On 15 Feb 2011, at 9:32, Alessandro Candini wrote: >> Is that a single query on that one DB compared to 4 queries on 4 DB's? How >> does a single DB with 4 parallel queries perform? I'd expect that to win >> from 4 DB's, due to the overhead those extra DB instances are generating. > > Maybe my configuration and test is not clear
It is clear. I gave you another suggestion for something to try instead. > Splitted instances: > 600 millions of records in total splitted into 4 postgresql instances (port > 5433, 5434, 5435, 5436), let's say more or less: > > 5433 ---> 150 millions of records > 5434 ---> 150 millions of records > 5435 ---> 150 millions of records > 5436 ---> 150 millions of records Try the above on a single DB using 4 threads. It will very probably perform better. To use your example: 5432 ---> 150 million records 5432 ---> 150 million records 5432 ---> 150 million records 5432 ---> 150 million records >> Did you read all the way to section 35.9.10? That explains how to create >> SRF's like yours, including examples. If the step from simple functions to >> SRF's is too large for you, create a few sample-functions to learn how the >> intermediary steps work. We can't upload the info to your brain, after all - >> you'll have to do the learning part by yourself. >> > Do you think is a good idea continue using libpq or should I abandon them and > focus on postgresql documentation examples? To learn how to write functions like these? Best to use a simple case so it's clear what's going on. If you try to alter already complicated code for that purpose you'll probably just get more confused. Alban Hertroys -- Screwing up is an excellent way to attach something to the ceiling. !DSPAM:737,4d5ac6b711731056715104! -- Sent via pgsql-general mailing list (pgsql-general@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
