Storage Management Subsystem (SMS) - IBM Mainframe

The Storage Management Subsystem (SMS) provides a range of data and space management functions. SMS improves storage space use, controls external storage centrally, and lets you manage storage growth. It makes it easier to convert to new device types. It takes advantage of what available hardware can do. With SMS, you can move toward system-managed storage.

SMS manages an installation's storage according to the currently active storage management policy. Through the Interactive Storage Management Facility (ISMF), you define an installation storage management policy in an SMS configuration. An SMS configuration contains the following:

  • Base configuration information
  • Classes and groups
  • Automatic class selection (ACS) routines
  • Optical library and drive definitions
  • Tape Library definitions

The base configuration identifies the systems that the SMS configuration is to manage. These systems constitute an SMS complex. The base configuration also contains installation defaults.

You can define more than one control data set, but only one at a time controls SMS. Each control data set defined for SMS is called a source control data set (SCDS). The control data set that is in effect at a given time is the active control data set (ACDS).

SMS classes and groups are lists of traits and characteristics that are associated with or assigned to data sets, objects and volumes. An SMS configuration can contain the following five types of classes and groups:

  • Storage group allows you to define a list of volumes and manage them as if they were one large, single volume. SMS applies the properties you assign to a storage group to all the volumes within the storage group.
  • Management class allows you to define different levels of migration, backup and retention services. Through management class, you can associate a level of service with a data set or object that is independent of the physical location of the data set or object. Also, you can identify an object characteristic that might trigger a class transition.
  • Storage class allows you to define different levels of performance and availability services. Through storage class, you can separate the level of service for a data set or object from physical device characteristics. You can also separate the level of service for an object from physical device characteristics with different storage classes used to place objects at various levels of the storage hierarchy.
  • Data class allows you to define allocation defaults. Through data class, you can simplify and standardize the allocation of new data sets.
  • Aggregate Group allows you to define groups of data sets for the purpose of backing up or recovering all data sets in a group in a single operation.

An SMS configuration can contain multiple constructs of each type. Data sets managed by SMS are called system-managed. Each system-managed data set or object must reside in a storage group. The system-managed data sets must have a storage class, and might also have a management class and a data class. The objects must have a storage class, a management class, and cannot have a data class.

You can assign the same name to various SMS classes and a storage group, if you wish. For example, a data class and a storage class can have the same name.

ACS routines determine the SMS classes and storage groups for data sets and objects. You can also use ACS routines to control the transition of data sets and objects to and from SMS management.

Advantages and Disadvantages of SMS

Data sets that are created by SMS are called SMS Managed data sets. The following are the advantages of SMS:

  • User is relieved of making decisions about resource allocation of data sets, since it is done by SMS.
  • SMS provides the capability or concatenating data sets of unlike devices. This capability is not available in non-SMS environment.
  • SMS Managed data set cannot be deleted unless they are first uncataloged. Due to this extra step, erroneous deletions of dataset are minimized.
  • Additional features are available in the use of IDCAMS in the SMS environment.
  • VSAM data sets created in an SMS environment offer more flexibility than those created through JCL in non-SMS environment.

The following are the disadvantages of SMS:

If SMS is used to request space for a data set and there is not enough space left for the secondary allocation, then the job will abend. For non-SMS managed data sets, subsequent volumes of the device for which the request is issued are searched, and the job does not abend.

In SMS, if a non-existent data set name is created and referenced in the same step, an error is encountered. A data set must exist before the start of a job, if it referenced in it. For non-SMS managed data sets, a data set can be created and referenced in the same job.

DISP Defaults in SMS*

In a SMS managed environment the statement DISP= (NEW, KEEP) defaults to DISP= (NEW, CATLG). In a non-SMS managed environment the CATLG sub-parameter has to be coded explicitly.

In a SMS managed environment the statement DlSP= (OLD, DELETE) will delete and uncatalog a data set. Some of the SMS parameters are given below:

STORCLAS Parameter

This is a keyword parameter. It is used to assign a data set to an SMS defined class. This parameter has significance only if SMS is active, otherwise it is ignored. The syntax of this parameter is 'STORCLAS=class', where class is an installation defined name and can be 1 to 8 characters long. Use of this parameter result in the data set defined in the DD statement within that job being SMS managed. Since the data set will be managed by SMS the VOLUME and UNIT parameters can be omitted since these values will be supplied by SMS.

DATACLAS Parameter

This is a keyword parameter. It is used to define any or all of the following parameters for a data set: LRECL, RECORG, RECFM, RETPD, EXPDT, volcount coded on the VOL parameter, SPACE and AVGREC. For SMS managed VSAM data sets, DATACLAS can be used to define the following parameters: CISIZE, IMBED, REPLICATE, SHAREOPTIONS and FREESPACE. The data set itself can be any one of the following type: Physical Sequential (PS), Partitioned (PO), VSAM, Direct Access (DA), etc. The syntax of this parameter is 'DATACLAS=class where class is installation defined and can be 1 to 8 characters long. Use of this parameter in a DD statement results in the pre-defined values for any or all of the above parameters being used for the associated data set.

MGMTCLAS Parameter

This parameter is used to provide a management class for the associated data set coded in the DD statement. A management class is used to control the migration of data sets, the frequency of back-ups, the number of back-up versions, and the retention criteria of back-up versions. The syntax of this parameter is 'MGMTCLAS=class where class is installation defined and can be Ho 8 characters long. Parameters defined for this class cannot be overridden. Use of this parameter in a non-SMS managed data set will result in a JCL error.

RECORG Parameter This parameter is used to create SMS managed VSAM data sets. It is significant only at the time that a VSAM data set is created. It is ignored if the data set already exists. The syntax of this parameter is 'RECORG=option', where 'option' can be any of the following: KS (Key Sequenced Data Set), ES(Entry Sequenced Data Set), RR (Relative Record Data Set) LS (Linear Space Data Set).

KEYOFF and KEYLEN Parameters

The KEYOFF parameter is used in conjunction with RECORG and KEYLEN parameters. It is required of KSDSs. It identifies the key offset, which is the position of the first byte, of the key field for the data set, relative to the first byte of the logical record. The syntax of this parameter is 'KEYOFF=offset'. The KEYLEN parameter is used to define the length of the key. The syntax of this parameter is 'KEYLEN = n', where 'n' is a number.

AVGREC parameter

This parameter is used to request space in units of number of records. The syntax of this parameter can be in any one of the following formats:

  • AVCREC=U - This is used to indicate that the primary and secondary allocations of the SPACE parameter will be equal to the number of logical records specified and multiplied by 1.
  • AVOREC=K - This is used to indicate that the primary and secondary allocations of the SPACE parameter will be equal to the number of logical records specified and multiplied by 1024.
  • AVGREC=M - This is used to indicate that the primary and secondary allocations of the SPACE parameter will be equal to the number of logical records specified and multiplied by 1,048,576.

LIKE parameter

This parameter is used to copy attributes from an existing cataloged data set to a new data set. The following attributes can be copied over: SPACE, RECFM, AVGREC and LRECL. The following attributes can be copied over VSAM data sets: RECORG, KEYLEN & KEYOFF. The syntax of this parameter is 'LlKE=model.dataset', where 'model.dataset' identifies the name of the model data set.


All rights reserved © 2018 Wisdom IT Services India Pvt. Ltd DMCA.com Protection Status

IBM Mainframe Topics