Re: unusual performance for vac following 8.2 upgrade

[Kim <kim@xxxxxxxxxx>]

Hey Tom,
We were running on 8.1.1 previous to upgrading to 8.2, and yes, we definitely have a heafty pg_class. The inheritance model is heavily used in our schema (the results of the group by you wanted to see are down below).  However, no significant problems were seen with vacs while we were on 8.1. Execution time for the vac seemed more linked to large table size and how active the table was with updates, rather than being universally over 10 minutes regardless of the vac's object. We will be doing an audit of our 8.2 install to try and make sure that it looks like a complete install, any tests you can think of that may further narrow things down for us?



 relkind | count
---------+--------
 v       |   1740
 t       |  49986
 c       |      4
 S       |     57
 r       | 108689
 i       | 374723
(6 rows)



Tom Lane wrote:

Kim <kim@xxxxxxxxxx> writes:
  

size of pg_class: 535,226
number of relations: 108,694
    

Oh, I shoulda read all the way to the bottom of your email :-(.  What
version of PG were you running before?  I would think that pretty much
any version of pgstat_vacuum_tabstats would have had a performance issue
with pg_class that large.  Also, could we see

	select relkind, count(*) from pg_class group by relkind;

			regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?

               http://archives.postgresql.org