DSXchange

If I use an IPC stage between a DRS stage and a hash file, will it improve the performance. I'm trying to load data from an Oracle DB table containing around 3 million records into a hash file and it took around 30 minutes, I want to improve upon this.

Do you use all the records you are loading into the hash file?
If not then doing a lookup without a hash file would improve your performance.
Please post your scenario, So that people can help you better.
There are a lot of ways to improve your current performance, if you can post the exact scenario.

First determine the bottleneck. If you replace the Hashed File stage with a Sequential File stage, who long does it take to run then?

This hash file acts as a shared lookup across different warehouses(In my case warehouses differ on the basis of line of business - Finances, HR, etc..). and all the rows are needed.

Are you using Static or Dynamic Hashed files?
Wrong settings in Static hashed files can reduce the performance drastically.

Were you able to identify the bottleneck using Craig's suggestion.
You optimization route will depend on that.