Parameters to create HIS CSV report

 

REXX program S113CSV creates the result HIS CSV reports. The program uses a set of parameters which are contained in <UID>.S113.JCLPRMS(HISREP). There are 3 Basic parameters:

Control Statement

Description

DEBUG,YES

Print debug information in Batch output

CALC,<rexx control statement>

The REXX control statement is interpreted by the interpret command of the REXX interpreter

CALC,$BEGIN

     <rexx control statement>

     <rexx control statement>

     . . .

$END

It is possible to enter a series of consecutive rexx control statements between the two keywords $BEGIN and $END

PRINT,<column title>,<length of title>,<rexx control statement>

Prints the columns. The sequence determines the columns in the output table.

<rexx control statement> can be a variable or a valid rexx statement containing variables.

 

All column headers in the resulting HISDATA file or summary of the HISDATA file (same headers) can be used as variables in the <rexx control statements> be using a # as prefix. It is also highly recommended to always use a # as prefix for all variables because this assures that no unwanted side effects occur with the REXX program. The REXX program S113CSV does not use any variables prefixed with a #.

When you look at the HISREP parameter set you can observe that the control statements calculate the result metrics from the input data by using the header titles as variables. The main formulas are described below.

Please note: For consistency reasons is the same notation used for z10 as for the newer systems. The L1.5 cache of the z10 is named L2 cache and the L2 cache L4 cache. The z10 doesn’t have a L3 cache.

Metric

z10

z196

zEC12

z13

CPI

#B00 / #B01

IPCG

Speed / CPI

#PRBS

(#P33 / #B01) * 100

#SUPS

(#B01 - #P33) / #B01 * 100

#LCPU

#B00 / #SPEED / #INTVAL / #PRC# / 10000  (#INTVAL=1 for current report; #PRC# = number of LCPs per processor type)

#B2B4

#B02 + #B04

#L1MP

#B2B4 / #B01 * 100

#L2Src

#E128 + #E129

#E128 + #E129

#E130 + #E131+ #E132

#E133 + #E136

#L2P

#L2Src / #B2B4 *100

#L3Src

N/A

#E150 + #E153

#E144+#E150+#E153+#E159

#E144+#E145+#E162+#E163

#L3P

N/A

L3Src / #B2B4 * 100

#L4Src

#E130 + #E131

#E135+#E136+#E152+#E155

#E145 + #E147 + #E151 + #E154 + #E156 + #E160

#E146 + #E147 + #E148 + #E164 + #E165 + #E166

#L4LP

#L4LSrc / #B2B4 *100

#L4RSrc

#E132 + #E133

#E138 + #E139 ) + #E134 + #E143

#E146 + #E148 + #E152 + #E155 + #E157 + #E161

 #E149 + #E150 + #E151 + #E152 + #E153 + #E154 + #E155 + #E156 + #E157 + #E167 + #E168 + #E169 + #E170 + #E171 + #E172 + #E173 + #E174 + #E175

#L4RP

L4RSrc / #B2B4 * 100

#MEMSrc

MemSrc = #B2B4 - #L2Src -#L4LSrc - #L4RSrc

MemSrc = #B2B4 - #L2Src -#L3Src - #L4LSrc - #L4RSrc

#MEMP

MemSrc / #B2B4 * 100

 

The next step is the Relative Nest Intensity which describes how expensive it is to obtain data or instructions relative to accessing the data from the L2 cache.

 

System

Relative Nest Intensity (#RNI)

z10

(1.0 x #L4LP + 2.4 x #L4RP + 7.5 x #MEMP) / 100

z196

1.67 x ( 0.4 x #L3P + 1.0 x #L4LP + 2.4 x #L4RP + 7.5 x #MEMP ) / 100

zEC12

2.3 x (0.4 x #L3P + 1.2 x #L4LP + 2.7 x #L4RP +8.2 x #MEMP) / 100

z13

2.6 x (0.4 x #L3P + 1.6 x #L4LP + 3.5 x #L4RP + 7.5 x #MEMP) / 100

 

And finally some additional metrics:

 

Metric

z10

z196

zEC12

z13

#B2B4

#B02 + #B04

#B3B5

#B03 + #B05

#EFCPI

#B3B5 / #B01 * 0.84

#B3B5 / #B01 * (0.59 + (0.1 * #RNI))

#B3B5 / #B01 * (0.54 + (0.04 * #RNI))

#E143 / #B01

#EICPI

#CPI - #EFCPI

#ESRC

#B3B5 / #B2B4 * 0.84

#B3B5 / #B2B4 * (0.59 + (0.1 * #RNI))

#B3B5 / #B2B4 * (0.54 + (0.04 * #RNI))

#E143 / #B2B4

#TMP

#E145 + #E146

#E130 + #E131

#E128 + #E129

#E143 / #B3B5

#TM2

#E138 + #E139

#E144 + #E145

#E133 + #E140

#E129 + #E134

#TLB1

#TMP / #B00 * 100 * 0.31

#TMP / #B00 * 100 * 0.61

#TMP / #B00 * 100 * 0.65

#TMP * (#E130+#E135) / #B00 * 100

#TLBC

#TMP / #TM2 * 0.31

#TMP / #TM2 * 0.61

#TMP / #TM2 * 0.65

#TMP * (#E130+#E135) / #TM2

#PTE

#E140 / # TM2 * 100

#E141 / #TM2 * 100

#E137 / #TM2 * 100

#CPEN

#B3B5 / #B2B4