[Monetdb-developers] JDBC/MonetDB performance test

Fabian Groffen Fabian.Groffen at cwi.nl
Tue Dec 6 16:03:54 CET 2005

On 14-10-2005 23:43:33 +0200, Fabian Groffen wrote:
> The results follow in the table below:
>                  PG8       My4       M4s     M4p       M5s     M5p
>    JDBC local
>    - DBTestI     37s       27s       90s     86s       62s     60s
>    - DBTest      19s       27s       60s     59s       46s     46s
>    - DBTestP     14s       27s       63s     64s       59s     60s
>    JDBC remote
>    - DBTestI     49s       37s       84s     85s       62s     62s
>    - DBTest      19s       37s       57s     58s       43s     43s
>    - DBTestP     14s       37s       66s     62s       53s     53s
>    JDBC remote [2]
>    - DBTestI    197s      168s         >1201s            >1545s
>    - DBTest      67s      168s       76s     77s       61s     63s
>    - DBTestP     68s      168s       75s     77s       66s     70s

We have gone through some extensive optimalisations on all levels of the
database, in order to improve the performance.  I conducted the same
experiments as before, excluding the remote over an SSH tunnel and the
tests with the 'progressive' JDBC driver.  The results can be seen

    JDBC local
    - DBTestI     37s       27s       49s     xxs       39s     xxs
    - DBTest      19s       27s       27s     xxs       19s     xxs
    - DBTestP     14s       27s       28s     xxs       23s     xxs
    JDBC remote
    - DBTestI     49s       37s       56s     xxs       49s     xxs
    - DBTest      19s       37s       28s     xxs       21s     xxs
    - DBTestP     14s       37s       29s     xxs       25s     xxs

A new round of optimisations, focussed on minimising the communication
overhead, is about to start, in which we hope to be able to get at least
a performance level between PostgreSQL and MySQL.  On a concluding note,
I see a 40-50% improvement on the previous numbers and the current
numbers, so I guess in total some major steps have been made.

More information about the developers-list mailing list