Data Set Characteristics Supplement report
The Data Set Characteristics Supplement report provides additional information on data sets. This report includes an entry for a data set if Strobe detected activity in the data set or computed a change in EXCP counts for it. If the EXCP count changed but Strobe did not detect activity in the data set, the report shows only the ddname and the EXCP count, and other I/O reports in the Performance Profile have no entry for the data set.
You can cross-reference this report with the Data Set Characteristics report by data definition name, access method, and data set name. If no information exists, Strobe suppresses the names of the fields. You can suppress the report by supplying the NODSCS parameter.
The report provides supplementary information in five sections:
- Identifying information
- File creation data
- System Managed Storage (SMS) data (if SMS manages the data set)
- VSAM data (if the data set was created through VSAM)
- Other information.
Data Set Characteristics Supplement Report
Identifying Information
This section of the report supplies the following information about the data sets.
DDNAME
The data definition name as it appears in the Data Set Characteristics report.
ACCESS METHOD
The access method as it appears in the Data Set Characteristics report.
DSNAME
The data set name as it appears in the Data Set Characteristics report. For HFS files that show *PATHNAME on the DATA SET NAME field of the Data Set Characteristics report, this field displays the full pathname as it appears on the SMF record.
OPEN INTENT\PROCESSING MODE
Describes parameters the program supplied when it opened the data set. This field can help you to understand better how the measured program accesses or processes the data set.
The following are the non-VSAM values:
EXTEND | The data set is treated as an OUTPUT data set, but records are added to the end of the data set regardless of what was specified on the DISP parameter of the dd statement. |
INOUT | The program accesses this data set first for input and later, without reopening the file, for output. The data set is processed as INPUT if it is a SYSIN data set or a PDSE or LABEL=(,,IN) is specified in the dd statement. |
INPUT | The data set is opened for INPUT. |
OUTIN | The program accesses this data set first for output and later, without reopening the file, for input. |
OUTINX | This data set is treated as an OUTIN data set, except that records are added to the end of the data set regardless of what was specified on the DISP parameter of the dd statement. For PDSEs, OUTINX is equivalent to OUTPUT. |
OUTPUT | The data set is opened for OUTPUT. For BDAM, OUTPUT is equivalent to UPDAT. |
RDBACK | This data set is an input data set positioned to read backward. This option is not allowed for DASD data sets; it is supported only for magnetic tape. |
UPDAT | This data set is to be updated in place or, for BDAM, blocks are to be updated or added. |
The following are the VSAM values:
DIR | This data set is accessed with direct access. |
INPUT | This data set is used for retrieval of records. (The ACB specifies the IN MACRF option.) |
DYN | This data set’s index component is processed randomly, which most frequently indicates that the COBOL program specified “ACCESS IS DYNAMIC”. For programs written in languages other than COBOL, the DYN parameter indicates that Strobe observed a VSAM POINT request. |
I-O | This data set is for input and output access. (The ACB specifies both IN and OUT MACRF value.) |
LOAD | Records are written to the data set in load mode. |
OUTPUT | This data set is used for storage of new records. (The ACB specifies the OUT MACRF value.) VSAM access allows both input and output operations against the file. |
SEQ | This data set is accessed with sequential access. |
SKP | This data set is accessed with skip-sequential access. |
FILE
Identifies some basic data set characteristics.
RECFM
This field provides and displays record and block format specifications.
The record formats are:
F | Fixed length |
U | Undefined |
V | Variable length |
The block formats are:
B | Blocked |
S | Spanned (variable length) or standard (fixed length) |
For print data sets, the following codes apply:
A | ANSI control characters |
M | Machine control characters |
EXTENTS
Displays the maximum number of extents allocated to the data set. Strobe reports on extents for concatenated data sets in this way:
- For concatenated sequential data sets, Strobe reports on extents for individual data sets in the concatenation. Strobe reports these extents only if it detects I/O activity for the data set.
- For concatenated partitioned data sets, Strobe reports extents for individual data sets in the concatenation.
EXTNDED FMT
Displays the presence of DFSMS extended format data sets and the attributes assigned to the data sets. The following values may appear in this field:
EA | The data set is defined with the extended addressability attribute. If it is followed by XRBA, this label indicates that the data set was accessed with the XRBA option. |
COMP | The data set is eligible for SMS compression. It may be followed by:
|
OTHER
Various file characteristics, such as CNV (control interval processing), ICI (improved control interval processing), UBF (user buffering), VLF (virtual lookaside facility), and VVDS (VSAM Volume Data Set).
SMS Data
System Managed Storage (SMS) data for the data set. The SMS option enables you to identify ways to reclassify a data set to improve its performance.
For VSAM, Strobe may not always provide SMS information. For example, if Strobe could not determine the data set’s cluster name, Strobe cannot provide SMS information. If only the component name is available and the component name has a last qualifier .INDEX, .I, .DATA, or, .D, Strobe assumes that the file’s cluster name is the component name without the last qualifier and attempts to find information for the cluster name.
Strobe does not collect SMS information for alternate indexes. SMS will assign an alternate index the same management class and storage class as its associated base cluster.
DATA CLASS
The data class to which SMS assigns a data set. DATA CLASS contains a collection of allocation and space attributes for a group of data sets. The storage administrator defines the attributes.
MANAGEMENT CLASS
The management class to which a data set is assigned. MANAGEMENT CLASS displays a collection of management attributes that control:
- Release of allocated but unused space
- Retention, migration, and backup of data sets
- Retention and backup of aggregate groups
- Retention, backup, and class transition of objects.
The storage administrator defines these attributes.
STORAGE CLASS
The storage class to which a data set is assigned. STORAGE CLASS provides a named list of storage attributes that identify performance goals and availability requirements as defined by the storage administrator. It provides the criteria that SMS uses to determine an appropriate location for a data set or object.
