pgsql-odbc doesn't seem a right place to post.
I forwarded my reply to pgsql-hackers also.
Michael Rudolph wrote:
>
> Hi,
>
> I have a performance-problem with my Postgresql-database: up to now I
> had centura sqlbase but for some reasons I have to change to Postgres.
> After copying the data of an existing database I made some tests for
> performance which where very disappointing. There is one table, let's
> say table1, with about 11.0000 rows. On doing a "select distinct var1
> from table1" via ODBC-Driver, I had to wait about 15 sec for the result.
> The same select on centura sqlbase lasted about 3 sec. Ok, I thought, it
> might be the ODBC-Driver with that bad performance and I did that query
> directly with psql. Not much better, it lasted around 10 sec.
>
How many rows are returned by the query in reality ?
Does the table *table1* have an index on the column *var1*
in your centura sqlbase ?
> This isn't what I can give to my users because in my application I have
> to do a lot of queries of that kind. The users would have a lot of time
> to hang around, waiting for the application, and thinking, how to kill
> me ;-) .
>
> I made some more tests with indices and with more indices but the
> improvement is not worth to mention.
>
> I am now at a point, wondering, if the transfer from centura to postgres
> is the right way but I can't imagine why the performance of centura is
> so much better. The conditions are nearly the same (256 MB RAM, Pentium
> III) beside of the OS, which is Novell Netware for Centura and Linux
> (Kernel 2.4.0) for Postgres. The version of Postgres is 7.0.3.
>
> Does anybody have an idea what else I can do? Thank yo very much for
> your help.
>
> Michael