2. PERFORMANCE REPORTING ANALYSIS › 2.4 I/O Configuration Analysis › 2.4.7 Tape Drive Simulation Analysis Inquiry › 2.4.7.2 Technique Tutorial › 2.4.7.2.2 Computational Methods Available
2.4.7.2.2 Computational Methods Available
The inquiry obtains data from the DETAIL levels of the BATPGM
File (for batch jobs) and from the BAT_ST File (for started
tasks). Optionally, TSO work from the optional BAT_TS file
or APPC information from the BAT_TP file can be included if
relevant to the study. Since the number of tape drives
allocated or reserved/allocated to a particular job can
change from one job step or step interval to another, and
since the amount of time during which each particular step
has exclusive use of tape drives must be known for this type
of analysis, the inquiry requires this detailed level of
data.
Using this data as input, the inquiry then performs a
"stepped" computation of the number of drives allocated or
reserved/allocated during each time interval analyzed. The
number of drives allocated during the interval represents the
number of drives allocated from the step start to step end
and under MVS control. The number of drives
reserved/allocated represents the number allocated (maximum
of JES3 MDS high water mark reservations (HIGHWATR) or the
MVS allocated number (PGMMXNTA)) from step start to step end.
You can use either one of two methods to perform this
analysis:
o Snapshot Method
o Prorating Method
SNAPSHOT METHOD
Using the snapshot method, if, for example, the interval
clock value of the inquiry is set to two minutes, then the
inquiry determines the number of allocated tape drives at
each two-minute interval from the beginning to the end of the
analysis period specified. In this way, it calculates
snapshots of the allocation status of the tape drive
subsystem.
This method has the advantage of providing an exact sample of
tape drive activity at the time of each sample interval. Its
disadvantage is that it misses any job step which starts
after one sample interval and completes prior to the next
sample interval.
To minimize this potential effect of overlooking jobs steps
shorter than the sampling interval, the inquiry does not
permit this method to use a sampling interval longer than two
minutes. If you do not specify a sample interval, the
default time interval used is one minute. Analyses of a
variety of data have shown that one minute sample intervals
are usually satisfactory for this method of analysis.
If you wish, you can specify sampling intervals shorter than
one minute. Although shorter sampling intervals improve the
accuracy of this method slightly, they also lead to greater
use of resources.
Reports and charts produced under the snapshot method
represent intervals within an hour.
PRORATING METHOD
The prorating method does not overlook tape drive utilization
by even very short job steps. It also credits each job step
which begins or ends part way through a sampling interval
with partial use of any tape drives that it has allocated
during this interval. This method provides an accurate
calculation of the average number of drives allocated during
any size sampling interval.
It is important to remember, when using this method, that
excessively long intervals tend to mask the real-world impact
of short-term peaks in tape drive utilization. This effect
is characteristic of any analysis based upon averages, and is
not unique to the inquiry.
You need to determine what the maximum useful sampling
interval is, based upon your understanding of your
installation's workload mix and operational objectives.
Analyzing the same data using several different sampling
intervals may help you determine what is the largest sampling
interval which still retains sufficient detail about peaks
and valleys in the allocation pattern for tape drives.
Reports and charts produced under this method are displayed
by day if the specified interval (INTV) is greater than 60
minutes; otherwise they are displayed by hour within day.
