I've been workinng on Teradata with the mainframes all through my experience. As per my current project, I had to use Datastage as ETL and I have a question here.
How do we collect statistics after an MLOAD or a Fastload when we are using ETL. Do we call the UNIX scripts to do this??
Just submit a COLLECT STATS in MLoad between END MLOAD and LOGOFF.
Before BEGIN and after END MLOAD you can submit almost any SQL, see "Teradata SQL Statements" in the MLoad manual.
In Fastload you need to call a BTEQ script.
In case you do not deal with mload scripts but call mload via the api I would expect that datastage is offering native SQL api as well where you can call collect stats commands.
I don't know data stage too well but all other ETL tools I saw so far offer these.
Many times collect stats are not required on daily basis .If you data doesn't change drastically,you can also consider having a separate fortnightly job collecting stats.
You can also collect stats by calling the procedure in the scripts which would also require an one time entry of the DB and tablename into the procedure view name.
In TeradataStages under Input tab Before and After tabs are there. You can specify Collect Statistics there. Depending on your requirment you can specify before mload or after mload.