MSR
The millisecond response (MSR) objective defined for this data set. MSR is defined for both direct and sequential processing modes and is reported as a three-digit value from 000 through 999.
Must Cache data sets are identified by MSR values that cannot be met without the use of cache. May Cache data sets are identified by a value of 000 (indicating that no MSR was specified) or a value that can be met without the use of cache. Never Cache data sets are identified by a value of 999.
BIAS
The BIAS value defined for the data set. BIAS is defined for both direct and sequential processing modes. The possible values for this field are READ, WRITE, or NONE.
STATS
Some response-time statistics for data sets. Values are in microseconds.
RESP
The average I/O response time. Examining this value, in conjunction with the value reported in the MSR field, provides a method of determining the cache usage of the data set. Generally, an average response time lower than 25 indicates that the data set was cached.
IOSQ
The average I/O Supervisor queue time. This value represents the average time spent waiting for service during an I/O operation while the device was busy with an operation from another system.
PEND
The average pending time. This value represents the average amount of time spent waiting to connect to a device.
DISC
The average disconnect time. This value represents the average time that the device on which the data set resides was logically disconnected from the channel subsystem. A low average disconnect time usually indicates that the data set has a good hit ratio.
VSAM Data
This section of the report appears if Strobe determines that the data set uses the Virtual Storage Access Method (VSAM) and can access the catalog information.
FREESPACE
The number of bytes of free space remaining in the component. When displayed in units of K (Kilobytes), M (Megabytes), and G (Gigabytes), this value will not be exact.
SHROPTS
The share options specified for the VSAM data set. This information may not be available for VSAM data sets processed under earlier versions of the Data Facilities Product (DFP).
CI/CA
The number of control intervals in a control area.
%CI FREE
The percentage of control interval set aside as free space when the cluster is initially loaded, during a mass insert. %CI FREE applies only to key-sequenced clusters and relative-record clusters with variable record lengths.
%CA FREE
The percentage of each control area set aside as free space when the cluster is initially loaded, during a mass insert. %CA FREE applies only to key-sequenced clusters and relative-record clusters with variable record lengths.
INDEX LVLS
The number of levels in the index component at the last time Strobe detected activity in the data set. More than three index levels indicates a possible problem with the VSAM cluster.
ATTR
The attributes of the VSAM definition parameter. The possible values are:
RCVY | The cluster was defined with the RECOVERY definition parameter, which causes VSAM to preformat each control area before data is written. |
SPEED | The cluster was defined with the SPEED attribute, which instructs VSAM not to preformat control areas. |
IMBD | The cluster was defined with the IMBED parameter, which specifies that the sequence-set record for each control area is written as many times as it will fit on the first track of the data control area. |
REPL | The cluster was defined with the REPLICATE parameter, which specifies that each index record is written on its own track as many times as it will fit. |
USER RECORDS
The total number of logical records in the data set the last time Strobe detected activity in the data set. If the total number of user records exceeds 999,999,999, the report displays the amount in units of G (billions of records).
LOGICAL OPERATIONS
The total number of the following types of VSAM operations:
- DELETES is the number of logical records that have been deleted during measurement.
- UPDATES is the number of logical records that have been updated and rewritten to the component during measurement.
- RETRIEVES is the number of logical records that have been retrieved from the component. This count includes “get for update” requests that occur via the REWRITE verb and the GET macro that specifies “for UPD”.
- INSERTS is the number of logical records that have been inserted during measurement. This count does not include ESDS or KSDS records that are loaded or added to the end of the data set. It also does not include records inserted during create (load) processing.
The report will not contain LOGICAL OPERATIONS values if the Performance Profile includes the CICS Performance Supplement.
BatchPipes Data
This section of the report appears if Strobe determines that a data set was processed by the BatchPipes subsystem.
WAIT COUNT
The number of times the job step was forced to wait because the pipe was either full (for a writer job) or empty (for a reader job).
WAIT TIME
The total amount of time spent waiting on a full pipe (for a writer job) or an empty pipe (for a reader job).
WAIT-FOR-OPEN TIME
The total amount of time spent waiting for a pipe partner (either a writer job or a reader job) to become active. If the value for this field is zero, this field does not appear.
PIPE PARTNER
The names of the jobs that have been designated as reader or writer partners for the measured job. The system name also appears if the pipe partner executed on another system in the BatchPipePlex (SmartBatch environments only).
This field displays a maximum of four pipe partners. The line ADDITIONAL PIPE PARTNERS EXIST appears if there are more than four pipe partners observed during the measurement session.
SUBSYSTEM
The system name, if the pipe partner executed on another system in the BatchPipePlex (SmartBatch environments only).
HFS/zFS Data
This section of the report appears if Strobe determines that the data set accessed is in the Hierarchical File System (HFS) or zFS File System. This section of the Data Set Characteristics Supplement report appears only if Strobe for UNIX System Services is installed at your site.
INODE
The file identifier.
DEVICE NUMBER
An identifier for the mounted file system. When a non-unique INODE is encountered within a measurement, the device number can be used further to identify the file. (An INODE is guaranteed to be unique only within an HFS.)
I/O BLOCKS
READ | The number of blocks read from the file during the measurement. |
WRITTEN | The number of blocks written to the file during the measurement. |
FILE TYPE
The definition of the file. This field can have one of two values: REGFILE for a normal file or FIFO for a named pipe.
BYTES
READ | The number of bytes read from the file during Strobe measurement. |
WRITTEN | The number of bytes written to the file during Strobe measurement. |
Strobe reports the I/O blocks and bytes statistics only if System Management Facility (SMF) is configured to produce type 92 records. For more information, refer to the Installing.