Thread: dblink questions

What's the best way to create a connection with dblink with persisted
connections for web applications?

For example:

I want to have a view that links to a table in another pg database so I
can mask the connection and make it easier to query.

so I would do this ...

create view myremote_pg_proc as
select *
from dblink('myconnection','select proname, prosrc from pg_proc')
as t1(proname name, prosrc text);

so I can do this ...

select * from myremote_pg_proc where proname like 'bytea%';

I understand that a connection can be made by dblink_connect(text
connstr) and
dblink_connect(text connname, text connstr), but how do you suggest
making the first connection from a web application?

select dblink('myconnection','127.0.0.1 port=5432 dbname=template1
user=postgres password=mypasswd')

Do you suggestion not using the persistent connection?  and do this?

create view myremote_pg_proc as
select *
from dblink('myconnection','127.0.0.1 port=5432 dbname=template1
user=postgres password=mypasswd','select proname, prosrc from pg_proc')
as t1(proname name, prosrc text);



Thanks a Bunch,

Brian

Brian Maguire wrote:
> What's the best way to create a connection with dblink with persisted
> connections for web applications?
>

It all depends on how you intend to use the dblink connections.

If your web page connects to Postgres, issues a single query using
dblink (via a view or whatever), and then completes, you don't need a
persistent dblink connection -- just use the explicit connection syntax.

If, on the other hand, your web page connects to Postgres, issues
several queries using dblink (via a view or whatever), and then
completes, it would make sense to use a persistent dblink connection. In
that case, execute dblink_connect() before you execute your queries. Use
named persistent connections if you need more than one, anonymous otherwise.

HTH,

Joe