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

from being created later in the PDB build because they

contain only control data not useful for analysis.

Excluding these tables in ASMRMFV processing will reduce

PDB build overhead by not creating SHD and RED table

output in the RMFBSAM file in the first place.

-NOTE: If you do not plan to develop your own USE/WAIT

analysis by type of WAIT, you can also use add NOUWD to

the PARM= or SYSIN file to further reduce PDB build

overhead.

-The ASMRMFV source prologue as well as the ADOCRMFV

documentation member are updated appropriately.

-REQUIREMENT: In order to receive these improvements the

current ASMRMFV utility program from this MXG change must

be installed. See MXG SOURCLIB member JCLASM3 for sample

JCL for the assembly and link-edit install steps.

Thanks to Peter Gray, HP Australia, for the DSNTYPE=LARGE suggestion.

Change 31.229 -Documentation. Regarding the inconsistent interval data

VMACXAM DURATION (seconds) mentioned, contact VELOCITY SOFTWARE

Oct 25, 2013 for a band-aid "fix" with your zWRITE data-collection

script. Explain that you want to see consistent

15-minute interval data for the entire 24-hour period.

This would be akin to the IBM RMF SYNC(59M|00M) option.

-Also, you want to be up-to-date on VELOCITY zWRITE/zVPS

maintenance, that being at least PROD4142 to address some

instances when not all data was being flushed from the

in-memory buffers at zWRITE data-close (typically

end-of-day) time.

Thanks to Scott Barry, SBBWorks Inc, USA.

Change 31.228 IMS zIIP CPU time is created in IMF CIMSTRAN dataset as

VMACCIMS CPUZIPTM=TRXZTCPU-TRXZONCP;

Oct 25, 2013

Thanks to Scott Barry, SBBWorks Inc, USA.

Change 31.227 These report members failed after the default argument

ANALAVAI PDB= was changed to &PDBMXG, which fails due to timing of

ANALBLSR macro variable resolution. This example shows the problem

ANAL72GO and that a direct circumvention requires THREE periods to

Oct 23, 2013 properly resolve to PDB.TYPE72GO:

Nov 12, 2013 %MACRO TESTIT(PDB=&PDBMXG);

%LET A=&PDB..TYPE72go;

%LET B=&PDB...TYPE72go;

%PUT A=&A;

%PUT B=&B;

%MEND TESTIT;

%TESTIT;RUN;

Why only these three, when MANY of the %MACRO members in

MXG have the same syntax? Because all of the others have

a %LET PDB=%UPCASE(&PDB); statement after the %MACRO, and

that (accidentally?) circumvents the macro timing, so if

these three had been properly created with the %UPCASE,

this peculiarity would not have been observed.

-ODS Graphics likes to override time formats with DATETIME

formats. YAXIS statement inserted to force it to use

the format as specified in the data. Nov 12 update.
Change 31.226 Support for Alebra Parallel Data Mover USER SMF record

EXTYPDM Performance Record creates TYPEPDM dataset:

IMACPDM dddddd Dataset Description

TYPEPDM TYPDM TYPEPDM Alebra PARALLEL DATA MOVER

TYPSPDM

TESSUSR1

VMACPDM

VMXGINIT

Oct 23, 2013
Change 31.225 Support for Emtex JESConnect product's SMF Performance

EXTYJESC Record creates TYPEJESC dataset:

IMACJESC dddddd Dataset Description

TYPEJESC TYJESC TYPEJESC JES2 CONNECT

TYPSJESC -VMXGIPV4 is created to convert IPV4 $CHAR4 fields into

TESSUSR1 dotted IPV4 text IP Address, similar to VMXGIPV6.

VMACJESC

VMXGIPV4

VMXGINIT

Oct 23, 2013

Thanks to Larry Stahl, IBM Global Services, USA.
Change 31.224 WPS Version 3.0 QA tests found these incompatibilities:

ANAL80A -INFILE operand TERMSTR, set in macro variable &MXGCRLF to

ANALAVAL CR/CRLF/LF for ASCII in VMXGINIT, to define the end of

ANALPATH line delimiter for input data files that can be either

ANALTAPE PC or Unix format, is not supported in WPS, so it is now

TYPESVC set only for execution under SAS. Currently, it is used

UTILEXCL only in TYPERACF, TYPETAPR and TYPEWWW.

