|
|
|
The following describes detail and performance data returned for non-VSAM and VSAM data sets.
The following panel shows an example of the Data Set Details panel containing data for a non-VSAM data set.
CA MAT --------------------- Data Set Details ------------ Row 1 to 19 of 64 COMMAND ===> SCROLL ===> CSR DDNAME: MDYU3000 Profile: DB2SAMP Access Method: VSAM Cluster: CAMAT.TUNCIC61.MDYUDATA.CC3000 Opened for: DIRECT Data: CAMAT.TUNCIC61.MDYUDATA.CC3000.DATA Share Options: 2 3 Index: CAMAT.TUNCIC61.MDYUDATA.CC3000.INDEX Organization: KSDS Data Component Information for: CAMAT.TUNCIC61.MDYUDATA.CC3000.DATA ------ SMS ------- ------ Allocation ------- ----------- Format --------------- DATACLAS: **NONE** Unit: CYL Rec Len: 43 CI Size: 4096 STORCLAS: **NONE** Primary: 5 # CIs per CA: 180 Strings: 0 MGMTCLAS: **NONE** Secondary: 2 CA Freespace: 0 Buffers: 2 Volume Count: 1 CI Freespace: 0 Total Extent Count: 1 Stripes: 0 ---- VSAM Statistics ---------------------------------------------------------- Number of records: 0 Records Read: 0 Records Updated: 0 CA Splits: 0 Records Added: 0 Records Deleted: 0 CI Splits: 0
For field related information for the Allocation and Format sections, see the previous section. Field descriptions for the Data Set Performance fields section for non-VSAM data set are described next.
For Systems Managed data sets, this class is a named collection of data set and space attributes.
For Systems Managed data sets, this class is a named collection of performance and device availability attributes.
For Systems Managed data sets, this class is a named collection of migration, backup, and expiration attributes.
Average data set response time during the monitoring interval.
Average number of I/Os per second being performed against the file during the monitoring interval.
Total number of EXCPs (I/Os) performed against the file during the monitoring period.
Total data set connect time during the monitoring interval.
This value is expressed in 128 milliseconds units and includes data transfer time, connect search time, protocol, and reconnect time.
Average IOS Queue Time experienced by the data set during the monitoring interval expressed in ms units. IOSQ TIME is computed by using the formula:
(QC / S)
Average IOS = ----------
queue time R
where
Average number of milliseconds that an I/O request remains queued to the channel.
This value reflects the amount of time that elapses between when the SSCH (Start SubChannel) function is accepted by the subchannel and when the first command that is associated with the SSCH function is accepted at the device.
Pending time accrues when the request is waiting for an available channel path or control unit, as well as when shared DASD contention occurs.
Average number of milliseconds that the device had an active I/O request but was disconnected from the channel path.
Disconnect time includes
Average number of milliseconds the device was connected to a channel path and actively transferring data between the device and central storage.
This value is calculated as follows:
Average Connect time = Total connect time / Measurement Even count
An example of the Data Set Details panel containing data for a VSAM data set follows.
CA MAT --------------------- Data Set Details ------------ Row 1 to 19 of 35 COMMAND ===> SCROLL ===> CSR DDNAME: NSUSSN3 Profile: TUNSAMP Access Method: VSAM Cluster: TAC.H159NPS.SESS.MGR.SESSN3.CLUS Opened for: DIRECT Data: TAC.H159NPS.SESS.MGR.SESSN3.DATA Share Options: 2 3 Index: N/A Organization: ESDS Data Component Information for: TAC.H159NPS.SESS.MGR.SESSN3.DATA ------ SMS ------- ------ Allocation ------- ----------- Format --------------- DATACLAS: STANDARD Unit: CYL Rec Len: 32752 CI Size: 32768 STORCLAS: FAST Primary: 800 # CIs per CA: 22 Strings: 0 MGMTCLAS: MA@NODEL Secondary: 20 CA Freespace: 0 Buffers: 16 Volume Count: 1 CI Freespace: 0 Total Extent Count: 29 Stripes: 0 ---- VSAM Statistics ---------------------------------------------------------- Number of records: 28776 Records Read: 1895 Records Updated: 835 CA Splits: 0 Records Added: 0 Records Deleted: 0 CI Splits: 0 ---- Data Set Performance --------------------- Avg Resp Time: 4.39 Avg IOSQ Time: 0.00 Active Rate: 0.68 Avg Pend Time: 0.29 Total EXCPs: 2727 Avg Disc Time: 1.51 Total Conn Time: 7062.53 Avg Conn Time: 2.59
The field descriptions for the Data Set Performance fields section for VSAM data set are described next.
Identifies the access method used to read from or write to the data set.
Intention of the program when performing I/Os for this data set. Valid values are:
Displays two numeric values.
The first value is the level (1, 2, 3, or 4) of cross-region sharing allowed by VSAM. The second value is the level (3 or 4 only) of cross-system sharing allowed by VSAM.
Physical organization of the data stored in this data set.
Can be one of the following types:
Full name of the data set.
If this organization is KSDS, CA Mainframe Application Tuner displays the names of the DATA and INDEX components.
Is the data set name of the DATA component of the VSAM file.
