MVS- Catalog

A catalog is the directory of data sets on an MVS system. It connects a data set name to the disk volume where the data set lives.

Without a catalog entry, the system cannot find your data set by name alone - you must also tell it the volume.

What the catalog is

• A catalog associates a data set with the volume on which it is located.
• A catalog entry stores: the data set name, the volume serial (volser), the unit (device type), and device information.
• Locating a data set needs the data set name; the catalog supplies the volume and unit for you.
• A typical system has one master catalog and many user catalogs.

Master catalog and user catalogs

• The master catalog contains entries for system data sets and entries that point to user catalogs.
• User catalogs contain entries for user data sets.
• There is only one master catalog, but any number of user catalogs.
• VSAM data sets must be cataloged. Non-VSAM data sets do not have to be, because their labels are stored in the VTOC.

Catalog aliases

• The high-level qualifier (HLQ) of a data set name usually decides in which user catalog it is cataloged.
• A user catalog can have several aliases, so files with different HLQs can live in the same user catalog.
• Example: aliases PAY.* and PAYROLL.* can both point to the same user catalog.
• If the HLQ has no alias, the data set goes to the master catalog by default - which most shops forbid for user data.

How a data set is located

• Step 1: the system searches the catalog for the data set name.
• Step 2: the catalog entry gives the volume serial and unit.
• Step 3: the volume is located (mounted if needed) and its VTOC is read.
• Step 4: the VTOC's DSCB gives the exact location of the data set on the volume.
• Catalog search goes: master catalog -> matching user catalog -> data set entry.

Three levels of data set allocation

• Level 1 - unit allocation: the system picks the device. You can give a generic name (like 3390) or a group (esoteric) name.
• A group name such as SYSDA (disk for general use) or TAPE (tape drives) is defined by each installation.
• Level 2 - volume allocation: for a new data set you may give a specific volser, or let the system choose (non-specific request).
• Level 3 - data set allocation: for an existing data set the labels are checked; for a new data set the labels are created, space is allocated, and the VTOC is updated.

Disposition: old or new data set

• The DISP parameter in JCL tells the system whether the data set already exists or is new.
• DISP=(NEW,...): create the data set.
• DISP=(OLD,...): use an existing data set, with exclusive access.
• DISP=(SHR,...): use an existing data set, shared read access.
• DISP=(MOD,...): add records to the end of an existing data set, or create it if it does not exist.
• The second part of DISP says what to do at the end: CATLG (catalog it), KEEP, or DELETE.

OPEN modes and CLOSE

• Before a program does I/O on a data set, it must issue an OPEN.
• INPUT mode: records can only be read from an existing data set.
• OUTPUT mode: records can only be written to a new or existing data set.
• I/O mode: records can be read, written, updated, and deleted.
• When finished, the program issues a CLOSE to disconnect from the data set.
• A data set is automatically deallocated when the job step ends.

LISTCAT: list catalog entries with IDCAMS

• IDCAMS LISTCAT displays catalog information about data sets.
• LISTCAT ENTRIES(name) shows one data set; ALL shows full details.
• LISTCAT LEVEL(qualifier) lists every entry starting with that qualifier.
• LISTCAT with CATALOG(name) searches a specific user catalog.

Binary, hexadecimal and EBCDIC notations

• Bits are binary digits: 0 (off) or 1 (on). Eight bits make one byte.
• Hexadecimal (hex) uses 16 digits: 0-9 and A-F. One hex digit represents 4 bits.
• Hex is used by mainframe debugging tools to display field contents.
• EBCDIC (Extended Binary Coded Decimal Interchange Code) is the character code used on IBM mainframes.
• Most other computers use ASCII (American Standard Code for Information Interchange).
• Examples: in EBCDIC, A = X'C1', space = X'40', digit 0 = X'F0'.

EBCDIC vs ASCII: collating sequence

• The character code decides the sort (collating) sequence of the computer.
• In EBCDIC, special characters come first, then letters, then numbers.
• In ASCII, special characters come first, then numbers, then letters.
• Moving data between mainframe and PC requires ASCII-to-EBCDIC conversion (or the reverse).
• The sort order of a file changes during the conversion, so re-sort after converting if order matters.