Hi Paula
The gotcha if you don't load
with any indexes and constraints is that you need a lot of temp
tablespace to build them but experience says that it will be a lot
faster overall to apply constraints and build indexes later
Just import the data. If you can grab an
export norows from the old database then you can use the ddlwizard to create
scripts to build indexes and constraints or any other schema objects for
your task it is a pretty cool assistant.
Cheers
Peter
Using Oracle 8.1.7, Solaris 2.9
Trying to load a large amount of data using import.
The schema needs to be the same and only the data reloaded.
There are triggers, constraints (pk, fk, etc.) and pk indexes.
I have been loading with pk's enabled and pk indexes (not fk and fk
indexes) and the process is very slow.
I have been running the import with statistics=n, indexes=n
I am tempted to disable all constraints and drop pk indexes but I am afraid
that I might have trouble creating the pk's properly. Can I used the
indexfile option to create the pk's properly and will this make a significant
difference in the import.
What about after insert triggers? What is the best way to handle the
after insert triggers?
Any help would be appreciated. I have been spending a lot of time
with this data load.
Thanks,
Paula
Groups are talking. We´re listening. Check out the
handy
changes to Yahoo! Groups.
