Procedure requirement for Concurrent access

Question

Hello all
i need to write a procedure with the following requirement
I got a table Employee , this table may contain 9 million records, and few users of the same company will be accessing this Table (Employee ).
Ie select * from Employee where company = AB order by datecreated ;
Suppose this query will fetch 25000 records , and the company AB has got 150 users .
My requirement is when the 150 users of the company AB fire the above query at the same time I have to fetch the first record which returns by the query and return it to the first user by updating the user_id column by the fist users, and second record to the second user etc. but more than 1 user should not get the same record.
To fetch the first row I have used the TOP N method and locking the record with FOR UPDATE
Select * from (select * from Employee where company = AB order by datecreated)
Where rownum=1 for update ;
But FOR UPDATE locks the row for the first user and update the use_id column with first userid when the second users waits for the lock release and when he gets it released it gets overwritten by the second users , my purpose is not getting satisfied.
I tried with FOR UPDATE NOWAIT and handled the exception and put it in loop . but as the number of SELECT his on DB is more its crashing it ..
Will SKIP LOCKED help me with this issue ?? but no one recommends that because its not documented in 10g.
This is for a web application , please suggest me an appropriate way to handle this issue ..
mishra

Answer 1

Hi Justin
the topic you have given me completely new to my knowledge and it is too waste , if you dont mind can give an introduction on this like , which part i have to look into for my requirement , and how i got to make it work to my requirement,?
Mishra