Greg Smith wrote:
> Greg Williamson wrote:
> > Our tests -- very much oriented at postGIS found Oracle to be between 5
> > and 15% _faster_ depending on the specifics of the task. We decided to go
> > with postgres given the price difference (several hundred thousand dollars for
> > Oracle in the configuration we needed vs. zip for postgres -- we already had
> > trained postgres DBAs).
> >
>
> Can always throw the licensing savings toward larger hardware too; $100K
> buys a pretty big server nowadays. At the FAA's talk about their
> internal deployment of PostgreSQL:
> https://www.postgresqlconference.org/2010/east/talks/faa_airports_gis_and_postgresql
>
> They were reporting that some of their difficult queries were
> dramatically faster on PostgreSQL; I vaguely recall one of them was 100X
> the speed it ran under Oracle Spatial. It was crazy. As always this
> sort of thing is very workload dependent. There are certainly queries
> (such as some of the ones from the TPC-H that big DB vendors optimize
> for) that can be 100X faster on Oracle too.
The FAA reported something like that at PG East about Oracle vs.
Postgres performance with GIS data.
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ It's impossible for everything to be true. +