File generation groups

ARCHIVE also supports file generation groups (FGGs). These are files which are chronologically connected and have identical file attributes. FGGs are mainly intended for the long-term storage of large data sets, e.g. of data from processing sequences that recur at regular intervals. FGGs thus also afford convenient support for data saving in accordance with the grandfather-father-son principle.

Each file generation is handled as a separate file. Both entire file generation groups and individual file generations can be saved with the usual save facilities (with or without directory, CHANGED=YES, etc.). However, partial saves are not possible.

The following should be noted when saving FG(G)s:

1. A file generation is saved if the following statement is entered:

   FILES NAME=<file-generation-group-name>(*n)

2. A file generation group is saved if the following statement is entered:

   FILES NAME=<file-generation-group-name>

   The generation index and all entries are processed:

   • If CHANGE=NO, the generation index and the entries are saved.

   • If CHANGE=YES, two cases have to be distinguished:

     • If the first and the last entry of the FGG have not been updated since the last full save, the generation index is not saved; the entries are saved FULL or CNS (see section "Directory file"), depending on whether or not they have been updated since the last full save.

     • If the first or the last entry of the FGG have been updated since the last full save, the generation index and all entries are saved.

   • If the ERASE operand is specified in a save run, the generation index is not deleted if not all entries have been saved.