11th January 2017

IBM z/OS Reference

Concepts

Files (datasets)

The dataset is the mainframe equivalent of a file. In its basic form, a dataset is record-oriented, by default it consists of a fixed number of fixed size records. Therefore, creating a dataset requires a lot of low-level parameters that would be handled by the filesystem on other operating systems. Datasets can be of various types, including advanced, database-like storages for higher level software like RACF, DB/2 or CICS; the default is “NONVSAM”.

Directories

There are no directories on mainframes. Instead, you can create a special dataset called “partitioned dataset” (“PDS”) which is essentially a container of flat files. This is what you’d use for COBOL source code, REXX scripts, JCL jobs and other basic documents.

Filenames

Under RACF, the first component of the filename is the user or group owning the file.


Commands

IPL

Instruct hercules to IPL from disk A80:

IPL A80

On the console terminal, reply to all questions asked:

r 0,r
r 0,...

Then, after it finished booting, login on a regular terminal as user FOO:

LOGON FOO

Shutdown

In the console, replying to questions that may be asked by these:

F OMVS,STOPPFS=NFS
F OMVS,SHUTDOWN

Then logout from all terminals/console and wait until all CPUs are stopped.

List datasets in catalog

LISTCAT

List dataset FOO.BAR.BAZ settings

LISTDS FOO.BAR.BAZ

Low-level disk editor to look at datasets in their on-disk state (before e.g. PDS interpretation)

DITTO

Emulation Software

IBM Mainframe Architecture Emulator

IBM 3270 Terminal Emulation


Back to Knowledge Base.

Ƿ