Re: Query tuning: partitioning, DISTINCT ON, and indexing

On Thu, Jun 20, 2013 at 9:13 PM, bricklen <bricklen@gmail.com> wrote:

On Thu, Jun 20, 2013 at 6:24 PM, Maciek Sakrejda <m.sakrejda@gmail.com> wrote:

SELECT

  DISTINCT ON (type) ts, type, details

FROM

  observations

WHERE

  subject = '...'

ORDER BY

  type, ts DESC;

First thing: What is your "work_mem" set to, and how much RAM is in the machine? If you look at the plan, you'll immediately notice the "external merge Disk" line where it spills to disk on the sort. Try setting your work_mem to 120MB or so (depending on how much RAM you have, # concurrent sessions, complexity of queries etc)

Good call, thanks, although the in-mem quicksort is not much faster:

                                                                                  QUERY PLAN                                                                                   

-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 Unique  (cost=471248.30..489392.67 rows=3 width=47) (actual time=32002.133..32817.474 rows=3 loops=1)

   Buffers: shared read=30264

   ->  Sort  (cost=471248.30..480320.48 rows=3628873 width=47) (actual time=32002.128..32455.950 rows=3628803 loops=1)

         Sort Key: public.observations.type, public.observations.ts

         Sort Method: quicksort  Memory: 381805kB

         Buffers: shared read=30264

         ->  Result  (cost=0.00..75862.81 rows=3628873 width=47) (actual time=0.026..1323.317 rows=3628803 loops=1)

               Buffers: shared read=30264

               ->  Append  (cost=0.00..75862.81 rows=3628873 width=47) (actual time=0.026..978.477 rows=3628803 loops=1)

                     Buffers: shared read=30264

...

the machine is not nailed down, but I think I'd need to find a way to drastically improve the plan to keep this in Postgres. The alternative is probably caching the results somewhere else: for any given subject, I only need the latest observation of each type 99.9+% of the time.