Re: pg_depend - Mailing list pgsql-hackers

From Bruce Momjian
Subject Re: pg_depend
Date
Msg-id 200107171601.f6HG1M608919@candle.pha.pa.us
Whole thread Raw
In response to Re: pg_depend  (Peter Eisentraut <peter_e@gmx.net>)
Responses Re: pg_depend
List pgsql-hackers
> Bruce Momjian writes:
> 
> > > I'm not seeing the point.  You're essentially duplicating the information
> > > that's already available in the system catalogs.  This is bound to become
> > > a catastrophe the minute a user steps in and does manual surgery on some
> > > catalog.  (And yes, manual surgery should still be possible.)
> >
> > But how then do you find the system table that uses the given oid?
> 
> It's implied by the column you're looking at.

Is it?  Are we going to record dependency both ways, e.g primary table
-> foreign table and foreign table -> primary table, or just one of
them.  And when we see we depend on something, do we know always what it
could be.  If I drop a table and I depend on oid XXX, do I know if that
is a type, function, or serial sequence?

--  Bruce Momjian                        |  http://candle.pha.pa.us pgman@candle.pha.pa.us               |  (610)
853-3000+  If your life is a hard drive,     |  830 Blythe Avenue +  Christ can be your backup.        |  Drexel Hill,
Pennsylvania19026
 


pgsql-hackers by date:

Previous
From: Lamar Owen
Date:
Subject: Re: What I do with PostgreSQL
Next
From: Bruce Momjian
Date:
Subject: Re: Idea: recycle WAL segments, don't delete/recreate 'em