|
|
|
The Monitor OverView report presents a high-level view of the application you are monitoring, gives you a preliminary indication of job performance, and helps you determine if the sample is valid for the unit of work you have targeted. See "Samples and observations" and "Sampling rules" for more information about the collected data.
Check the following fields to determine if the monitor is valid.
Number of samples should be enough to ensure accurate data.
The number should be greater than 1000. A number closer to 10000 is preferred.
Name of the program that you intended to monitor displays.
Number of recoverable errors is less than one percent of the total number of samples.
For message TN8091I, ensure that samples were collected; for example, ensure the monitor target address space was not swapped or non-dispatchable each time the DIE routine was executed.
Message TN8091I indicates:
The fields of the Monitor Overview panel are described next.
Name of the data set containing the sample data from the monitor.
Name of the profile that you specified when you created the monitor definition.
Type of analysis in effect: NORMAL, ALL, ACTIVE, or WAIT.
An asterisk (*) following the value indicates that tasks or transactions have been selected for analysis. Place the cursor on the Options field value and press Enter to display a pop-up panel listing the selected tasks/transactions.
Name of the job that was monitored.
Step name on the EXEC statement that invoked the program (if no cataloged procedure is used).
Name of the step that invoked the procedure (if a cataloged procedure is used).
Procedure step name (if a cataloged procedure is used).
Name of the program on the EXEC statement of the JCL for the job.
Address space identifier of the job that was monitored.
Identifier of the user who requested the monitor.
If the monitor was invoked by a batch job using TUNCALL, this field displays the name of the batch job.
JES job identifier of the job that was monitored.
The 3-digit system and 4-digit user completion codes for the job being monitored.
Release level of the CICS system that was used by the job.
This field is not displayed unless CICS was detected for the job being monitored.
Release level of the DB2 system that was used by the job.
This field is not displayed unless DB2 was detected for the job being monitored.
Release level of the IMS system that was used by the job.
This field is not displayed unless IMS was detected for the job being monitored.
Release level of the MQ Series (MQS) system that was used by the job.
This field is not displayed unless MQS was detected for the job being monitored.
Release level of the SAP® system that was used by the job.
This field is not displayed unless SAP was detected for the job being monitored.
Release level of the UNIX® System Services (USS) system that was used by the job.
This field is not displayed unless USS was detected for the job being monitored.
Release level of the WebSphere Application Server (WAS) system that was used by the job.
This field is not displayed unless WAS was detected for the job being monitored.
Release level of the CA IDMS system that was used by the job.
This field is not displayed unless CA IDMS was detected for the job being monitored.
Release level of the CA Datacom system that was used by the job. This field is not displayed unless CA Datacom was detected for the job being monitored.
Release level of the CA Ideal system that was used by the job. This field is not displayed unless CA Ideal was detected for the job being monitored.
Release level of the Adabas system that was used by the job. This field is not displayed unless Adabas was detected for the job being monitored.
Release level of the Natural system that was used by the job. This field is not displayed unless Natural was detected for the job being monitored.
Requested region size for the step being measured below the 16 MB line.
Requested region size for the step being measured above the 16 MB line.
Amount of storage requested by the job being monitored.
Eser-specified dynamic linklist, if using a linklist other than that specified during IPL.
Amount of CPU time that was used by execution of code in task mode during the monitored period.
This value is obtained from the ASCB.
Amount of SRB time used by this address space during the time it was being measured.
The SRB time used by CA Mainframe Application Tuner is not reflected in this value.
Amount of CPU time used within this address space in HH:MM:SS.TH format.
Amount of CPU time used by this address space in a zSeries Application Assist Processor (zAAP) in HH:MM:SS.TH format.
Total time that there was Java work eligible to run on a zSeries Application Assist Processor (zAAP).
Within a particular application, some work might be Java and the JVM has designated the work as zAAP-eligible, while other work might be non-Java, or might be Java work that was not marked zAAP-eligible by the JVM. The zAAP-eligible work can be switched to a zAAP for processing, and when the work that is not zAAP-eligible is ready to execute, this work would execute on a standard CP. A Switch Service is in place to work with the z/OS Dispatcher, managing the dispatching of zAAP-eligible work between the standard CPs and the zAAP Engines.
This field can help identify opportunities where a zAAP processor would benefit processing as well as identify issues with available zAAP processors not being fully utliized. New zAAP users should refer to the JVM Startup Options for zAAPs such as -Xifa: as well as new parameters in the IEAOPTxx of SYS1.PARMLIB such as IFACrossover.
Amount of CPU time that was spent actually executing on an Integrated Information Processor (zIIP) in HH:MM:SS.TH format.
Total time that there was work eligible to run on a zSeries Integrated Information Processor (zIIP) engine.
Note that only enclave SRB work is eligible to run on the zIIP. DB2 and IPSEC are two of the initial exploiters.
Within a particular application or subsystem, only the Enclave SRB work is designated as zIIP-eligible. A Switch Service is in place to work with the z/OS Dispatcher, managing the dispatching of zIIP-eligible work between the standard CPs and the zIIP engines.
This field can help identify opportunities where a zIIP processor would benefit processing, as well as identify issues with available zIIP processors not being fully utliized. There is one external tuning option, zIIPAWMT, which controls the needs help dispatcher.
Also note APAR OA20045 that enables the IIPHONORPRIORITY=Yes/No Specification for z/OS 1.8.
Amount of time the address space was observed to be swapped out during the measurement in HH:MM:SS.TH format.
Amount of time the address space was observed to be nondispatchable (the ASCB nondispatchability flags were on) during the measurement in HH:MM:SS.TH format.
Amount of time the address space was observed to be delayed because of either LPAR delay or other disabled work on the processor during the time of the measurement in HH:MM:SS.TH format.
