Exit programs
An exit program is a program to which control is passed from a calling application program or system program. Exit programs can be used to customize particular functions to your needs.
Exit programs are usually user-written programs; however, a few are system-supplied (such as a few of the Operational Assistant exit programs).
To transfer control to an exit program, you do an external call as you would to any other program.
There are no general requirements for using exit programs. For any specific requirements, see the
documentation for the specific exit program. 
Using SQL within an exit program
should be avoided if there is any situation where the exit program could be called while running an
SQL statement.
Exit points
An exit point signifies the point in a system function or program where control is turned over to one or more exit programs to perform a function.
The registration facility provides a central point to store and retrieve information about IBM i and non-IBM i exit points and their associated exit programs. This information is stored in the registration facility repository and can be retrieved to determine which exit points and exit programs already exist.
You can use the registration facility APIs to register and deregister exit points, to add and remove exit programs, and to retrieve information about exit points and exit programs. You can also perform some of these functions by using the Work with Registration Information (WRKREGINF) command.
The exit point provider is responsible for defining the exit point information, defining the format in which the exit program receives data, and calling the exit program.