We take a copy of our production database and make a copy in our non-prod server once a month for QC and make another copy once a quarter for testing. Thus, we are taking a copy of the production database from tape and copying it to two different databases on the same server. What I've noticed is that when this is done, expression statistics in each of the two non-prod databases end up sharing common names between the two databases. When I try to recollect statistics for these expression statistics, the collect process fails because the express stat name is not unique.
The only options I've identified is: A) Drop the expression stat and then recollect (which can take a long time) or B) Spin off the Statistic Values for the expression stat, edit the result changing the name, drop the original stat and then reload from the stat values. This is faster but can still take a while.
1) Is there a better way to approach this so that we don't have name collisions with our expression statistics?
2) Any possibility of getting "Rename statistics" functionality to simplify the remedy?
3) It seems a little odd that teradata statistics names appear to be expected to be unique, but there is nothing that enforces this expectation. Wouldn't it make sense to have statistic names be required to be unique at the table level rather than at the database level (or is it at the system level)?