Re: High db file sequential reads during imp?

Many thanks to all for the responses. Of course, I should have posted the
parfile I used:

USERID=**/**
FILE=**.DMP
LOG=**.log
COMMIT=n
FULL=n
FEEDBACK=100000
IGNORE=Y
ROWS=y
grants=n
statistics=none
constraints=y
TOUSER=**
FROMUSER=**

I think the obvious thing I'm not getting - as several people identified -
is the amount of reading on the index which is needed just to carry out
inserts to it and this is clarified by the Lewis article. I've been aware of
the INDEXES=N methood for a while but haven't had to try it yet.

Cheers,
cam

Other related posts: