Hi Joel,
It's possible that on the source there were no statistics on temporary
tables because the volume can change everyday and they probably relied on
dynamic sampling to compute dynamically accurate statistics. By having
statistics calulated now for these table you won't be able to get a good
plan for each different workload.
Concerning the fact that you have statistics now on your temporary tables,
are you sure they were calculated during the import?
Le 3 déc. 2015 23:20, "Joel Slowik" <jslowik@xxxxxxxxx> a écrit :
Hi Folks,
I'm not quite sure what to make of this.
I'm migrating a handful of our databases from windows servers (2003) to
RHEL 7 x64 servers. These are the basic steps I'm performing:
1) create database (using that very command, scripted out)
2) run catalog, catproc, utlrp, and then finally run catqm
3) run a script to setup our users and data schema
4) expdp full from target database (database is Oracle 64bit 11.2.0.3)
5) impdp schemas (just one) to source database (database is Oracle 64bit
11.2.0.4)
Somehow along the way (and this is not happening in all of our database),
statistics are being gathered during import for our GTTs and set precisely
to 2000 num_rows, 2000 sample size, 100 average row length, and 100 blocks.
Statistics do not exists for our GTTs in our source database.
What's going on here?
-joel slowik
________________________________
Confidentiality Note: This electronic message transmission is intended
only for the person or entity to which it is addressed and may contain
information that is privileged, confidential or otherwise protected from
disclosure. If you have received this transmission, but are not the
intended recipient, you are hereby notified that any disclosure, copying,
distribution or use of the contents of this information is strictly
prohibited. If you have received this e-mail in error, please contact
Continuum Performance Systems at {203.245.5000} and delete and destroy the
original message and all copies.
--
//www.freelists.org/webpage/oracle-l
