Defining volume and data set attributes for data class

From Page 1 of the Data Class Define panel, issue the DOWN command to view Page 2, on which you specify volume and data set attributes.

Data Set Name Type

Specifies the format in which the data set is to be allocated. When you specify extended format, you can also select requirements for the data set, including the need for extended addressability.

See z/OS DFSMS Using Data Sets for detailed information on assigning and allocating data classes for extended format data sets.

The available values are:

EXT

Specifies that the data set to be allocated is in the extended format. All VSAM data set types can be allocated in the extended format (except for key range data sets, temporary data sets and system data sets).

If you specify EXT, you must also specify a value for the If Ext subparameter. The other subparameters are primed with the following defaults: Extended Addressability = No, and Record Access Bias = USER.

Sequential extended format data sets can be version 1 or 2. The user can specify the version number with the second value of DSNTYPE on the DD statement or dynamic allocation equivalent. The PS_EXT_VERSION keyword in the IGDSMSxx member in PARMLIB can change the default for the version number.

Note: Encrypted data sets are created in version 2 format, regardless of user specification, if a key label is present.

If the sequential extended format data set is striped or if it resides only on one volume, then the version number does not matter. If it is not striped and it resides on multiple volumes, then the version number affects whether the system can use FlashCopy®. FlashCopy can handle version 2 in that case but not version 1. PTFs are available to allow proper handling on releases earlier than z/OS® 2.1.

HFS

Specifies that the data set is a hierarchical file system (HFS) data set.

LARGE

Specifies that the data set is a large format data set. Large format data sets are physical sequential data sets with the ability to grow beyond the limit of 65 535 tracks per volume.

The BLOCKTOKENSIZE=REQUIRE option in the IGDSMSxx member of SYS1.PARMLIB affects what programs can open these data sets. See the description of the IGDSMSxx member in z/OS MVS Initialization and Tuning Reference.

These data sets cannot be written on a system older than z/OS 1.7. These data sets cannot be read on a system older than z/OS 1.7 if they have more than 65535 tracks on each volume.

LIB

Specifies that the data set is a PDSE.

PDS

Specifies that the data set is a partitioned data set.

blank

Leaves the Data Set Name Type attribute value unspecified. This is like coding DSNTYPE=BASIC on the DD statement but you cannot specify BASIC on this panel.

If you specify a Data Set Name Type of EXT for data sets allocated in extended format, you can specify additional attributes:

If Ext

Specifies whether allocation in extended format is preferred or required. If you specified EXT for the Data Set Name Type attribute, you must also specify one of the following:

P (preferred): The data set allocation is attempted in nonextended format if the necessary system resources for extended are not available.
R (required): The data set allocation fails if the data set cannot be allocated in extended format.

Extended Addressability

Specifies that a VSAM data set with the extended format attribute be allocated using extended addressability. Having extended addressability allows the data set to be allocated or grow beyond the 4 gigabyte (4 GB) size. The extended addressability data set can be a VSAM data set in any record organization but must be allocated as extended format. The only exception is for linear VSAM data sets which can have the extended addressability attribute and be non-SMS managed which means the data class does not have the extended format attribute set. To allow a VSAM LDS to be allocated or grow beyond 4 GB and be non-SMS managed, the following data class values are required:

Data Set Name Type is blank.
The If Ext value is blank.
Extended Addressability value is Y.
Recorg value is LS.

The available values are:

Y: Extended addressability is used.
N: Extended addressability is not used. This is the default.

Record Access Bias

Specifies whether to let VSAM determine how many and which type of buffers to use when accessing VSAM extended format data sets by batch processing. This is known as system-managed buffering, and is available to VSAM data sets in any record organization which are allocated in the extended format.

Tip: You can use JCL to override the buffering algorithms that are determined by VSAM.

