[Mondrian] RE: TPC-H for Mondrian

Julian Hyde jhyde at pentaho.org
Sat Dec 22 15:24:02 EST 2007


> Since there are strict rules about reporting TPC-H results of 
> any kind, 
> we should make it clear via disclaimers and naming that this is an 
> unofficial multidimensional adaptation and not the real TPC-H 
> benchmark.

Agreed.

I would also like to avoid acrimonious debate about performance numbers for
mondrian running on various platforms. Our goal is to make mondrian the OLAP
best engine for all operating systems and databases, not to play favorites
or allow people to advertise how wonderful their system is.

So, I propose that people don't publish numbers in this forum; DBMS vendors
are of course welcome to publish numbers on their own forums.

It may be necessary to discuss numbers for development purposes, e.g.
"Running MySQL, I added an index and query A went from 100s down to 30s".
This is acceptable as long as the specifics (hardware platform, memory, disk
specification) are kept to a minimum necessary to have a technical
discussion, and direct comparisons to competing platforms (e.g. Oracle, or
Linux vs. Windows) are not made.

Julian




More information about the Mondrian mailing list