* copyright (C) 1984-2019 merrill consultants dallas texas usa

Nov 21, 2006 the variable lists (SUM=, etc.) that VMXGSUM must parse.

Using the full line for your variables, up to 72 bytes,

will maximize the number of variables that will fit in

the 99*200=19800 bytes for each argument, enough for 2200

variables with 8-byte names.
Change 24.232 The last field in subtype 2, ACTRREPQ is only 45 bytes,

VMACENTX not the 48 bytes documented by the vendor.

Nov 20, 2006

Thanks to Chris Taylor, GMAC Insurance, USA.

Change 24.231 Label for variables SOV2WMNT was corrected to read:

VMACHPDM SOV2WMNT='ACCUM*DELAY*WAITING*FOR MOUNT'

Nov 18, 2006

Thanks to Tom Elbert, Assurant, USA.

Change 24.230 SAS procedures BLKCOPY & FCOPY (used during Installation

FORMATS and Service Pack Updates, and MIGRATE are now recognized

Nov 14, 2006 by the $MGSASPR format. Even though the official SAS doc

(only found deep in the SAS Support site) says you cannot

use PROC MIGRATE from SAS 6.09E to SAS 9, that particular

conversion IS supported, but only under SAS on z/OS.

Thanks to Jim Horne, Lowe's Companies, USA.
Change 24.229 The program worked fine if output to //WORK was used, but

TYPEIMS7 if you used either of the examples in the comments, to

Nov 9, 2006 send either IMS0708 or IMSSUMRY to the //IMSTRAN ddname,

that failed with ERROR 455-185 DATA SET WAS NOT SPECIFIED

ON DATA STATEMENT.

Thanks to Denise L. Jeffers, CIGNA, USA.

Change 24.228 Support for HyperPAV APAR OA12865.

VMAC74 -TYPE74 dataset: New variables created:

VMAC78 HYPERPAV='HYPERPAV*BASE*DEVICE?'

Nov 11, 2006 SMF74HPC='HYPERPAV*ALIASES*CONFIGURED*THIS LSS'

SMF74PSM='SUCCESSFUL*PAV*SAMPLE*COUNTS'

Variable SMF74TMS is deleted, as it never existed and was

input by MXG in error.

-TYPE78CU dataset: New variables added for each LCUID:

R783HCU ='HYPERPAV CU IDENTIFIER'

R783HNAI='TIMES I/O NO START*NO HYPERPAV*AVAILABLE'

R783HTIO='HYPERPAV I/O*REQUESTS*FOR THE LSS'

R783HAIU='HWM*IN-USE*HYPERPAV*ALIASES'

R783HCAD='HWM*ALIASES*IN USE*ONE BASE'

R783HIOQ='HWM*OF IO-S QUEUED'

Variable PCTALLBY set missing, per Change 19.203, instead

of generating a missing value note for each TYPE78CU obs.

Thanks to Dr. H. Pat Artis, Performance Associates, USA.
Change 24.227 Optional ESS GPARMKEY='4A'x caused INPUT STATEMENT

IMAC6ESS EXCEEDED LENGTH error if there were more than one

VMAC6 addressee. MXG now keeps four (ESSMACC1-ESSMACC4)

Nov 7, 2006 and alerts you if there were more, with a note.

Nov 8, 2006 -Support for ESS GPARMKEY='4C'x creates ESSMFROM variable

Nov 9, 2006 and for ESS GEPARMKY='42'x creates ESSOFSYF variable.

-Support for ESS GEPARMKY='2023'x creates ESSOUTBN.

-Length of DEPT, TITLE, BUILDING increased to $60.

Thanks to Dr. Alexander Raeder, ATOSORIGIN, GERMANY.
Change 24.226 Variable QWACWLME, Service Class Name, is now kept in the

VMACDB2 PDB.DB2ACCTP dataset for analysis. Note that APAR PK37312

Nov 7, 2006 corrects blank value in QWACWLME after zIIP maintenance.

