Re: [GENERAL] Materialized view vs. view
От | Nicolas Paris |
---|---|
Тема | Re: [GENERAL] Materialized view vs. view |
Дата | |
Msg-id | 20170110213008.GC5870@gmail.com обсуждение исходный текст |
Ответ на | Re: [GENERAL] Materialized view vs. view ("David G. Johnston" <david.g.johnston@gmail.com>) |
Список | pgsql-general |
Le 10/33/2017 à 21:33, David G. Johnston écrivait : > On Tue, Jan 10, 2017 at 1:01 PM, Melvin Davidson > <[1]melvin6925@gmail.com> wrote: > > Can we all agree that the "Materialized View" should be faster > > Yes. The OP told about a 500K rows view. Every select queries on that view will have to fetch those 500K rows before - in any case this is quite slow. However, 500K rows do not represent a huge physical space to materialize. But do not forget to index/refresh the materialized views, depending on the select set of queries to run. Answer would have been different with 50K rows I guess. > > If you add in the condition that the same answer has to be returned > whether you query the view or the mat-view then no, I wouldn't agree; > and the original question is, IMO, unanswerable (or at least, if > answered, requires making assumptions that tend to go unstated). > > is just an opinion and too late to change at this point? > > Your opinion, or the name? :) > David J. > > References > > 1. mailto:melvin6925@gmail.com
В списке pgsql-general по дате отправления: