[unixODBC-support] Problem in using cursor to fetchmillionsofrows
jlam at sandvine.com
Thu Apr 7 18:49:08 BST 2005
In My sql statement, it asks for the id which has the datatype of int. Using isql, same bebavior happens (wait for 2-3 minutes before it starts flushing out the records), ~ 153 M consumed. (see below). The table has 4 integer fields, 4 bigint fields and 1 timestamptz.
last pid: 69042; load averages: 0.74, 0.77, 0.41 up 3+04:17:40 17:47:46
51 processes: 1 running, 50 sleeping
CPU states: 0.0% user, 0.0% nice, 0.3% system, 0.8% interrupt, 98.9% idle
Mem: 178M Active, 338M Inact, 166M Wired, 8K Cache, 112M Buf, 321M Free
Swap: 1056M Total, 13M Used, 1043M Free, 1% Inuse
PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU CPU COMMAND
68985 jlam 2 0 153M 153M poll 2 6:01 28.52% 28.52% isql
68986 pgsql 2 0 17556K 10888K sbwait 0 0:14 3.32% 3.32% postgres
From: unixodbc-support-bounces at easysoft.com
[mailto:unixodbc-support-bounces at easysoft.com]On Behalf Of Nick Gorham
Sent: Thursday, April 07, 2005 1:36 PM
To: Support for the unixODBC project
Subject: Re: [unixODBC-support] Problem in using cursor to
Jerry Lam wrote:
>Yes, the memory usage stay fixed. It released a portion back to the OS when I call "db disconnect" about 80M. It hurts because it used 166M of memory and I'm expected it to use much less than that ~20M since I want 1 row at a time by using the cursor.
And does it still do the same using isql ?
I have just returned 39286 rows in isql from postgres and the memory
usage of isql doesn't go about 31mb. Are there any blobs or long
datatypes in the table ?
unixODBC-support mailing list
unixODBC-support at easysoft.com
More information about the unixODBC-support