2. USAGE GUIDELINES › 2.1 Data Analysis

2.1 Data Analysis

Several methods of operational and workload analysis using
the CA MICS Batch and Operations Analyzer are presented in
this section.


ADMINISTRATION

The CA MICS Batch and Operations Analyzer provides
performance, throughput, and availability information that is
vital in establishing service level goals and agreements.  In
addition, the product can be valuable in determining the
magnitude of service degradation.  Unauthorized use of the
computer and other security violations can also be identified
with product data.


RESOURCE ACCOUNTING AND CHARGEBACK

By providing input data to the CA MICS Accounting and
Chargeback Product or user-developed accounting systems, the
CA MICS Batch and Operations Analyzer allows for cost
recovery of computer resources.  In addition to its interface
with the CA MICS accounting product, batch and operations
data, including elapsed time, CPU time, memory utilization,
occupancy time, and I/O device utilization, can be integrated
into any accounting algorithm.


REPORTING BATCH JOB SERVICE

Batch job groups are a CA MICS structure that allows you to
analyze batch job service and performance with respect to the
service requested and received.  You can use this
installation-defined information to determine if batch job
service levels have been missed, met, or exceeded.


TREND ANALYSIS

You can identify trends and enforce standards by comparing
measurements taken in the current month with similar
measurements for the previous month.  Some uses of this
technique include analyzing CPU time, I/O counts, and service
units.


HARDWARE CONFIGURATION ANALYSIS

You can use the CA MICS Batch and Operations Analyzer to
support utilization studies for measuring hardware capacity
or to determine if the configuration is adequate for the
needs of the enterprise.  If the utilization is inappropriate
for the users' needs, appropriate budgetary actions can be
taken.

Such an analysis performed daily or weekly will indicate dead
hardware on the system and in addition will provide the
necessary information for operations to properly distribute
the workload over each of the devices.  This will minimize
the under-utilization or over-utilization of devices.


APPLICATION PROGRAM ANALYSIS

The CA MICS Batch and Operations Analyzer can support a
performance study of the programs that are used most
frequently to analyze their impact on the total workload.
Once these programs are identified, they may be streamlined
to reduce program run times and overall data center workload.


PROGRAM CONVENTION ANALYSIS

Data from the CA MICS Batch and Operations Analyzer can
assist in performance studies to identify programming
practices that result in wasted or poorly used computer
resources, with the objective being a reduction in total
required computer time for the workload being studied.  Such
a study can identify memory and devices allocated but not
used, tape devices used temporarily with extremely low
volume, direct access space allocated but not used, data sets
on direct access using secondary allocation, the profile of
usage patterns for online systems, and abnormal
terminations.


BOTTLENECK ANALYSIS

The CA MICS Batch and Operations Analyzer can be a valuable
tool in bottleneck studies to identify problems that result
in poor usage of the data center's capacity.  The study's
results can be used to maximize system usage and balance the
distribution of the workload over the available resources.

The product can support the study of usage trends for average
job turnaround time, average job costs, total costs
recovered, number of unsuccessful runs, number of jobs
submitted, memory utilization, direct access device
utilization, magnetic tape drive utilization, spooling
volume, and operator mount activity.


OPERATIONAL PROCEDURE ANALYSIS

Using batch and operations data, your installation can
identify operating practices that result in wasted resources
or poor service.  The product can be used to identify
resources lost as a result of system failures, jobs sitting
in a wait status within the system, periods of the day when
memory and devices are at low utilization levels, abends that
are caused by the operations staff, job failures because of
resource contention, and jobs being scheduled into the wrong
partitions or regions.


APPLICATION SYSTEM ANALYSIS

The CA MICS Batch and Operations Analyzer can assist in a
performance study to identify programs whose system design,
data flow, and database handling results in a wasteful use
of computing system power.  By first identifying the most
resource-intensive application systems and processing steps,
it can target those processes that merit further study.
Spending the time to streamline the most resource-intensive
jobs will produce the most benefit to the data center.


INSTALLATION ENVIRONMENT

The activities of installation environment include the
monitoring of the I/S production environment to maintain the
operational readiness of hardware, software, communications
network, and physical facilities.  The percentage of system
abends, the number of system IPLs in a specified time period,
the number of lines written to spool, and TSO messages
processed indicate operational readiness.  The product can
provide performance information by assessing, for example,
average CPU time per batch job.


DEVELOPMENT

Development activities pertain to the creation, procurement,
and modification of applications, hardware, networks,
software, management systems, facilities, and other
resources.  The CA MICS Batch and Operations Analyzer files
can aid project management by providing assessments of
resource usage for anticipated development work.  This may
include counts on the number of jobs submitted for a
development project, the usage of a specific piece of
software, or the error rates encountered.


DISTRIBUTED PROCESSING

APPC/MVS allows transaction programs (TPs) to establish
conversations with other TPs in both z/OS and non-z/OS
environments.  The CA MICS Batch and Operations Analyzer
files can be used to track the frequency and resource
utilization of any partner TPs executing under z/OS.


PRINTER UTILIZATION ANALYSIS

By analyzing the amount of time the writer was actively
processing a print output data set, you can study the
components of job turnaround time and estimate printer
utilization.