Row level security
Hi,
I am on 10.1.3.4
I do not currently have LDAP configured, but I would switch to it if it were the only way.
The documentation / blogs are clear enough when it comes to how to implement row level security, but can anyone advise on the best way to do this when users are not associated with single values, but instead multiple values, and multiple values so great that it would prohibit striping the source data in the data warehouse to resolve the tables to single row security.
i.e.
We have a table Cost_Centres.
Say
100000
200000
300000
400000
500000
600000
700000
800000
(but in practise there are many thousands of these)
Cost Centres are alpha numeric strings of length 6 characters.
We have tables Balances, Budgets, Transactions etc - all of which contain a single cost centre for each row.
As an example
Balances
PERIOD COST CENTRE BALANCE
APR-10 700000 500
APR-10 100000 1200
APR-10 800000 900
APR-10 300000 1900
We have a user Joe Bloggs.
Joe Bloggs is configured to the following cost centres => 200000, 300000, 800000 (but in practise he may have 300+ cost centres)
My question is how to set up row level security such that I do not have to stripe all of my tables (which are already very large) and without a massive detioration in speed?
So when Joe Bloggs runs a report against balances he only sees
APR-10 800000 900
APR-10 300000 1900
thanks for your input,
Robert.
Edited by: Robert Angel on 09-Nov-2010 04:11