Re: pgsql: Expose the estimation of number of changed tuples since last ana

On Fri, Jul 5, 2013 at 4:23 PM, Cédric Villemain <cedric(at)2ndquadrant(dot)com> wrote:
> Le vendredi 5 juillet 2013 15:44:55, Magnus Hagander a écrit :
>> On Fri, Jul 5, 2013 at 3:43 PM, Cédric Villemain <cedric(at)2ndquadrant(dot)com>
> wrote:
>> >> Expose the estimation of number of changed tuples since last analyze
>> >>
>> >> This value, now pg_stat_all_tables.n_mod_since_analyze, was already
>> >> tracked and used by autovacuum, but not exposed to the user.
>> >
>> > I though you agreed on a shorter name ? (n_mod_tuple ?)
>>
>> No, I suggested a shorter name, but was convinced that this would
>> likely be more confusing..
>
> I understand (I read the review thread). is it correct to interpret that as
> the number of tuples bloating the table ?

No. That would n_dead_tup.

It's the number of tuples that changes the distribution of values and
would thus require ANALYZE. not VACUUM.

--
Magnus Hagander
Me: http://www.hagander.net/
Work: http://www.redpill-linpro.com/