Teradata dump - Prod to dev


Teradata dump - Prod to dev

Hi All,

We are trying to setup a teradata server for our dev environment and we would like to replicate the production environment but with only sample data(limited no of records in each table). Needless to say, our dev env has less space compared to teh production.

What is the easiest/Industry standard way of doing this?

Also, is there any way we can parameterize the no of records from tables in particular database (Ideally, we would like to copy all the data from the process control/metadata tables but only samples from the real data tables)

I have heard of the arcmain utility but i am not sure if we can get samples of data from each table in the arcmain dump.

Appreciate any help on this!


Re: Teradata dump - Prod to dev

You can't restore sample data using arcmain, but you could restore the whole table then create a table using sample data

create the table first or use create table as syntax..

select * from databasename.tablename sample 100;

Then drop the full table. Start with the biggest table(s) otherwise you may not have the space at the end of the task.

If you do not have the space to do this on the dev, if you have the access rights and space on the live box, you can create sample tables on the live using the same techniques and then archive them to tape using arcmain and restore to dev.

Another method would be to export the sample data to excel (a csv file) and then import it to the dev box. Depends how much data you want to use.

Re: Teradata dump - Prod to dev

with Fastload - store  prod data in a  file then fastload into an emtty table in Dev

with Unix scripts - Get the teradata table restore scripts from the DBAs and run that.