
> extra points, use *only one* test case. Perhaps this paper can be
> described as "comparing an F-15 to a 747 on the basis of required
> runway length
We used that analogy for comparing database benchmarks as far back as
1989-90 at Teradata. My memory is it was invented to counter claims that
DB2 "was faster" after some disastrous initial benchmark results while
attempting a straight database migration. The contrast was that the
client/server overhead of each request *was* higher, though the parallel
database could perform actions much faster when it eventually got
started. The original analogy was a comparison of the passenger carrying
capacity, since a jet fighter could only carry ~1 person while the
airliner could carry 100s, yet the jet fighter could obviously deliver 1
person much faster to a destination. (At the time, the concept of
client/server was widely laughed at). Joke -> Threat -> Obvious.
That thought led to the development at BA of a system specifically
designed to offload large SQL queries from the mainframe DB2 system. BA
knew how to judge database systems and use them for what they were good
at. (They continued to use TPF also, because of its speed of hash index
implementation, amongst other optimisations).
--
Simon Riggs www.2ndQuadrant.com
PostgreSQL Training, Services and Support