The available values are:

Keyword	Action
User (the default)	System-managed buffer is not used.
SO	System-managed buffering with sequential optimization is used.
SW	System-managed buffering weighted for sequential processing is used.
DO	System-managed buffering with direct optimization is used.
DW	System-managed buffering weighted for direct optimization is used.
System	Specifies that VSAM chooses how many buffers to use and how they are processed. System-managed buffering only takes effect if the application requests the use of non-shared resources (NSR). It is not effective for applications requesting local shared resources (LSR), global shared resources (GSR), or record-level sharing (RLS). The number of buffers and the buffering technique are determined by the system or the user: The system, based on application specifications (ACB MACRF=(DIR,SEQ,SKP)) and the values for direct and sequential millisecond response (MSR) and bias specified in the storage class. The user, as specified on the JCL DD card (AMP=('ACCBIAS=). If sequential processing is to be used, the system optimizes the number of buffers and uses the NSR buffering technique. If direct processing is to be used, the system optimizes the number of buffers and uses the LSR buffering technique. You can change the defaults for space and the relative amount of hiperspace when direct optimization is used by using the following keywords in the AMP= parameter: SMBVSP=, SMBHWT=, and SMBDFR=.

Force System Determined Blocksize

If the data class or the user specify DSORG=PS or PO for a new data set, then the system attempts to calculate an optimal value for the maximum block size (BLKSIZE). Normally, the user can override this by coding the BLKSIZE keyword on the DD statement or dynamic allocation equivalent. This might be because the user has a special requirement for an inefficient BLKSIZE value. To specify whether the system ignores the user-specified maximum block size and calculate the data set's maximum block size when allocating space for the new data set, select one of the following options:

Y: (1) If the user codes the BLKSIZE keyword and unallocates the new data set before opening it for writing, then the system discards the BLKSIZE value that the user coded. This implies that the storage administrator knows that the user specified an inefficient value for BLKSIZE. The data set label retains the system-determined value until a program opens it for writing. When a program eventually opens the data set for writing, that program can use or override the system-determined block size.; (2) If a program opens the new data set for writing while the DISP value is NEW or the data set is new with DISP=MOD, then this data class option has no effect and OPEN use the BLKSIZE value that the user coded on the DD statement or dynamic allocation. In this case, the BLKSIZE value that the user specified overrides the value that the system determined. This implies that the user had a reason to override the system-determined value for BLKSIZE.
N: The user can override normal system-determined block size logic when the data set is allocated and when it is opened for writing. N (No) is the default.

RMODE31

RMODE31 allows you to specify whether to allocate the buffers and control blocks in 31-bit addressable storage. Start of change

This option applies only to VSAM data sets. For other kinds of data sets, the DCBE macro provides a similar option for assembly language programs. End of change

The values you might specify for RMODE31 are:

ALL: Control blocks and buffers above the line.
BUFF: Buffers (only) above the line.
CB: Control blocks (only) above the line.
NONE: Control blocks and buffers below the line.

Dynamic Volume Count

Dynamic Volume Count is used during allocation processing to determine the maximum number of volumes a data set can span. The number can be in the range 1 through 59. The default value is 1. 59 is the z/OS volume limit.

During define processing, Dynamic Volume Count allows for a larger number of volumes to be considered without increasing the number of candidate volumes stored in the catalog. During existing data set allocation, it provides a way to increase the number of TIOT/JFCB entries that are created, so that more volumes can be dynamically allocated as required during the lifetime of the allocation. For VSAM data sets, the Dynamic Volume count is the maximum number of volumes that all components in the sphere being allocated can span.

Use caution when defining a Dynamic Volume Count greater than 1 for a data class that might be used for data sets that are allocated to the master scheduler address space (such as data sets in the LNKLST during PROGxx processing). This address space has a TIOT size of 12K, which allows for approximately 600 unit allocations. Use of a Dynamic Volume Count greater than 1 for those data sets reduces the number of data sets that can be allocated to the master scheduler address space. For further information regarding TIOT space calculations, refer to the description of the TIOT SIZE parameter in ALLOCxx in z/OS MVS Initialization and Tuning Reference.

Note: Dynamic volume count support when a data set extends is a function of access method end-of-volume (EOV) processing and not the data set type. For products that do not use the standard IBM® access method EOV interface for their data sets, DVC might not be supported. Consult with your product representative to see if it supports dynamic volume count.

