data pump impdp is SLOOOOOOW [message #463122] |
Tue, 29 June 2010 14:12 |
lawadm1
Messages: 3 Registered: June 2010
|
Junior Member |
|
|
I'm hoping that someone can help me with this. I've tried using data pump (imdp) rather than imp to see if import time speeds up. I hear it's suppose to anyway.
It has taken almost 10 hours to import the data using data pump, and it usually takes 4 hours the old way with imp.
Below is the command that I'm using, and hopefully someone will see something that stands out where I have something incorrect. I thought using the CONTENT=DATA_ONLY would also help speed things up.
impdp lawdev/lawdev90 schemas=lawprd REMAP_TABLESPACE=LAWPRD:LAWDEV REMAP_SCHEMA=LAWPRD:LAWDEV DIRECTORY=IMPORT DUMPFILE=gmkprd2_exp%U.dmp CONTENT=DATA_ONLY EXCLUDE=CONSTRAINT,REF_CONSTRAINT,INDEX TABLE_EXISTS_ACTION=TRUNCATE logfile=LAWDEV_impdp.log
Thanks,
Jeff
|
|
|
|
Re: data pump impdp is SLOOOOOOW [message #463131 is a reply to message #463125] |
Tue, 29 June 2010 15:05 |
|
Michel Cadot
Messages: 68728 Registered: March 2007 Location: Saint-Maur, France, https...
|
Senior Member Account Moderator |
|
|
You can use PARALLEL clause if your hardware, OS workload and configurationallow it.
You have to know that TRUNCATE actions can take long (above all if you have a large buffer cache) and this is not include in former imp so you cannot directly compare.
Regards
Michel
[Updated on: Tue, 29 June 2010 15:07] Report message to a moderator
|
|
|
Re: data pump impdp is SLOOOOOOW [message #463137 is a reply to message #463131] |
Tue, 29 June 2010 15:10 |
lawadm1
Messages: 3 Registered: June 2010
|
Junior Member |
|
|
Do you have an example of how I would use that clause? We're on an HP Integrity server running 11iv3, and Oracle 10gr2 (64 bit). This is a test system, so the OS workload is minimal, and usually things run pretty darn fast on this server.
|
|
|
|
Re: data pump impdp is SLOOOOOOW [message #463140 is a reply to message #463138] |
Tue, 29 June 2010 15:23 |
lawadm1
Messages: 3 Registered: June 2010
|
Junior Member |
|
|
Once I mentioned this to the DBA, he said that he forgot about the PARALLEL parameter. Then he said that we actually use that parameter (PARALLEL=4) on the export, which is probably why we have 4 different dump files. And then here I am using the default of 1 on the import. I'll try this on the next test refresh of our production data, and hopefully we'll see a significant difference.
Thanks for the help.
Jeff
|
|
|
|
|