Sign up for our newsletter
Technology / AI and automation


Anyone who assumed that our couple of pages last week had wrung IBM’s latest major announcement dry is living in cloud cuckoo land, but we should be able to despatch the rest of the mainframe software that was announced today. Probably the most important of the remaining announcements is Data Facility Hierarchical Storage Manager – DFHSM if you must – Version 2 Release 4, which is claimed to bring major storage management improvements. The new release implements new and improved DFHSM space and availability functions that manage data sets according to new service attributes, rather than according to the attributes of the volume on which they reside. DFHSM Version 2 Release 4 and other Data Facility family products, plus the RACF product, provide the complementary functions that make up Data Facility Storage Management System, DFSMS. The new release implements customer-defined storage management policies through the Storage Management System storage groups; management classes; and automatic class selection routines. In addition, whether you have the Storage Management System, the new release adds deletion of all expired backup versions with a single command; automatic system reblocking of sequential data sets; and parallel migration paths to DFHSM-owned tape or disk. The new release operates with DFSMS to manage each data set according to its assigned DFSMS service attributes. These attributes are defined by the storage administrator and written in a common DFSMS service repository, and class attributes are preserved by DFHSM space and availability management functions when data sets are migrated or backed up. MVS/DFP Version 3 automatic class selection routines select volumes within a storage group. The common DFSMS service repository and DFHSM-preserved class attributes are used by these routines to select a volume when DFHSM recalls or recovers a data set. Storage Group service attributes determine which DFHSM management policies – automatic migration, incremental backup, and dump – should be applied to the disk volumes in that group, eliminating the need for a storage administrator to add or delete volumes for management by DFHSM. Management Classes determine how DFHSM management policies should be applied to each data set or groups of data sets and provide the type of flexible data set controls that customers have requested. Service for DFHSM managed and owned data sets includes: release of unused data set space on user disk volumes; eligibility for automatic/command migration from user disk volumes to DFHSM-owned disk and/or tape volumes based on elapsed days since last used; early migration of generation data sets and migration or deletion of rolled-off generation data sets; data set deletion based upon elapsed days since last use, creation, a specific date, or no limit; eligibility for automatic or by-command back-up with minimum frequency in days; number of back-up versions to be retained while data set exists and after deletion; retention in days of current and noncurrent backup versions after DFHSM determines that data set deletion has occurred. Automatic Class Selection Routines can be customised to convert data sets on DFHSM-owned volumes to volumes in DFSMS storage groups and data sets that are migrated from non-DFSMS volumes can be recalled to DFSMS volumes, and data sets that were backed up when they were on non-DFSMS volumes can be recovered to DFSMS volumes. DFHSM 2.4 will be available in the US from December 31 and will be a one-time $38,400 for Group 40 CPUs, $24,000 for Groups 20 and 30.

White papers from our partners

This article is from the CBROnline archive: some formatting and images may not be present.

CBR Staff Writer

CBR Online legacy content.