Creating and Mounting the Root File System

Create an HFS data set to contain the root file system by running a job to allocate the data set. During allocation, z/OS® UNIX builds a basic root directory, which you can alter to meet your specific needs. You specify DSNTYPE=HFS to designate an HFS data set.

   //STEP1  EXEC PGM=IEFBR14
   //MKFS   DD   DSNAME=OMVS.ROOT,DISP=(NEW,CATLG),
   //            DSNTYPE=HFS,SPACE=(CYL,(1,1,1))

Tip: To make the HFS data set SMS-managed, use the ACS routines or specify the STORCLAS parameter in the JCL. HFS data sets do not have to be SMS-managed.

Write operations present the greatest exposure to file system damage. For this reason the root file system should be small, minimizing the amount of write activity to it, thus offering the least exposure to damage.

Additionally, if all users' files are in file systems that are mounted on the root file system, rather than defined as part of the root itself, and users are denied write access to the root, the root is further protected from inadvertent damage. Damaged user directories or files can be unmounted and replaced without causing z/OS UNIX system services to fail.