Thanks to Chuck Hopf, Bank of America, USA.

Change 24.225 LPARs with no current share (no current weight points)

VMAC7072 had LPARSHAR=0 in PDB.TYPE70. Those LPARs are treated

Nov 7, 2006 now as Current Share = Initial Share.

Thanks to Rudolf Sauer, T-Systems Enterprise Services GmbH, GERMANY.

Change 24.224 Setting the Clock Back an hour, without quiescing for an

VMAC7072 hour, produces duplicate values of STARTIME in all RMF

VMXG70PR datasets, plus negative execution/elapsed/durations/etc

Nov 7, 2006 in many other records, and in many cases it is impossible

Nov 17, 2006 to even recognize the duplication/overlap.

However, for the 70 and 72 records, adding GMTOFFTM to BY

lists, after SMF70GIE, and to the KEEP= lists, appears to

prevent the duplicate STARTIME values from being summed

(which caused doubling of the values of PARTNCPU and

CPCMUS, among other errors), although duplicates will

still exist in these datasets.

Nov 17: Typo, APPCLAX in KEEP= in VMAC7072 should have

been APPCMAX, which caused APPCMAX to be not kept.

Jan 21: See Change 24.

Thanks to Rudolf Sauer, T-Systems Enterprise Services GmbH, GERMANY.

Thanks to Jim Horne, Lowe's Companies, USA.

Change 24.223 Total Virtual Storage Above the Bar was captured in the

VMAC78 VSDGxxxx variables in TYPE78VS dataset, but the Shared

Nov 7, 2006 bytes were not INPUT. And because IBM reused the VSDG

Dec 19, 2006 prefix for both sets of variables, this change changes

the variables VSDGxxxx to TOBYxxxx for the Total Byte

fields, and now creates SHBYxxxx variables for Shared

byte fields above the bar.

Dec 19: Corrected long line for R783HNAI input.

Thanks to Ralph C. Baechle, John Deere, USA.
Change 24.222 CICS Statistics variables DSGEJST and DSGSRBT were INPUT,

VMAC110 and correctly calculated/formatted, but were not KEPT in

Oct 24, 2006 the CICDS dataset.

Thanks to Helmut Rose, Com-Software, GERMANY.

Change 24.221 -Support for changed field lengths in SAR/EXD SMF type 6

IMACEXD optional data.

Oct 24, 2006

Thanks to Joe Kimberly, Kansas City Southern, USA.

Change 24.220 -Support for NTSMF OBJECT='DATABASE ==> INSTANCES creates

EXNTDATI new DATABASI dataset; previously, those objects were

IMACNTSM output into the DATABASE dataset, which caused nearly-

VMACNTSM duplicate observations.

VMXGINIT -Support for DATABASE object with NRDATA=191.

Oct 25, 2006

Thanks to Paul Billick, Harleysville Insurance, USA.
Change 24.219 The optional ARZGS GSACCT segment for CICS can have any

UTILEXCL length; MXG's INPUT statement expected 8, which caused

Oct 21, 2006 errors when the real length was 12. Now, UTILEXCL prints

the CMODLENG and a message to compare your actual length

with MXG's, and to change IMACICU2 if needed.

Thanks to Richard Hilber, Allgemeines Rechenzentrum GmbH, AUSTRIA.

Thanks to Peter Gschirr, Allgemeines Rechenzentrum GmbH, AUSTRIA.
Change 24.218 Support for NTSM Beta Version 3.0.0.8 (COMPATIBLE), adds

VMACNTSM SUMRYINT and ORGANIZN to NTCONFIG dataset.

Oct 18, 2006
Change 24.217 If you want to process on the CICSTRAN data to create the

ASUMUOW PDB.ASUMUOW dataset without reading DB2ACCT, when there

Oct 18, 2006 are observations in DB2ACCT, you can use

%INCLUDE SOURCLIB(BUILDPDB);RUN;

%LET MACKEEP=

MACRO _NOOBS % MACRO _YESOBS % ;

MACRO _LDB2ACC WORK.DB2ACCT %

;

%INCLUDE SOURCLIB(ASUMUOW);

and MXG will use only CICSTRAN as input to PDB.ASUMUOW.

