Carlos Henrique Reimer <carlos.reimer@opendb.com.br> wrote:=0A=0A> Anyway i=
t does not seam related to statistics as the query plan=0A> is exactly the =
same for both scenarios, morning and evening:=0A=0A> Morning:=0A=0A>=A0=A0 =
Index Scan using pagpk_aux_mes, pagpk_aux_mes, pk_cadpag,=0A>=A0=A0=A0=A0 p=
k_cadpag, pk_cadpag, pagchavefunc00 on cadpag=0A=0A> Evening:=0A=0A>=A0=A0 =
Seq Scan on cadpag=0A=0AThose don't look exactly the same to me.=0A=0A> I'v=
e used this query just as an example but the general feeling=0A> is that ev=
erything takes more time to process in the evening.=0A> Evening is also the=
period with more tasks in the the database.=0A=0AWell, if the resources ar=
e split among more processes, each process=0Awill take longer to run.=A0 No=
surprise there.=A0 Of course, there=0Acould be some interaction -- queries=
which do more work because=0Astatistics are out of date could contribute t=
o the slowdown of=0Aeverything else.=0A=0A-- =0AKevin Grittner=0AEnterprise=
DB: http://www.enterprisedb.com=0AThe Enterprise PostgreSQL Company
