I have a table with 3,000,000 of rows, I have the following indexs:
create index regist_idx on regist(pname,sname,plastname,slastname);
create index regist_idx1 on regist(pname,plastname,slastname);
create index regist_idx2 on regist(ci);
create index regist_idx3 on regist(plastname);
create index regist_idx4 on regist(pname);
create index regist_idx5 on regist(pname,plastname);
create index regist_idx6 on regist(pname,slastname);
create index regist_idx7 on regist(sname,plastname,slastname);
create index regist_idx8 on regist(sname,slastname);
create index regist_idx9 on regist(sname);
create index regist_idx10 on regist(slastname);
if the query is on 4 columns of the idx1, the result is speed, but if teh
query is on two columns of the idx1 the rsult take 10 minutes.
My postmaster run with option -S -o -F for best performance, but the
results is very slow.
Teh machine have: 256 MB memory, SWAP 512, System Operate: Red Hat Linux
6.2, posgresql version 6.5.3 the table struct is:
table regist(tipo varchar(1),cod_barr integer,ci varchar(13),jrv integer
not null,plastname varchar(15),slastname varchar(15),pname
varchar(15),sname varchar(15),lug_nac smallint,fec_nac varchar(10))
I already use vacuum analize. I ran explain with the select and the result
is:
Index Scan using regist_idx5 on regist (cost=25265.64 rows=6 width=94)
How can I have best performance?, Thanks in advance
Regards,
_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.
Share information about yourself, create your own public profile at
http://profiles.msn.com.