Thank to Christian Hodel, SwissCom, SWITZERLAND.

====== Changes thru 24.216 were in MXG 24.08 dated Oct 18, 2006=========
Change 24.216 Using SUPPRESS with TYPETMNT failed; logic in UTILBLDP

UTILBLDP didn't protect non-standard-named-tokens, now corrected.

Oct 18, 2006

Thank to Robbie A. McCoy, Salt River Project, USA.

Change 24.215 Incorrect values for ICMP Statistics variables TSICDUTM

VMAC119 thru TSICOUAR, because -3 was incorrectly subtracted

Oct 18, 2006 twice from OFF11905 when subtype 6 logic was added.

Thank to Thomas Heitlinger, FIDUCIA IT AG, GERMANY.

====== Changes thru 24.214 were in MXG 24.08 dated Oct 17, 2006=========
Change 24.214 Processing z/OS 1.6 with SPG processing enabled failed

ASMRMFV with an 0C4, as there were no SPG records in z/OS 1.6.

Oct 16, 2006 Correction bypasses SPG for old versions.

Thank to Betty Wong, Bank of America, USA.

Change 24.213 Variable MCDFBID is now formatted HEX8. vice HEX4.; the

VMXGHSM field was INPUT as PIB4.

Oct 16, 2006

Thank to Sam Bass, McLane Company, USA.

Change 24.212 Support for NetSpy Version 11 was added in August, 2005,

VMACNSPY when test data validated that records were unchanged.

Oct 16, 2006

Thank to Brian Conway, IBM Global Services, CANADA.

Change 24.211 INVALID DATA messages for SCBGN, SYSIUL, SYSCIU because

VMACQACS MXG had &NUM vice &PD, and incorrect lengths for those

Oct 15, 2006 QAPMSYST variables.

Thanks to Jim Wertenberger, Antares Solutions, USA.

Change 24.210 Support for SMF 99 Resource Group section now creates new

EXTY99RG TYPE99RG dataset.

IMAC99

VMAC99

VMXGINIT

Oct 13, 2006

Thanks to Claude Breault, Centre de Services Partages Quebec, CANADA.
Change 24.209 Syntax error corrected; the previous circumvention to run

UTILBLDP as a two step process should no longer be required with

Oct 13, 2006 the Oct 17th edition.

Oct 17, 2006

Thanks to Ralph Gifford, AIG, USA.

Thanks to Bruce Whittington, TIAA-CREF,USA.

Change 24.208 LPAR share variables were destroyed by Dedicated CPUs;

VMAC7072 the test was expanded for their calculation to bypass:

Oct 13, 2006 IF SMF70CIN='CP' AND LCPUSHAR NE 0FFFFX THEN DO;

These variables were impacted if you have Dedicateds:

TOTSHARE TOTSHARC LPARNSW LPARSHAR LPARSHAC

Thanks to Rudolf Sauer, T-Systems Enterprise Services GmbH, GERMANY.

Change 24.207 TYPE72GO CPUTCBTM (and hence CPUTM) will incorrectly have

VMAC7072 included the zAAP CPU time, if zIIP APAR OA13499 was put

Oct 13, 2006 on, but you did not also install the z/OS web-deliverable

FMID JBB722S (or JBB66S9) for that zIIP APAR. The APAR

extended the segment to 576 bytes, adding zIIP fields,

but also two new IFA Service Units in R723CIFA,R723CIFC.

The two IFA fields are created by the APAR, but with only

the APAR installed, they are always zero. IBM says this

is working as designed, that I should only use the new

fields if they are non-zero. Unfortunately, IBM did not

document that "design" feature in the SMF manual!