Elapsed time that this monitor spent waiting.
Normally, this is the time that is not reflected in the CPU time, Swapped Out, Non Disp, and LPAR/DIS Delay fields.
CPU Service units, which are an interval of CPU time normalized by a CPU processor specific adjustment rate.
Number of EXCPs or I/Os issued by the address space during the duration of the measurement.
Number of EXCPs or I/Os per second being issued by the address space during the duration of the measurement.
High-water-mark value for the used region size below the 16 MB line.
This value might actually be larger than the requested size if modified by other products or SMF exits.
High-water-mark value for the used region size above the 16 MB line.
This value might actually be larger than the requested size if modified by other products or SMF exits.
Number of page-in operations for the job.
A high value in this field can indicate job delays.
Number of page-in operations per second.
A high rate can indicate job delay due to storage constraints.
When reference is made to a page of virtual storage that does not appear in central storage, the page is brought in from DASD or expanded storage to replace a page in central storage. The newly referenced page is said to have been paged in. It is the page-in (not page-out) rate that is of primary concern because page-in activity occurs synchronously (for example, the task stops until the page fault is resolved).
Date the monitor started.
Time the monitor started.
Duration of time the address space was measured by CA Mainframe Application Tuner for this monitor in HH:MM:SS format.
Final rate at which CA Mainframe Application Tuner observed the activity of the address space.
Note: This value was formerly known as the sampling rate. See "Specifying additional monitoring criteria" for more information.
Number of requested observations in the monitor profile as requested by the user.
Actual number of observations made by CA Mainframe Application Tuner.
Number of samples used from the measurement.
See "Specifying additional monitoring criteria" for more information.
Percentage of samples used that were recorded when the application was actively using CPU.
Percentage of samples used that were recorded when the application was not using CPU and was waiting on a resource.
Average number of TCBs that were actively using the CPU during an observation.
This value can be from 0 up to the number of online CPUs.
High-water-mark amount of common (ECSA) storage used by CA Mainframe Application Tuner to measure the job.
This value can be used for planning purposes on setting the maximum active number of monitors allowed.
Eight-character system name identifier for the z/OS image on which the monitor was executed.
Four-character System Management Facility identifier for the z/OS image on which the monitor was executed.
Release level of the MVS system where the target job executed.
Name of the operating system.
Release level of the operating system.
Name of the operating system owner.
CPU model number of the system on which the monitor was executed.
CPU version of the system on which the monitor was executed.
Serial number of the processor on which the monitor was executed.
Processor-specific adjustment rate used in the calculation of CPU service units.
This value is based on the number of instructions per CPU second for a given processor.
The Working Set Size Chart displays the average frame count usage over time. The total monitor period is divided into 50 equal intervals. Each + on the chart represents the average frame count for an interval.
The chart contains ten rows whose range is determined by the minimum and maximum average frame count used for the monitored period. The number of frames per row is calculated as
(Max Avg. Frame Count - Min Avg. Frame Count)/9
Approximate percentage of time the application spent either running on a CPU or being delayed for data over the life of the monitor run. The total monitor run is split into 50 equal periods, each period being 2% of the total run. The activity is summarized into each slot by resource type. In this way, you see an overview of the resource demands during the monitor run.
Task name, file name, or database ID that uniquely identifies the resource being reported.
Type of resource, which can be one of the following types:
Fifty columns of numbers and/or symbols depicting the time spent by the program either executing on a CPU or delayed for data for each period.
The values shown represent from 0% to 100% of the total resource demand. When the percentage of resource demand is 0%, a blank displays. When the percentage of resource demand is greater than 0% but less than 5%, a period (.) displays. When the percentage of resource demand is 5% or greater but less than 95%, a single digit from 1 to 9 displays. When the percentage of resource demand is 95% or greater, a plus sign (+) displays.
Messages that are issued by the Server Space during the monitored period are displayed here. These messages provide additional information about execution and completion.
Monitor options that you specified in the Monitor Criteria panel are displayed here.
Monitoring parameters specifying DB2 data collection information are displayed here. The values are defined in UTRPARM member TUNSSP00 (see the Installation Guide for more information about parameters in TUNSSP00).
---------------- DB2 Monitoring, Harvester, and Flag settings ----------------- DB2 Explain .... DB2EXPL = YES Call DB2 Explain for sampled SQL. DB2HEXPL = YES Call DB2 Explain for all Harvested SQL. DB2 Monitoring and Harvesting data granularity control .... DB2HVLOC = YES Add Requester Location to the summarization key. DB2HVCOR = YES Add Correlation ID to the summarization key. DB2HVIID = YES Add OPER ID to the summarization key. DB2HVTHD = NO Do not add Thread address to the summarization key. DB2 Harvester .... DB2HRVST = YES The DB2 Harvester is Active. DB2 Harvester summarization control .... DB2HVSUM = 0 DB2 Harvester data summarization is not Active. DB2 Harvester Restart/ABEND control .... DB2HVMXA = 1000 DB2 Harvester collection will end after 1000 ABENDs. DB2HVMXM = 100 DB2 Harvester ABEND messages will stop after 100 messages. DB2 data UNICODE control .... DB2TCSID = -1 To Coded Character Set IDentifier (CCSID). DB2 SQL collection control .... DB2CTSQL = YES Obtain SQL from the DB2 catalog. DB2HVSQL = YES Collect Harvested static and dynamic SQL. DB2HVDYN = NO This parameter is ignored when DB2HVSQL=YES. DB2SPSQL = YES Collect sampled static and dynamic SQL. DB2SPDYN = NO This parameter is ignored when DB2SPSQL=YES.
| Copyright © 2011 CA. All rights reserved. | Tell Technical Publications how we can improve this information |