VMXGINIT -In WPS, PROC APPEND BASE=PDB.CICSDICT NEW=CICSDICT; does

Oct 22, 2013 not propagate the Data Set Label from CICSDICT when the

BASE does not exist, while SAS does propagate that Label.

Using BASE=PDB.CICSDICT (LABEL='CICDIC: CICS DICTIONARY')

NEW=CICSDICT; also does not work with WPS while it does

work under SAS. So, an extra DATA step was needed after

the PROC APPEND to label the dataset.

(The absence of the Data Set Label caused the MXG QA job

to complain in DOCVER when the label was not found.)

-TYPESVC/TYPSSVC - WPS does not support NAMED INPUT.

-ANAL80A - WPS does not support PROC REPORT.

-ANALAVAL- WPS does not support PROC CALENDAR.

-ANALMPL,ANALTAPE - WPS does not support VREVERSE in PLOT.

-ANALPATH- WPS doesn't support OVERPRINT in PUT statement.
Change 31.223 CICS variable WMQASRTM is added into the CPUTM variable

VMAC110 in CICSTRAN datasets; back in CICS/TS 4.1 IBM added the

Oct 20, 2013 field (WMQASRBT, clock and count) and documented it:

The WebSphereMQ SRB time this transaction spent

processing MQ API requests. This field is zero for

point-to-point messaging activity, but is non-zero if

MQ API requests result in publish and subscribe type

messaging. Note: WebSphere MQ only returns this value

to CICS when MQ Class 3 accounting information is being

collected in WebSphere MQ. To collect the MQ Class 3

accounting info, START TRACE(ACCTG) DEST(SMF) CLASS(3)

must be issued in WebSphere MQ.

This SRB CPU time is already included in the CPUTCBTM in

the SMF 30s for the CICS region.

Note: VMAC110/UTILEXCL/DOCVER have these MQ variables

MQGETWTM MQGETWCN MQREQS MQWTCBUS MQQONTCPU MQREQUS

that were originally decoded in 2003 in Change 21.212,

from CMODNAME='MQSERIES' and CMODHEAD='MQGETW/MQREQS/

MQWTCBUS/MQONTCPU/MQREQUS' but I can not find any of

those names in CICS Data Areas, nor have those names

appeared in any CICS Dictionaries in recent versions so

I presume they are no longer created, and the newer

WMQGETTM WMQGETTM WQMREQCT WMQASRTM WMQASRCN variables

have replaced the 2003 variables. They exist in DOCVER

only because I force them to be part of the MXG QA, and

since they are ONLY created by UTILEXCL and ONLY IF the

CMODHEAD/CMODNAME are found, it is unlikely they will

even exist in your CICSTRAN dataset.

Thanks to Scott Barry, SBBWorks Inc, USA.
Change 31.222 INPUT STATEMENT EXCEEDED RECORD for RMF III SPG record if

VMACRMFV the system had more than 1361 devices; NRVOLSTV=1674 was

Oct 19, 2013 used for the VOL count, but that is the total number of

VOLs in ALL SPG records. Instead, SPGVOLNR is the count

of vols in this record, and is now used for the loop.

Thanks to MP Welch, Bank of America, USA.

Change 31.221 Revisions to ODS support. Revised VMXGODSO/ODSC members

ANAL72GO and to the programs that use them.

ANALACTM -VMXGODSO (Open ODS destination) now supports TYPE=PDF

ANALAVAI and passes the ODSTYPE to VMXGODSC.

ANALDB2B -VMXGODSC now actually correctly closes the current ODS

ANALGRID destination and restores the original ODS settings using

ANALHTML two new user-supplied parameters:

ANALID ODSORIGTYPE= the type of ODS output you want set after

ANALIDMS close. If you leave it blank, it will

ANALINIT revert to HTML on ASCII and to LISTING

ANALNAT in all other cases.

ANALNATR ODSORIGFILE= if you want to direct the output after

ANALPKGS close to a specific destination; can be

ANALRAID blank

ANALRANK -ANAL72GO adds the ODSORIGTYPE and ODSORIGFILE parameters

ANALS225 for VMXGODSC and also adds ODS GRAPHICS if you are on SAS

ASUMRAID 9.3 or greater and only uses SAS/GRAPHS for earlier SAS

GRAFCEC versions (if SAS/GRAPH is detected) and WPS.

