On 09/26/2012 15:03, FFW_Rude wrote:
Here is the answer to Ray Stell who send me the wiki page of Slow Query. I hope i detailed all you wanted (i basicly pasted the page and add my answers). Full Table and Index Schema: schema tables_adresses "Tables" tables_adresses.adresses_XX (id (serial), X(Double precision),Y (Double precision)). "Indexes" adresses_XX_pkey (Primary key, btree) calcul_XX (non unique, Btree on X,Y) schema tables_gps "Tables" tables_gps.gps_XX (id (int),x_max(numeric(10,5)), y_max (numeric(10,5)),x_min(numeric(10,5)),y_min(numeric(10,5))) "Indexes" calculs_XX (non unique Btree x_min,x_max,y_min,y_max) gps_10_pkey (Primary key on id btree) Approximate rows 250000. No large objects in it (just data) No NULL receives a large number of UPDATEs or DELETEs regularly is growing daily I can't post an EXPLAIN ANALYZE because of the 6hour query time. Postgres version: 9.1 History: was this query always slow, : "YES" Hardware: Ubuntu server last version 32bits Daily VACUUM FULL ANALYZE, REINDEX TABLE on all the tables. WAL Configuration: Whats a WAL ? GUC Settings: i didn't change anything. All is standard. shared_buffers should be 10% to 25% of available RAM (it's on 24MB and can't go higher. The server has 4Gb) effective_cache_size should be 75% of available RAM => I don't now what this is.
before looking further, please configure shared_buffers and effective_cache_size properly, it's fundamental you'll probably need to raise SHMALL/SHMMAX, take a look at: http://www.postgresql.org/docs/current/static/kernel-resources.html for 4GB of RAM I start with shared_buffers to 512MB and effective_cache_size to 2GB
Test changing work_mem: increase it to 8MB, 32MB, 256MB, 1GB. Does it make a difference? "No"
default work_mem is very small, set it to something like 16MB
-- View this message in context: http://postgresql.1045698.n5.nabble.com/Same-query-doing-slow-then-quick-tp5725486p5725491.html Sent from the PostgreSQL - performance mailing list archive at Nabble.com.
-- No trees were killed in the creation of this message. However, many electrons were terribly inconvenienced.
begin:vcard fn:Julien Cigar n:Cigar;Julien org;quoted-printable:Belgian Biodiversity Platform;ULB (Universit=C3=A9 Libre de Bruxelles) adr:;;;Brussels;;;Belgium email;internet:jcigar@xxxxxxxxx tel;work:+32(0)26505752 x-mozilla-html:FALSE url:http://www.biodiversity.be version:2.1 end:vcard
-- Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-performance
