On Wed, Oct 23, 2013 at 1:07 PM, Jeff Janes <jeff.janes@gmail.com> wrote:
> How does max not answer "is this query ever really slow?"? But good point,
> if we have a max, then I think a time-stamp for when that max was obtained
> would also be very useful.
I'm concerned about the cost of all of this. And like Stephen, I'm not
too impressed by the idea of a permanent max - it's going to be some
value from before the cache was warmed the large majority of the time.
I think that there are some big savings to be made now that the query
text is only useful to humans, and isn't compared directly for the
purposes of matching and so on. Generally speaking, a human will
inquire about query execution costs far less frequently than the
system spends aggregating them. So fixing that problem would go a long
way towards resolving these concerns. It would also probably have the
benefit of making it possible for query texts to be arbitrarily long -
we'd be storing them in files (with a shared memory buffer). I get a
lot of complaints about the truncation of query texts in
pg_stat_statements, so I think that'd be really valuable. It would
make far higher pg_stat_statements.max values practical to boot, by
radically reducing the amount of shared memory required.
All of this might be a bit tricky, but I suspect it's well worth it.
--
Peter Geoghegan