Prior to the INPUT of R723CIFA/CIFC, MXG created IFAUNITS

from the CPUIFATM, but then the zero in R723CIFA was put

in IFAUNITS, so those service units were not subtracted

from the raw CPUUNITS, causing CPUTCBTM/CPUTM to include

CPUIFATM. This condition can be detected in your data,

and/or corrected in PDB.TYPE72GO with this test/logic:

DATA TYPE72GO;

SET PDB.TYPE72GO;

IF IFAUNITS=0 AND CPUIFATM GT 0 THEN DO;

NFOUND+1;

IF NFOUND=1 PUT 'INCLUDED ZAAP TIME WAS FOUND';

CPUTCBTM=CPUTCBTM-CPUIFATM;

CPUTM=CPUTM-CPUIFATM;

END;

Jan 18: This change in MXG 24.08 was a CRITICAL CHANGE,

and was the final change to the SPLIT70 redesign. For one

site, it corrected negative CPUOVHTM in PDB.RMFINTRV and

the associated error messages RMFINTRV was created that

could occur with MXG 23.23 thru MXG 24.07.

Thanks to Tom Draeger, Aurora, USA.

Thanks to Heimir Hauksson, Barclays Bank, UK.

Change 24.206 Support for NTSMF Version 3.0.0.7 adds two new objects:

EXNTDTBU NTDTBU DTSBUFFU NT DTS.BUFFER USAGE

EXNTDTPL NTDTPL DTSPERFL NT DTS.PERFORMANCE LIBRARY

IMACNTSM

VMACNTSM

VMXGINIT

Oct 11, 2006
Change 24.205 Support for DATABASE ==> INSTANCES object with NRDAT=152.

VMACNTSM Two new variables created in NTSMF dataset DATABASE:

Oct 11, 2006 LOGCKPDP='LOG*GENERATION*CHECKPOINT*DEPTH'

SBPRDRT ='STREAMING*BACKUP*PAGES*READ PERSEC'

Thanks to Paul Billick, Harleysville Insurance, USA.
Change 24.204 Support for IMPLEX Version 4.10 (INCOMPATIBLE, fields are

EXMPLXAR expanded and new ones were inserted), with new variables

IMACMPLX and new IMPLEXAR dataset for the subtype 8 Alert Record.

VMACMPLX Sorts were updated to ensure duplicates are removed.

VMXGINIT

Oct 11, 2006

Change 24.203 If the number of workloads requested to be graphed in the

GRAFWRKX sample program exceeded 20, the program failed, and the

Oct 9, 2006 values for 21,31,41,51,61,71,81, and 91 were not defined.
Change 24.202 -New macro variable &SMFEXIT is added in VMACSMF to each

VMXGINIT statement INFILE &SMF &SMFEXIT .... in preparation for

VMACSMF MXG support for compressed SMF records. Soon, &SMFEXIT

Oct 7, 2006 will name the to-be-provided "INFILE EXIT" that will

Jan 28, 2007 decompress SMF records "on the fly". &SMFEXIT defaults

to blank in VMXGINIT. Stay tuned for a later change.

Note, however, SAS Infile Exits only exist under z/OS.

-While intended for a different purpose, this new macro

variable, since it is inside the MXG INFILE statement,

can be used also to pass INFILE options. In particular,

you can limit the observations that will be READ from the

SMF file, using

%LET SMFEXIT FIRSTOBS=nnn OBS=mmm;

%INCLUDE SOURCLIB(....);

This is very useful for MXG members that have to invoke

PROC SORTs to deaccumulate (like TYPEDB2,TYPEVMXA, etc),

because you can NOT use an OPTIONS FIRSTOBS=mmm OBS=nnn

global statement with those code members; the global will

restrict the INFILE processing, but they then need to be

reset to FIRSTOBS=1 and OBS=MAX prior to the SORTs, and

there is not simple way to do that for these members.

