Alternative way of Updating of Subinventory Accounts

Question

Hi.
Issue at hand : How to update (alternative way) the subinventory accounts of FIN and STAGE?
Facts :
1. Current setup includes - OM, Financials, INV ( but only to a very minimum level, having 2 Subinventories FIN and STAGE used for Order-to-Cash Cycle )
2. Recently, MFG and INV are being implemented.
3. Standard Costing
Since, there are already onhand and transactions using FIN and STAGE, we are looking at updating the accounts (setup) of FIN and STAGE to a different set.
I am aware of the procedure provided by Oracle on updating Subinnventory accounts, creating a new subinv and transferring onhand to that new subinv, handle reservations, etc., then update the account setup, and return the onhand, etc.
One team member suggested that we update the subinventory accounts directly from the database, similar to the script below:
Update MTL_SECONDARY_INVENTORIES
set material_overhead_account = 78373,
overhead_account = 78373,
resource_account = 78373,
outside_processing_account = 78373
where organization_id in (1234, 9999) and secondary_inventory_name in ('FIN', 'STAGE');
Question: What do you think is/are the implications of this process?
Your help would be very much appreciated.
Thank you.

Answer 1

Basically I have windows server with Oracle 11g (11.2.0.1) server and the database on this server needs to retired and move the data of it to a new server has Linux 64 and Oracle 11g built on it. Here are my answers below:
What exactly have you tried and done?
A.  I used "Cross-Platform Transportable Database" to transfer data from Windows to Linux and I am already done, database on the new server (Linux) has been tested and upgraded to latest release (11.2.0.4), my next step is the cutover but the data is too old (2 weeks ago) so I need to take a new data from the current prod server.
How long did it take and what is your downtime requirement for this migration and what how do you define a "huge database"?
A. RMAN convert took approximately 5hrs to convert data files from Windows to Linux (Please note that I used rman convert on source server)
and It took me a day and half to copy the converted files remotely from source in which files been placed to the destination server (the network bandwidth was about 100kb/s so it was so slow) next time I will copy data files on external hard disk then mount them on the new server if I use RMAN convert again.
End users accepts that the database can be in read only mode for one day business hours but they can't take it more so I'm really concern about the time when taking a new snapshot of data. Also, the size of data is about 380G and that's why I asked again if data pump will be more useful and faster than RMAN for copying data to the new platform.
Sorry if I made you confused!
Thanks again for all the support on here.