Restarting the RACF subsystem

When the RACF® subsystem is active, it detects subsystem failures and attempts to restart itself. This processing occurs multiple times. If the subsystem is unable to restart successfully, it eventually terminates. (For a description of the information that can be lost when the RACF subsystem address space is not active, see Stopping the RACF subsystem address space.)

If the subsystem terminates, after you resolve the error condition you can restart the subsystem manually in one of two ways. The first method you should try is to enter the following command at the MVS™ operator's console:

START xxxx,SUB=MSTR

where xxxx is the subsystem name chosen by the installation. Of the two restart methods, this method is the least destructive to current work executing in the subsystem.

If your attempt to restart the subsystem with the preceding command fails, it might still be possible to restart the RACF subsystem by entering the following command at the MVS operator's console:

START xxxx,SUB=MSTR,PARM=INITIAL

where xxxx is the subsystem name chosen by the installation. This second form of the START command might be necessary in rare circumstances where residual data is incompatible and cannot be reused.

The RACF ASID is not reusable. Do not specify REUSASID=YES on the START command.

When you use command prefix registration (CPF), the RACF prefix is registered only once, during subsystem initialization at IPL. Restarting the RACF subsystem does not alter the command prefix registration. If you restart the RACF subsystem after a failed attempt to register the command prefix with CPF, the subsystem uses the default command prefix (the subsystem name) and the prefix is not registered with CPF.