But now, you can use the preceding example to restrict

the INFILE but not the subsequent PROC SORTs.
Change 24.201 Support for E-Thales Security product's five user SMF

EXTHALCD records (poor choice: they should have created a single

EXTHALEX SMF record and used five subtypes!) creates these seven

EXTHALHS datasets:

EXTHALSA DDDDDD MXG MXG

EXTHALSD DATASET DATASET DATASET

EXTHALSN SUFFIX NAME LABEL

EXTHALVI

IMACTHAL THALCD THALCDS CDS PARAMETER RECORD

TYPETHAL THALHS THALHSMD HSMD ENTRY IN CDS RECORD

TYPSTHAL THALEX THALEXCE DETAIL EXCEPTION RECORD

VMACTHAL THALVI THALVIOL SECURITY VIOLATION RECORD

VMXGINIT THALSA THALSUMA APPL IN SUMMARY RECORD

Oct 5, 2006 THALSD THALSUMD DEVICE IN SUMMARY RECORD

THALSN THALSNAP SRM SNAPSHOT RECORD

Thanks to ??? , Public Bank, MALAYASIA

Thanks to Patrick Yap Chee Keong, SAS Institute, MALAYASIA
Change 24.200 Support for SMF 82 Subtype 22 (TRUSTED BLOCK CREATE CALL)

EXTY8222 creates TYPE8222 dataset.

IMAC82 Variable SMF82SXT in Subtype 21 is now input as numeric

VMAC82 numeric variable, and FORMATed DATETIME21.2, as the

VMXGINIT time value is TODSTAMP and not the documented $CHAR8.

Oct 10, 2006

Thanks to Greg Burt, 5th3rd Bank, USA.
Change 24.199 -Support for ThruPut Manager Version 6 adds new variables

VMACTPMX for IBM/STK/VTAPE/COPYCROSS virtual tape usage, and new

Oct 5, 2006 "Drive Booking Services":

CAC7IDID CA7INALI JXDBSPR JXDBSUU JXDBSWG JXSERVIC

JXIMPORT VOLVIBM VOLVIBMR VOLVIBMN VOLVSTK VOLVSTKR

VOLVSTKN VOLVVTS VOLVVTSI VOLVVTSN VOLVCPC VOLVCPCO

VOLVCPCM

-Change 24.147 incorrectly inserted NOT in the test for

JBAFF; that NOT is removed, that change text updated.

If the two formats in your IMACTPMX do not correctly

map your SYSPLEX and SYSTEMs, that will cause JBAFF to

be blank, but the original MXG logic is correct.

Thanks to Scott Barry, SBBWorks, Inc.
Change 24.198 Variable INTETIME (Interval End) in TYPE30_6 was wrong

VMAC30 if GMT offset was non-zero; the subtype 6 doesn't contain

Oct 4, 2006 SMF30IST, which was used to calculate GMTOFF30. Now, that

is calculated as GMTOFF30=SMFTIME-INTETIME for subtype 6.

Thanks to Leendert Keesmaat, UBS, SWITZERLAND.

Thanks to Michel Denervaud, UBS, SWITZERLAND.

Change 24.197 Variables DCVDVTYP DCVDPTYP from the VOLS record are now

VMACDCOL kept in both DCOLDSET and DCOLCLUS datasets, so the type

Oct 3, 2006 of device is known. For example,

DCVDVTYP='3390' DCVDPTYP='33909'

DCVDVTYP='3390' DCVDPTYP='2105'

Thanks to Brian Harvey, Blue Cross Blue Shield of Illinois, USA.

Change 24.196 TYPE1415 INPUT STATEMENT EXCEEDED RECORD with z/OS 1.7

VMAC1415 plus ptf's, or with z/OS 1.8 due to MXG coding error that

Oct 3, 2006 was introduced in Change 24.094 for new PDSE Cache Stats.

Dec 5, 2006 SKIP=SKIP-32; is required in place of the SKIP=SKIP-16

added by Change 24.094 (in MXG 24.04).

Thanks to Diane Eppestine, AT&T Services, Inc, USA

Thanks to Frank Debree, Dexia, BELGIUM.
Change 24.195 Variables QW0143UR and QW0144UR printed funny values.

