March 30, 1979 When you do not specify a data set name, AftSERV obtains the name from the AMSERV file. In the case of defining or deleting space, no data set name is needed; the FILE parameter corresponding to the ddna.e is all
that is necessary, and AMSERV assigns a default data set name to the
space. When you define space on a minidisk using AftSERV, CftS does not check
the extents you specify to see whether they are greater than the number
of cylinders available. As long as the starting cylinder is a valid
cylinder number and the extents you specify are on cylinder boundaries,
the DEFINE function completes successfully. However, you receive an
error message when you use an AftSERV function that tries to use this
space.
To define a cluster for VSAM space that has already been allocated, you
need (1) an AMSERV file containing the control statements necessary for
defining the cluster, and (2) the master catalog (and, perhaps, user
catalog) volume, which will point to the cluster. The volume on which
the cluster is to reside does not have to be online when you define a
suballocated cluster.
For example, the file CLUSTER AMSERV contains the following:
DEFINE CLUSTER ( NAME (BOOK.LIST) - VOLUMES (123456) - TRACKS (40) -
FILE (BOOK) - KEYS (14,0) RECORDSIZE (120,132) )­
DATA (NAME (BOOK.LIST.DATA) ) - INDEX (NAME (BOOK. LIST. INDEX) )
To execute this file, you would need to enter the following command sequence (assuming that the master catalog, on volume 123456, is in your
virtual machine at address 310): access 310 b "_lttll •• ,tlltr dlbl ijsysct b (perm 1 •• GlI amserv cluster Bote that to define a suballocated cluster, you do not need to provide a
DLBL command to define it to AMSERV. For a unique cluster (one defined with the UNIQUE attribute), you must define the space for the cluster at the same time you define its name and attributes; thus the volume or volumes on which the cluster is to
reside must be mounted and accessed when you execute the AMSERV command. You must supply extent information for the cluster's data and index
portions separately.
To execute an AMSERV file named UNIQUE which contains the following
(the ellipses indicate that the AMSERV file is not complete): 206 IBM VM/370 CMS User's Guide

DEFINE CLUSTER - (NAME (PAYROLL) ) -
DATA ( FILE (UDATA) - UNIQUE - VOLUMES (567890) - CYLINDERS (40) -
. .. ) - INDEX ( FILE (UINDEX) ) - UNIQUE - VOLUMES (567890) - CYLINDERS (10) -
. .. )
the command sequence should be:
access 350 c
dlbl udata c (extent DMSDLB331R ENTER EXTENT SPECIFICATIONS: 800 800 c
dlbl uind tent 600 200 c
amserv unique When you use AMSERV to delete a VSAM cluster, the volume containing the
cluster does not have to be accessed unless the volume also contains the
catalog in which the cluster is defined. In the case of data spaces and
user catalogs or the master catalog, however, the volume(s) must be
mounted and accessed in order to delete the space. When you delete a cluster or a catalog, you do not need to use the
DLBL command, except to define the master catalog; AMSERV can obtain the
necessary file information from the AMSERV file. In the case of data
spaces, you must supply a ddname (filename) with the DLBL command, but
you do not need to use the DSN parameter. You should be particularly careful when you are using temporary disks
with AMSERV, that you have not cataloged a temporary data space or
cluster in a permanent catalog. You will not be able to delete the space
or cluster from the catalog. USING THE REPRO, IMPORT, AND EXPORT (OR EXPORTRA/IMPORTRA) FUNCTIONS You can manipulate VSAM files in CMS with the REPRO, IMPORT, and EXPORT functions of AMSERV. You can create VSAM files from sequential tape or
disk files (on OS, DOS, or CMS disks) using the REPRO function. Using REPRO, you can also copy VSAM files into CMS disk files or onto tapes.
For the IMPORT/EXPORT process, you have the option (for smaller files)
of exporting VSAM files to CMS disks, as well as to tapes. You cannot, however, use the EXPORT function to write files onto as
or DOS disks. Nor can you use the REPRO function to copy ISAM (indexed
sequential) files into VSAM data sets, since CMS cannot read ISAM files. When creating a VSAM file from a disk file, the device track
size must be the maximum BLOCKSIZE in the INFILE statement. AMSERV expects a DOS or OS file as input and will not open a disk when the Section 10. Using Access Method Services and VSAM 207