Archive masks

An archive mask is a template for creating archive data set names.

A mask is composed of variables and literal values in the format of a data set name. A mask value can be up to 44 characters in length. When an archive data set is allocated, the variable symbols in the archive mask are replaced with specific values to generate a data set name.

You can specify two types of variable symbols in an archive mask:
  • Product symbols, which are provided by Output Manager
  • z/OS® system symbols, which are provided by your z/OS operating system

Each variable name in an archive mask begins with an ampersand (&). If you specify a series of variables, insert a period (.) at the end of one variable name and before the next variable name (for example, &JOBNAME..&JOBNUMB). In this situation, the period is used as a concatenation character and will not be part of the archive data set name.

Ensure that the combination of literal values and variables that you specify for an archive mask will result in valid, unique archive data set names. Otherwise, errors or unpredictable results will occur.

To help ensure that these names are unique, you can include the product symbol &N in your archive masks. This symbol causes a generated sequential number to be added to the archive data set names for every instance of the archive.

Important: If you are using multiple output classes for a job, you must include the &N variable in the archive mask to ensure that unique archive data set names are generated for each output class. You should also include the &N variable if you are basing your archive data set names on writer names (that is, you set the Archive Name Rule Source field to Win the associated selector rule or subselector rule).