Identify and report on only those programs using at least ten different I/O devices for more than one hour of elapsed time.
Build GROUP and GROUPC statements as follows:
position 1 2 3 4 5 8
1........0.........0.........0.........0.........0.. ...0
points to Elapsed Time
in Basic Accounting Table
│
│ points to # of Devices Used
│ in Basic Accounting Table
│ │
│ │ ┌─ compared field is in
▼ ▼ ▼ packed decimal format
┌──┐ ┌──┐
DGROUP 1713XR3632PR
▲ ▲
compared field is └── rejection indicator
in hexadecimal format
zeros in integer portion of Elapsed Time
▼
┌────┐
DGROUPC 1000000
DGROUPC 2000 009
--- ---
▲ ▲
zero to nine devices
Example 4 shows you how to employ the format indicator while using the Grouping feature. The first definition on the GROUP statement points to the elapsed time field in the Basic Accounting Table. The test is to be made against the first three bytes of this field in hexadecimal format represented by six digits. Therefore, the GROUPC statement which corresponds to the first definition must contain an upper limit consisting of six EBCDIC characters. CA JARS automatically converts the six characters into the necessary hexadecimal format for the compare. Any records matching the compare (000000 hours) are rejected as requested per the rejection indicator.
Selected records pass to the next test as defined by the second definition on the GROUP statement. This test is against the field containing the number of I/O devices used. The two-byte integer value is in packed decimal format. Therefore, the GROUPC statement which corresponds to the second definition must contain a lower limit consisting of three EBCDIC numbers, left-justified (000 followed by 5 blanks) and an upper limit of 009 followed by 5 blanks.
CA JARS automatically converts the three-digit numbers to two-byte packed decimal fields for the comparison. Any records within the range (from 0 to 9 I/O devices) are rejected as requested by the rejection indicator (value R). All unidentified records (no match on either test) are selected for reporting.
In the case of packed decimal compares, note that a logical comparison is made, eliminating the possibility of a data exception (S0C7). Also, the sign portions of the converted fields are adjusted so that the sign of the data field (usually positive) has no effect on the result of the comparison; that is, the compare is, in effect, against the absolute value of the data field. If it is desired to specifically identify positive or negative packed fields, the hexadecimal format, which is converted exactly as specified, should be used.
| Copyright © 2012 CA. All rights reserved. |
|