GRAFRAID -ANALACTM. Added ODSORIGTYPE and ODSORIGFILE parameters

GRAFRMFI for VMXGODSC. The report of the WLM configuration needs

GRAFWRKX landscape orientation; to make that happen with TYPE=PDF

VGETWKLD the OPTIONS ORIENTATION=LANDSCAPE; statement is added

VMXG2DTE before the ODS PDF statement. Since ORIENTATION is an

VMXGODSC option in the format NAME=VALUE, which was not previously

VMXGODSO used in VMXGOPTR, it was also revised to support it.

VMXGOPTR -ANALAVAI. Adds the ODSORIGTYPE and ODSORIGFILE parameters

VMXGRMFI for VMXGODSC and corrects a macro variable resolution

Nov 2, 2013 problem discovered in testing.

Dec 8, 2013 -ANALDB2B. Adds the ODSORIGTYPE and ODSORIGFILE parameters

for VMXGODSC.

-ANALGRID. VMXGODSO/VMXGODSC execution is now conditional

so you can combine the output of ANALGRID in a single

file with other reports and graphs, such as:

%VMXGODS(TYPE=PDF,FILE=MYREPORT.PDF);

%ANALGRID(ODSTYPE=NO,other parameters);

PROC SGPLOT;

etc...

%VMXGODSC;

-ANALHTML. Adds the ODSORIGTYPE and ODSORIGFILE parameters

for VMXGODSC.

-ANALID. If you invoked ANALID without running TYPEID,

it would fail without really telling you why because it

could not find the _WTYID _LTYID datasets. Now if it

detects that TYPEID has not been run it automatically

includes VMACID to create those macros.

-ANALIDMS. Adds the ODSORIGTYPE and ODSORIGFILE parameters

for VMXGODSC.

-ANALINIT. Adds the ODSORIGTYPE and ODSORIGFILE parameters

for VMXGODSC, and some reformatting for PDF output.

-ANALNAT and ANALNATR. PLOTs changed to SGPLOT. Comments

added to show how to create PDF output as well as HTML.

-ANALPKGS. PDF comments added and HTML commented so that

you can choose which you prefer

-ANALRAID. Added ODSORIGTYPE and ODSORIGFILE parameters

for VMXGODSC. Modified for PDF output.

-ANALRANK. Adds the ODSORIGTYPE and ODSORIGFILE parameters

for VMXGODSC. Reduce to a single report.

-ASUMRAID modified for use with VMXGODSO and VMXGODSC. Now

makes GRAFRAID obsolete since it creates the same output

using HTML or PDF, but does it without SAS/GRAPH.

-ANALS225. Uses DB2STATS as input since the IFCID=225 data

in completely contained in PDB.DB2STATS datasets and the

T102S225 is not created. DB2 interval records are now

created every minute (cannot be changed) so the revised

report provides for summarization to any chosen interval.

ODSORIGTYPE and ODSORIGFILE parameters added.

-GRAFCEC modified to use VMXGODSO and VMXGODSC for HTML

or PDF and to use SGPLOT when on SAS 9.3 or higher.

-GRAFRAID - marked as obsolete. ASUMRAID creates the same

output more cleanly and without requiring SAS/GRAPH if

you use PDF or HTML.

-GRAFRMFI modified to use VMXGODSO and VMXGODSC for HTML

or PDF and to use SGPLOT when on SAS 9.3 or higher.

-GRAFTAPE. Adds the ODORIGTYPE and ODSORIGFILE parameters

for VMXGODSC and adds ODS GRAPHICS if you are on SAS 9.3

or later, and uses SAS/GRAPH for earlier SAS and WPS.

-GRAFWRKX (hourly bar chart of each workload in RMFINTRV)

now uses and describes each of the SGPLOTs; ODSORIGTYPE

and ODSORIGFILE added to parameters for VMXGODSC.

Dec 8: ODS statement parameter ODSSTYLE added pointing at

STYLES.MXGSTYLE1.

-VMXG2DTE. Option added to CYCLE parameter to enable tests

at any time. CYCLE=FORCE will override the input DD in

the MTD database.

