RE: Temp space in 10g and data insert speed
- From: "Hollis, Les" <Les.Hollis@xxxxxx>
- To: <ineyman@xxxxxxxxxxxxxx>, <RROGERS@xxxxxxxxxxxxx>, <oracle-l@xxxxxxxxxxxxx>
- Date: Tue, 11 Jan 2005 09:11:02 -0600
Sort_area_size and _retained increase would definitely help.
Just set sort area size in your session before running your inserts.
Set it to something in the magnitude of 30M to 50m....since you are
already constrained by only having 764M, much larger would probably send
you into a panic.
-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Igor Neyman
Sent: Tuesday, January 11, 2005 8:24 AM
To: RROGERS@xxxxxxxxxxxxx; oracle-l@xxxxxxxxxxxxx
Subject: RE: Temp space in 10g and data insert speed
Why do you think tempfile should be increasing? May be it's big enough?
To check usage of tempfile look into V$TEMPSTAT.
Also, increasing sort_area_size and sort_area_retained size might help
with the performance.
Igor Neyman, OCP DBA
ineyman@xxxxxxxxxxxxxx
-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Ron Rogers
Sent: Tuesday, January 11, 2005 8:18 AM
To: oracle-l@xxxxxxxxxxxxx
Subject: Temp space in 10g and data insert speed
List,
OS Redhat AS 3.0 u3 memory 750 Meg swap 1.5 Gig
Oracle 10.1.0.3
I have created the database and included the temporary temp tablespace.
The sga is created using the sga_target and sga_max parameters using
685 Meg
and there still is free memory via the TOP command. The swap space does
not show as being used.
The tables have been created along with the indexes and I am now
starting to insert the data.
My question. In the past I have had to create large temp tablespaces to
handle the inserts with indexes but on this database I see no increase
in the size of the temp tablespace. The manuals describe the temp
tablespace as being used for the "user" created objects and sorts
created by the "users".
The inserts are taking a long time ( hours) to insert 4 million rows.
What could be the possible option that I have missed in creating the
database that
would allow for inserts and index creation to be faster. I know that I
should drop the indexes during dataloads but using my play/test box with
9i is a lot faster inserting data into the same table.
=20
Listed is my init.ora---
# create to replace the spfilelnxdb.ora REDHATDB DATABASE
########################################################################
#######
# database identification
db_name=3Dredhatdb
db_domain=3Dglc
instance_name=3Dlinuxdb
service_names=3Dredhatdb.glc
# database files
control_files=3D("/data/control01.ctl","/u01/data/control02.ctl","/u02/da=
t
a/contro
l03.ctl")
open_cursors=3D700
db_block_size=3D8192
# db_cache_size=3D130023424
# database memory usage
sga_max_size =3D 685500000
sga_target =3D 685500000
#shared_pool_size=3D184549376
# large_pool_size=3D16777216
# java_pool_size=3D8388608
sort_area_size=3D65636
sort_area_retained_size=3D65636
# misc database parameters
processes=3D200
# log_buffer =3D 163840
remote_login_passwordfile=3DEXCLUSIVE
os_authent_prefix=3D""
compatible=3D10.1.0.2.0
fast_start_mttr_target=3D300
# audit_trail =3D false # if you want auditing
timed_statistics =3D true # if you want timed stastics
# archive information
log_checkpoint_timeout=3D1800
log_archive_dest_1=3D'LOCATION=3D/u02/archlogs'
log_archive_format=3Darch_%t_%s_%r.arc
# define directories to store trace and alert files
background_dump_dest=3D/oratrace/bdump
core_dump_dest=3D/oratrace/cdump
user_dump_dest=3D/oratrace/udump
max_dump_file_size=3D10000
utl_file_dir =3D /DATA
utl_file_dir =3D /DATA/DATA
utl_file_dir =3D /DATA/LOG
utl_file_dir =3D /DATA/PAR
utl_file_dir =3D /oratrace/bdump
NLS_DATE_FORMAT =3D "MM-DD-YYYY"
replication_dependency_tracking =3D FALSE
# db_file_multiblock_read_count=3D16
# job_queue_processes=3D10
#query_rewrite_enabled=3DFALSE
# star_transformation_enabled=3DFALSE
# sga_target=3D203423744
# dispatchers=3D"(PROTOCOL=3DTCP) (SERVICE=3DlinuxdbXDB)"
# pga_aggregate_target=3D67108864
# system managed undo and rollback segments
undo_management=3DAUTO
undo_retention=3D600
undo_tablespace=3DUNDOTBS
# asynch i/o configuration
disk_asynch_io =3D true
tape_asynch_io =3D true
filesystemio_options=3Dasynch
Thanks for the input as I continue to search the doc's and metalink.
Ron=20
--
//www.freelists.org/webpage/oracle-l
--
//www.freelists.org/webpage/oracle-l
--
//www.freelists.org/webpage/oracle-l
Other related posts:
