Greg Stark <gsstark@mit.edu> writes:
> On Mon, Dec 21, 2009 at 5:02 PM, Greg Smith <greg@2ndquadrant.com> wrote:
>> Right; that's exactly the way I'm computing things now, I just have to crawl
>> way too much catalog data to do it. �I also agree that if we provide
>> pg_table_size, the issue of "pg_relation_size doesn't do what I want" goes
>> away without needing to even change the existing documentation--people don't
>> come to that section looking for "relation", they're looking for "table".
>>
>> Bernd, there's a basic spec if you have time to work on this.
> What about, the visibility maps and free space maps?
Those would be included for each relation, I should think. The
objective here is not to break things down even more finely than
pg_relation_size does, but to aggregate into terms that are meaningful
to the user --- which is to say, "the table" and "its indexes".
Anything you can't get rid of by dropping indexes/constraints is
part of "the table" at this level of detail.
regards, tom lane