VMAC102 They are now FORMATted $HEX12. and input as $CHAR6.

Oct 2, 2006 instead of $EBCDIC6. (required for ASCII execution, makes

no difference when executing MXG on z/OS), like the other

QW0nnnUR Unit Recovery Token variables.

Thanks to Larry Stahl, IBM Global Services, USA.

Change 24.194 -CPUTM and QWSnXXXX variables in PDB.DB2STATS Statistics

VMACDB2 dataset was wrong, with large positive or negative values

Sep 29, 2006 due to incorrect login in MXG code.

Oct 10, 2006 -New QWSnZSRB variable, Preemptable SRB Time on zIIP is

now created and kept in the PDB.DB2STATS dataset.

-ADOCDB2 CPU variables were updated to indicate whether or

not they contained zIIP CPU time.

-Variable QWHUCPU is now kept in DB2ACCT.

-zIIP variables QW0231ZI and QW0231ZE for IFCID 231 are

now created and kept in T102S231.

Thanks to Jim Robson, HighMark, USA.

Thanks to John Paul, HighMark, USA.

Change 24.193 Support for 3592 Tape Devices (no change); they have the

VMAC30 same DEVTYPE as the 3590s, so usage will automatically

Sep 27, 2006 be stored in the xxxx3590 variables.
Change 24.192 Using %READDB2, you could not override the output DDNAME

READDB2 using %LET PDB2ACC=DB2ACCT; %READDB2(WANTONLY=ACCOUNT);

Sep 27, 2006 to create observations only in DB2ACCT.DB2ACCT. Now, if

PDBOUT= argument is null, the original _Ldddddd defs will

be used, so the %LETs can be used to changed output DD.

NOTE BENE: CHANGE 25.189 REVISED PDBOUT= options, and

THIS CHANGE IS NO LONGER CORRECT. SEE PDBOUT= option in

the READDB2 or ANALRMFR member for revised impacts.

NOW, if PDBOUT=, (the default) was specified, ALL output

datasets will be written to //WORK DD. PDBOUT=YES is now

required if you want your user tailoring honored.
Change 24.191 Support for HP MeasureWare for Windows/NT for Collector

EXMWNTAP Versions C.03.65.00 and C.04.50.00. C.04 has additional

EXMWNTCO variables that will be missing with C.03 data records.

EXMWNTDS These datasets are created:

EXMWNTGL DDDDDD DATASET DATASET

EXMWNTLA SUFFIX NAME LABEL

EXMWNTPR MWNTAP MWNTAPPL HPMWA MWNT APPL RESOURCES

EXMWNTTT MWNTCO MWNTCONF HPMWA MWNT CONFIGURATION

EXMWNTVL MWNTDS MWNTDSK HPMWA MWNT DSK ACTVTY FRM DIS

IMACMWNT MWNTGL MWNTGLOB HPMWA MWNT GLOBAL ACTIVITY

TYPEMWNT MWNTLA MWNTLANS HPMWA MWNT LANS

TYPSMWNT MWNTPR MWNTPROC HPMWA MWNT PROCESS RESOURCES

VMACMWNT MWNTTT MWNTTRAN HPMWA MWNT TRANSACTION TRACKER

VMXGINIT MWNTVL MWNTVOLS HPMWA MWNT VOLUME ACTIVITY

Oct 2, 2006

Thanks to Bobby Greer, Automobile Association of Michigan, USA.

Thanks to Dominik Covens, KBC Bankverzekeringsholding, BELGIUM
Change 24.190 Executing the output of %UTILBLDP as a separate step, if

UTILBLDP you specified both BUILDPDB=YES and EXPDBOUT= text that

Sep 25, 2006 had a %INCLUDE, caused an error deep inside SAS, either

a syntax error, or expression exceeded 64000 bytes error.