Is the data set name of the INDEX component of the VSAM file.
VSAM KSDS data sets are indexed using a multilevel structure. The number of levels can have an impact on the performance of the data set if there are insufficient index buffers.
Specified length of the records that are residing in this data set.
Count of CIs allocated within a control area for the data in this data set.
Percentage of free space in the control area.
Defined when the data set was created.
Percentage of free space in the control interval.
Defined when the data set was created.
For extended sequential data sets, the number of stripes that are used to read/write the data.
This number will be nonzero if the extended sequential data set is allocated with multiple stripes.
Size of the control interval for the data in this data set.
Number of strings for the data set.
This value indicates the number of concurrent operations that might be started for the data set.
Total number of records that were retrieved during the monitored period.
Total number of records that were added during the monitored period.
Total number of records that were written during the monitored period.
Total number of records that were deleted during the monitored period.
Total number of control area splits that have occurred since the file was defined.
Total number of control interval splits that have occurred since the file was defined.
Average data set response time during the monitoring interval.
Average number of I/Os per second being performed against the file during the monitoring interval.
Total number of EXCPs (I/Os) being performed against the file during the monitoring period.
Total data set connect time during the monitoring interval.
This value is expressed in 128 milliseconds units. Includes data transfer time, connect search time, protocol, and reconnect time.
Average IOS Queue Time that was experienced by the data set during the monitoring interval expressed in ms units. IOSQ TIME is computed by using the formula:
(QC / S)
Average IOS = ----------
queue time R
where
Average number of milliseconds that an I/O request remains queued to the channel.
This value reflects the amount of time that elapses between when the SSCH (Start SubChannel) function is accepted by the subchannel and when the first command associated with the SSCH function is accepted at the device. Pending time accrues when the request is waiting for an available channel path or control unit, as well as when shared DASD contention occurs.
Average number of milliseconds that the device had an active I/O request but was disconnected from the channel path.
Disconnect time includes
Average number of milliseconds that the device was connected to a channel path and actively transferring data between the device and central storage.
This value is calculated as follows:
Average Connect time = Total connect time / Measurement Even count.
Pool identifier that was used to satisfy requests for this file.
Size of the buffer that was used from the pool identified by ‘Pool ID’.
Percentage of READ requests that were satisfied from the buffer pool without an EXCP.
Number of buffers that are available in the shared pool.
Number of non–user-initiated WRITES that were initiated for this buffer pool.
A value here usually indicates a serious performance consideration. (LSR/GSR).
For either a non-VSAM or VSAM data set, you can scroll down to see additional volume information:
Volume Information for Volume PUBA01 ---------------------------------- Unit No.: 8359 Avg Resp Time: 40.41 Avg IOSQ Time: 086.30 Dev Type: 3390-380 Active Rate: 239.66 Avg Pend Time: 23.71 Alloc Cnt: 12 Dev Res: 0.03 Avg Disc Time: 1.61 Open Cnt: 1 Dev Util: 400.01 Avg Conn Time: 15.08 Mounted: PUBLIC Cache Act.: CFW CAC DFW
The field descriptions for the Data Set Performance fields section for VSAM data sets are described next.
These statistics are for the entire volume and not just for the data set. This way, you can use CA Mainframe Application Tuner to determine reasons for data sets not performing correctly and compare the performance of the data set with the entire volume.
Volume serial name
UCB Unit number (address)
Device type
Number maximum number of users allocated to the volume during the monitoring period.
Maximum number of open data sets on the volume during the monitoring period.
The open count directly influences the IOS Queue time to the device.
Method that the volume was mounted to z/OS. Can be PRIVATE, PUBLIC, or STORAGE. This value is used to determine how data sets are allocated on the device.
Indicates active caching options for the device. If present, the option is active. The options are as follows:
These options can greatly affect the performance of the volume.
Average data set response time during the monitoring interval.
This value is in milliseconds units.
Average number of I/Os per second being performed against the volume during the monitoring interval.
Percentage of time the device was found to be reserved during the monitored period.
Percentage the device was in use during the monitoring period.
This value includes both the time the device was performing I/O operations and the time that the device was reserved, but not necessarily active. The calculation is
Device Utilization Percentage = 100 * [(DCT +TDT / T) + (RS / S)]
where
Average IOS Queue Time experienced by the data set during the monitoring interval expressed in ms units. IOSQ TIME is computed by using the formula:
(QC / S)
Average IOS = ----------
queue time R
where
Average number of milliseconds that an I/O request remains queued to the channel.
This value reflects the amount of time that elapses between when the SSCH (Start SubChannel) function is accepted by the subchannel and when the first command that was associated with the SSCH function is accepted at the device. Pending time accrues when the request is waiting for an available channel path or control unit, as well as when shared DASD contention occurs.
Average number of milliseconds that the device had an active I/O request but was disconnected from the channel path.
Disconnect time includes
Average number of milliseconds the device was connected to a channel path and actively transferring data between the device and central storage.
This value is calculated as follows:
Average Connect time = Total connect time / Measurement Even count
| Copyright © 2011 CA. All rights reserved. | Tell Technical Publications how we can improve this information |