Data set level mirroring to accomplish a volume merge/migrate in a digital storage system

Abstract

Users of Mainframe computers running under IBM's MVS operating systems have a need to merge migrate data from multiple smaller DASD devices (disk volumes) to larger DASD devices, and/or to migrate data in order to combine data set extents. A method is disclosed for migrating data where a volume merge is involved, and/or when combining of extents is desired, that minimizes the down time of applications using the involved data while satisfying MVS rules for volume and catalog meta-data files as well as data extent limitations. The method may be implemented such that source and target volumes need not be in like manufacturer DASD devices.

1 . A method for data set-level migrating a selected source data set under a Multiple Virtual Storage (MVS) operating system while minimizing downtime of the source data set, the method comprising: 
 (a) identifying a set of source data set tracks on which the source data set is stored;    (b) initially allocating a target data set with at least sufficient space to store the source data set tracks;    (c) copying the source data set tracks to the target data set;    (d) while the copying proceeds, monitoring to detect any changes to a source data set track after the track was last copied, thereby identifying as a modified track one that requires re-copying;    (e) continuing the copying of source data set tracks until all of the source data set tracks have been copied, and all modified tracks have been re-copied;    (f) after completion of the copying, closing and unallocating the source data set;    (g) responsive to an indication that the source data set has been closed and unallocated, effecting a final re-synchronization step by re-copying any modified track not yet re-copied;    (h) in response to detecting an allocation change having been made to the source data set after the target data set was initially allocated, adjusting the target data set to accommodate the allocation change;    (i) updating a corresponding catalog and system meta-data describing the source data set to point to the target data set; and then    (j) signaling release to open the target data set.    
     
     
         2 . A method for data set-level migrating according to  claim 1  wherein the detected allocation change to the source data set after initially allocating the target data set comprises an extension to the source data set, and said adjusting the target data set comprises, if necessary, extending the target data set so as to accommodate the source data set extension.  
     
     
         3 . A method for data set-level migrating according to  claim 1  wherein the detected allocation change to the source data set after initially allocating the target data set comprises a release of space in the source data set, and said adjusting the target data set comprises a corresponding release of space allocated to the target data set.