If you use %UTILBLDP this way in the two-step process,

then you must add this statement at the first //SYSIN

in the saved output that will be executed:

%LET BLDPOUT= xxxxxx ;

where xxxxx is the text in the EXPDBOUT= argument.

Thanks to Robert Carballo, Office Depot, USA.

Change 24.189 The options NOOPD and NOSPG did not suppress the OPD/SPG

ASMRMFV records, causing OPD or SPG records to be output when you

Sep 25, 2006 had intended to not write them. And, the NOCSR flag was

incorrectly set when NOSPG was specified.

Thanks to Jerry Urbaniak, Acxiom CDC, USA.
Change 24.188 Support for new TNG object from NT and SOLARIS platforms:

EXTNT131 dddddd Dataset Description

EXTSO029 TNT131 NT131 SLM METRICS

IMACTNG TSO029 SO029 CA PROCESS GROUP

VMACTNG and additional variables in AI019, AI022, AI024, and the

VMXGINIT NT035 datasets, and labels were corrected.

Sep 26, 2006

Thanks to Michael Kynch, International Paper, USA.

Change 24.187 ASUM70PR with INTERVAL=HOUR (or any duration) can create

VMXG70PR observations with incorrect DURATM (50 vice 60 minutes,

Sep 23, 2006 which then caused PCTCPUBY and PCTOVHD to be too large),

Oct 8, 2006 because MXG's heuristic test IF FIRST.DURATM to recognize

a new group to store SYSDUR failed when adjacent DURATMs

happened to be identical. That test is enhanced to also

recognize a group when the new LCPUADDR is less than the

old LCPUADDR, a far more robust criteria. This error can

NOT occur with the defauit INTERVAL=DURSET in ASUM70PR,

which does not summarize by time, so this error can occur

only if you have a tailored ASUM70PR member. Five out of

twenty-four intervals were wrong in one day's data, but

only those three variables were wrong; all of the other

data were correct.

-Oct 8: Notes about uninit OLDSTART, OLD70GIE eliminated.

Thanks to Scott Weiner, WPS Insurance Corporation, USA.

====== Changes thru 24.186 were in MXG 24.07 dated Sep 22, 2006=========
Change 24.186 Support for zIIP variables in PDB.RMFINTRV dataset; I had

VMXGRMFI overlooked this addition. MXG 24.07 was redated with the

Sep 17, 2006 change so all of the "mainstream" MXG datasets now have

the additional sets of ZIP/ZIE variables.

Thanks to Jonathan M. Miller5, JohnDeere, USA.
====== Changes thru 24.185 were in MXG 24.07 dated Sep 21, 2006=========
Change 24.185 Support for DMF Product that creates SMF 110 records with

VMAC110 SMFPSRVR=41.1, i.e., CICS/ESA 4.1.1 (from 1994). While

Sep 17, 2006 MXG supported 4.1.0 and 5.1.0, there never was an MXG

site with 41.1, and it may be their creation, as it has a

seventh TCB, while 5.1.0 only had six.

Thanks to Vernon Stanton, Government of South Australia, AUSTRALIA.

Change 24.184 -Variables ORIGWAIT in PDB.TYPE70PR was not populated for

VMAC7072 IFAs and ZIPs, but it is needed so that both the "CPU"

Sep 17, 2006 busy (LCPUPDTM-based) and the "MVS" busy (ORIGWAIT-based)

can be calculated for analysis of logical ready queues.

Now, in the observations with PARTISHN=LPARNUM ("this"),

ORIGWAIT will be populated for SMF70CIN='IFA' or 'IIP'.

-PDB.TYPE70 existing variables PCTIFBYx and PCTZIBYx are

the "MVS" percentages, but MXG did not create the "CPU"

(LCPUPDTM-based) percentage for IFAs nor ZIPs. Now, new

PCTCIBYx variables are created with the "CPU"/"Logical"

percentages, for the IFA or ZIPs, and variable IFATYPx

identifies if the CPU is an IFA or a ZIP. This permits

the analysis of Logical Ready Queuing for IFAs and ZIPs,