How to handle the locks (table, rowlevel, record level, data

What do you mean when you ask "handle" the locks? Locking is at 'generally' at a record level, where you lock a record, do updates and release/downgrade the lock. It is essentially handled within the source code.

There is only one thing you need to do and that is to be sure that the size of the lock table is large enough. In the promon database monitoring program there is a screen that shows the lock table high water mark and the current number of locks in use. The high water mark is the maximum number of locks that were in use at the same time. You just have to look at that occasionally to see that it is comfortably below the lock table size.  If an application consumes all the locks, then it will get an error and the current transaction will fail.