All the following must be true for the Dynamic Volume Count value to be valid:

The Dynamic Volume Count is larger than the current volume count of the data set.
The data set is SMS-managed.
The Space Constraint Relief value is Y.

Dynamic Volume Count is not supported for the following:

Non-SMS managed data sets or data sets with no associated data class
System data sets:
- Page data sets
- Single-volume data sets such as a catalog BCS or VVDS, VSAM temporary data sets, or any system data set with ACBSDS=ON
SAM striped data sets
VSAM data sets with the IMBED or KEYRANGE options
VSAM EOV Snapshot processing

The Dynamic Volume Count field has significance only when it is larger than administrator-specified data class Volume Count, and any user-specified volume information provided by JCL, IDCAMS, or TSO.

Compaction

Specifies whether the data set is to be compressed format. You can compress data that is being written to an extended format data set or to a PDSE. This field is ignored for DASD data sets if the data set name type is not EXT or LIBRARY. You can compress data when it is being written to tape. A compressed data set cannot reside on the same tape cartridge as a data set that is not compressed.

The possible values are:

Y: Extended format or PDSE data sets are compressed and tape volumes are compacted. The type of DASD compression depends on the COMPRESS option in parmlib member IGDSMSxx. Tape volumes are compacted unless compaction is overridden by the user through JCL/dynamic allocation.
N: Data sets are not compressed. Tape volumes are not compacted, unless compaction is requested by the user through JCL/dynamic allocation.
T: Extended format data sets are compressed using tailored dictionaries. This overrides the COMPRESS option in parmlib member IGDSMSxx.
G: Extended format data sets are compressed using generic dictionaries. This overrides the COMPRESS option in parmlib member IGDSMSxx.
ZR: Indicates zEDC Required, meaning that the system should fail the allocation request if the zEDC function is not supported by the system, or if the minimum allocation amount requirement is not met.
ZP: Indicates zEDC Preferred, meaning that the system should not fail the allocation request, but rather create either a tailored compressed data set if the zEDC function is not supported by the system, or a non-compressed extended format data set if the minimum allocation amount requirement is not met.
blank: Data sets are not compressed. Tape volumes might be compacted depending on what was specified by the user on JCL/dynamic allocation, the installation with the COMPACT option in parmlib member DEVSUPxx, or the allocated hardware model. This is the default.

T, G, ZR, and ZP override the compression option set in the IGDSMSxx PARMLIB member, and let you select the type of compression on a data set level. Current® users of generic compression can move to using tailored or zEDC compression one data set at a time, as new data sets are created.

For an interactive starting point for Integrated Accelerator for zEDC, and access to a variety of technical resources, see Integrated Accelerator for zEDC.

Spanned / Nonspanned

Specifies whether a data record can span control interval boundaries. This applies to both system-managed and non-system-managed data sets. Specify one of the following:

Spanned: Specifies that if the size of a data record is larger than a control interval, the record can be contained on more than one control interval. This lets VSAM select a control interval size that is optimum for the DASD.
When a data record that is larger than a control interval is put into a cluster defined for spanned record format, the first part of the record fills a control interval. Subsequent control intervals are filled until the record is written into the cluster. Unused space in the record's last control interval is not available to contain other data records.

Restriction: Do not use this attribute for a variable-length relative record data set (VRRDS).
Nonspanned: Specifies that a record must be contained in one control interval. VSAM selects a control interval size that accommodates the largest record in the data set. This is the default.

System Managed Buffering

Specifies the amount of virtual storage for SMB Direct Access Bias obtained for buffers when opening the data set. The possible values are:

1K to 2048000K
1M to 2048M

EATTR

A data set level attribute specifying whether a data set can have extended attributes (format 8 and 9 DSCBs) and optionally reside in EAS.

NO: No extended attributes. The data set cannot have extended attributes (format 8 and 9 DSCBs) and cannot reside in EAS. This is the default behavior for non-VSAM data sets.
OPT: Extended attributes are optional. The data set can have extended attributes (format 8 and 9 DSCBs) and can optionally reside in EAS. This is the default behavior for VSAM data sets.