1. imp with indexs=n, this will not create indexes and make the import faster. Then u can do imp whith indexfile=index.sql, this will generate your index creation script which you can run later.
2. Create few monster rollback segments in different tablespace for the import and keep other rollback segments offline.
3. give buffer=large value. (be carefule or it will eat up your memory)
HTH
Regards
Amar
Amar "There is a difference between knowing the path and walking the path."
Thanks all for your imput. I did increased the buffer size. I also changed the parameter file as in oracle note id:#93763.1 unfortunately none of above helped with my problem. I am guessing its resource and contention issue since our upgrade.
Amar, setting index=n is time consumming. I did create large rbs before the import and kept others offline.
MH, can you expanciate more on the followin:
Did you switch optimizer modes?
When was the last time the tables were analyzed?
Indexes fragmented?
Is Pluto aligned with Jupiter?
Did you switch optimizer modes?
When was the last time the tables were analyzed?
Indexes fragmented?
Thanks all,
katie. [/B]
Well, I was just informed that optimizer mode has nothing to do with this problem so please disregrd.
If you are truncating the tables prior to the import then the analyze and index issues are no longer valid.
Stick to the metalink #93763. Aside from that I think there is little you can do.
Bookmarks