From: | "Marko Kreen" <markokr(at)gmail(dot)com> |
---|---|
To: | "f(dot)zamboni(at)mastertraining" <f(dot)zamboni(at)mastertraining(dot)it> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: accessing multiple databases using dblink |
Date: | 2007-12-13 13:31:24 |
Message-ID: | e51f66da0712130531v1b02341bk70d6396802d71309@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 12/13/07, f(dot)zamboni(at)mastertraining <f(dot)zamboni(at)mastertraining(dot)it> wrote:
> Good morning to everybody,
> I've to resolve this situation: I've a collection of many different databases, all identical, and the name of those databases is stored inside a table in another "central
> management" database.
> In an ideal world, I'd like with a single query to be able to recover the name of each database from the table that does contains them, and use it to do a query on each table on
> each databse joining the results.
> At the moment I'm doing it by external code, but performaces are awfull...
> At first I tough something similar to
> select * from tab_databases,dblink('dbname=' || tab_databases.name,'select count(id) from tab_data')
> could work, but it seems its not allowed to reference rows from another table inside the from level.
>
> Putting dblink outside that level allow me to use the dynamic dbname, but in that case I would need a row resource, while dblink give back a table resource...
>
> Is it possible to resolve this inside the databse?
> Thanks a lot...
This seems to be appropriate task for pl/proxy:
https://developer.skype.com/SkypeGarage/DbProjects/PlProxy
--
marko
From | Date | Subject | |
---|---|---|---|
Next Message | Bill Moran | 2007-12-13 13:37:39 | Re: Need LIMIT and ORDER BY for UPDATE |
Previous Message | Alvaro Herrera | 2007-12-13 13:31:19 | Re: [GENERAL] Slow PITR restore |