I agree with Rajesh Arora if you can use hints that will be good
enough. Also, if you can effort you can make a table in your local
database, so that the performance will increase, as if you are
interacting the table from remote database, the performance will be
down. So if you can effort to do, create a table in local database
with using COPY command, as if the table will be large and if in
chance you used CREATE TABLE...AS SELECT the Rollback segment will
be the problem too as it will grow more and more.. so better to use
the COPY command

COPY FROM SUNIL/SUNIL@REMOTE
TO SUNIL/SUNIL@LOCAL
CREATE <<TABLENAME>> USING SELECT *FROM <<REMOTE DATABASE TABLENAME>>

also remember to use "SET ARRAYSIZE 10000", so total 10000 will
retrieve in each batch.


---
In OraTechSupportGroup@yahoogroups.com, "dayananda_in"
<dayananda_in@y...> wrote:
>
>
> Hi,
> my name is dayananda shenoy, i am working as an vb/oracle
developer.
> currently we are developing a large scale centralised banking
> software in vb and oracle.
> our application has to handle many independent oracle databases.
> currently we are using oracle dblinks for accessing remote data.
> we are observing that cpu utilisation goes to maximum when we
query
> against dblinks.
> is there any parameters or efficiant way of fetching data from
> distributed databases
>
> thanks in advace
>
> dayananda shenoy.k