the inputs first, to maximize cache hit/miss ratio, such as
"CREATE INDEX extension_idx ON file (reverse(name)) ORDER BY name" ?
On Mon, 26 Aug 2002, Josh Berkus wrote:
> Jeff, h012,
>
> > CREATE FUNCTION fn_strrev(text) returns text as '
> > return revers
Hi,
CREATE TABLE file (name varchar(255));
I have a couple of milion filenames. I need to build index based on
extensions. A couple of possibilities come to mind:
CREATE INDEX extension_idx ON file (reverse(name));
-- but I didn't find a function called "reverse"
CREATE INDEX extension_
I've even launched the backend with "-F" and removed BEGIN/COMMIT and
LOCK TABLE and FOR UPDATE, but I still get slow response.
only when count(*) from file is 16000, I get about 2-3 rows / second on
average. When count(*) from file was 100, I get about 20-30 rows / second.
Help !
T
and I forgot to mention that my stats are available at:
http://John.Vicherek.com/slow/times.query.txt
John
On Sun, 11 Aug 2002 [EMAIL PROTECTED] wrote:
>
>
> Hi,
>
> I must be doing something silly. I have a 900MHz, 384MB RAM, and
> this thing is slow. (Postgresql-7.1.2).
>
> And
Hi,
I must be doing something silly. I have a 900MHz, 384MB RAM, and
this thing is slow. (Postgresql-7.1.2).
And growing exponencially slower.
SQL: http://John.Vicherek.com/slow/schema.sql (save to /tmp/schema.sql)
am.dat: http://John.Vicherek.com/slow/am.dat (save to /tmp/am.dat )
pe
Hi,
I realize that a relational database may not be ideal for storing (and
retrieving) tree-like strucutres, but it looks like you guys are doing
with PostgreSQL the impossible anyway.
Having table t of all nodes:
CREATE SEQUENCE nodeIDseq START 1;
CREATE TABLE t(
id int PRIMARY
Great !
I'll try patching it.
Thanks Oleg & Teodor for doing all this great work !
Long live PostgreSQL !
see yea,
John
On Fri, 9 Aug 2002, Oleg Bartunov wrote:
> Hmm,
>
> you dont' need to use GiST supporting functions !
> We've posted a patch to current CVS, it has everyt
thanks for the pointer, I'm now able to use the operators in expressions!
w.r.t. modifying the array contents: I looked through
/usr/lib/pgsql/contrib/intarray/_int.sql.in , but it didn't make me any
smarter. Are the "..._union" and "..._picksplit" functions supposed to add
into and remove
Hi,
I would like to say:
create table test1 (array1 int4[]);
insert into test1 values ('{123,234,345}');
insert into test1 values ('{456,567,678}');
now what I don't know how to do:
-- consider contents of array:
select * from test1 where array1 CONTAINS 567;
-- APPEND '789' to array in
