Are there any triggers, indexes, or other peculiarities to these tables?
Assuming you are not deferring segment creation: How much space is being
pre-allocated to each table? Are the tablespaces already large enough
for the defined extents, or are they being grown incrementally? Does
the storage have journalling and is that impacting the previous two things?
No data does not necessarily mean empty.
/Hans
On 2016-12-08 3:28 PM, Will Beldman wrote:
I am trying to run an impdp job from a database with thousands of tables. I
have ~20000 tables that are empty. I still want to create the tables but
there's no actual data to import.
The output from the impdp job seems to be spending anywhere between 2-5
seconds per table. Times 20000 tables I'm looking at anywhere between 12 and
24 hours to import the entire database.
I could understand large tables, but I would expect fractions of a second for
empty tables. Does this sound normal? Any strategies I could employ to speed
this up?