-VMXGOPTR. Enhanced to handle options where the format is

OPTION optionname=value (LINESIZE PAGESIZE VMXGOPTR),

needed for ODS TYPE=PDF when landscape is desired.

-VGETWKLD modified to end if no workloads are found and

fill in all macro variables expected downstream.

-VMXGRMFI trend logic modified to stop if no workloads are

found.

-VMXGODSO Dec 8, 2013: Cleanup and realignment of code and

added new ODSSTYLE= argument (especially for PDF output).
Change 31.220 Enhancement to CRYPTOGRAPHIC PROCESSOR support.

VMAC7072 -TYPE7002 dataset variables R702ASDC R702ASDB R702ASDI

Oct 16, 2013 were misspelled (R702AEDC,R702AEDB,R702AEDI).

-IBM RMF Report tokens for the OVERVIEW/EXCEPTION reports

are now added as variables to the crypto datasets:

TYPE7002 - PCCIC

CRYCKR='COPROCESSOR*KEY-GEN*RATE'

CRYCTE='COPROCESSOR*TOTAL*AVG EXECUTE*TIME'

CRYCTR='COPROCESSOR*TOTAL*RATE'

CRYCTU='COPROCESSOR*TOTAL*PCT UTILIZATION'

TYPE70X2 - PCICA

CRYAC1E='ACCELERATOR*1024BIT-CRT*AVG EXECUTE*TIME'

CRYAC1R='ACCELERATOR*1024BIT-CRT*RATE'

CRYAC1U='ACCELERATOR*1024BIT-CRT*PCT UTILIZATION'

CRYAC2E='ACCELERATOR*2048BIT-CRT*AVG EXECUTE*TIME'

CRYAC2R='ACCELERATOR*2048BIT-CRT*RATE'

CRYAC2U='ACCELERATOR*2048BIT-CRT*PCT UTILIZATION'

CRYAC3E='ACCELERATOR*4096BIT-CRT*AVG EXECUTE*TIME'

CRYAC3R='ACCELERATOR*4096BIT-CRT*RATE'

CRYAC3U='ACCELERATOR*4096BIT-CRT*PCT UTILIZATION'

CRYAM1E='ACCELERATOR*1024BIT-ME*AVG EXECUTE*TIME'

CRYAM1R='ACCELERATOR*1024BIT-ME*RATE'

CRYAM1U='ACCELERATOR*1024BIT-ME*PCT UTILIZATION'

CRYAM2E='ACCELERATOR*2048BIT-ME*AVG EXECUTE*TIME'

CRYAM2R='ACCELERATOR*2048BIT-ME*RATE'

CRYAM2U='ACCELERATOR*2048BIT-ME*PCT UTILIZATION'

CRYAM3E='ACCELERATOR*4096BIT-ME*AVG EXECUTE*TIME'

CRYAM3R='ACCELERATOR*4096BIT-ME*RATE'

CRYAM3U='ACCELERATOR*4096BIT-ME*PCT UTILIZATION'

TYPE70Y2 - CPACF

CRYIADO='AVG TIME*COPROCESSOR*CALLS*AES*DECRYPT'

CRYIADR='RATE*OF*AES*DECRYPT*SERVICE*CALLS*SENT'

CRYIADS='AVG BYTES*PER*AES*DECRYPT*SERVICE*CALL'

CRYIAEO='AVG TIME*COPROCESSOR*CALLS*AES*ENCRYPT'

CRYIAER='RATE*OF*AES*ENCRYPT*SERVICE*CALLS*SENT'

CRYIAES='AVG BYTES*PER*AES*ENCRYPT*SERVICE*CALL'

CRYIH2I='INSTR*TO*HASH*DATA*SHA-224*SHA-256*ALGO'

CRYIH2R='HASHING RATE*SHA-224*SHA-256*ALGORITHM'

CRYIH2S='HASHING SIZE*SHA-224*SHA-256*ALGORITHM'

CRYIH5R='HASHING RATE*SHA-384*SHA-512*ALGORITHM'

CRYIH5S='HASHING SIZE*SHA-384*SHA-512*ALGORITHM'

CRYIHAI='INSTR*TO*HASH*DATA*SHA-1*ALGORITHM'

CRYIHAR='HASHING RATE*SHA-1*ALGORITHM'

CRYIHAS='HASHING SIZE*SHA-1*ALGORITHM'

CRYIMGI='INSTR*TO MAC*GENERATE'

CRYIMGR='MAC*GENERATION*RATE'

CRYIMGS='MAC*GENERATION*SIZE'

CRYIMVI='INSTR*TO MAC*VERIFY'

CRYIMVR='MAC*VERIFY*RATE'

CRYIMVS='MAC*VERIFY*SIZE'

CRYIPTR='PIN*TRANSLATION*RATE'

CRYIPVR='PIN*VERIFY*RATE'

CRYISDDI='SINGLE*DES*INSTR*DECIPHER*THE*DATA'

CRYISDDR='SINGLE*DES*DECRYPT*RATE'

CRYISDDS='SINGLE*DES*DECRYPT*SIZE'

CRYISDEI='SINGLE*DES*INSTR*ENCIPHER*THE*DATA'

CRYISDER='SINGLE*DES*ENCRYPT*RATE'

CRYISDES='SINGLE*DES*ENCRYPT*SIZE'

CRYITDDI='TRIPLE*DES*INSTR*TO*DECIPHER*THE*DATA'

CRYITDDR='TRIPLE*DES*DECRYPT*RATE'

CRYITDDS='TRIPLE*DES*DECRYPT*SIZE'

CRYITDEI='TRIPLE*DES*INSTR*TO*ENCIPHER*THE*DATA'

CRYITDER='TRIPLE*DES*ENCRYPT*RATE'

CRYITDES='TRIPLE*DES*ENCRYPT*SIZE'

TYPE70Y3 - CE4XP

CRYPAGE='GENERATION*ASYMMETRIC-KEY*AVG EXEC TIME'

CRYPAGR='GENERATION*ASYMMETRIC-KEY*RATE'

CRYPAGU='GENERATION*ASYMMETRIC-KEY*PCT UTILIZATION'

CRYPFAE='FAST*ASYMMETRIC-KEY*AVG EXEC TIME'

CRYPFAR='FAST*ASYMMETRIC-KEY*RATE'

CRYPFAU='FAST*ASYMMETRIC-KEY*PCT UTILIZATION'

CRYPSAE='SLOW*ASYMMETRIC-KEY*AVG EXEC TIME'

CRYPSAR='SLOW*ASYMMETRIC-KEY*RATE'

CRYPSAU='SLOW*ASYMMETRIC-KEY*PCT UTILIZATION'

CRYPSCE='COMPLETE*SYMMETRIC-KEY*AVG EXEC TIME'

CRYPSCR='COMPLETE*SYMMETRIC-KEY*RATE'

CRYPSCU='COMPLETE*SYMMETRIC-KEY*PCT UTILIZATION'

CRYPSPE='PARTIAL*SYMMMETRIC-KEY*AVG EXEC TIME'

CRYPSPR='PARTIAL*SYMMETRIC*RATE'

CRYPSPU='PARTIAL*SYMMMETRIC-KEY*PCT UTILIZATION'

CRYPTE='PKCS11*COPROCESSOR*AVG EXEC TIME'

CRYPTR='PKCS11*COPROCESSOR*TOTAL*RATE'

CRYPTU='PKCS11*COPROCESSOR*TOTAL*PCT UTILIZATION'

Thanks to David Wilbur, State of Washington CTS, USA.

Thanks to Steve Majerick, State of Washington CTS, USA.
Change 31.219 -If you only want to know which character variable(s) in

VMXGSRCH which dataset(s) contains a text string, RESULTS=COUNT

Oct 15, 2013 produces the simple report of names and counts.

-Specifying RESULTS=PRINT prints all variables in those

found observations (printing the first NOBS observations

with both variable name and label as heading) and now

also produces the COUNT report after the prints.

-A log message is printed when observations are found.

-The VALUE= argument does not need to be the exact value

of a variable. VALUE=XYZ will find XYZ123 or ABCXYZ or

any variable that has XYZ anywhere in its value.
Change 31.218 INVALID DATA FOR SCACTIME in ID=119 SUBTYPE=32 because

VMAC119 variable SCORIGIN was overlooked and not INPUT causing

Oct 10, 2013 the misalignment by one byte.

Thanks to Thomas Heitlinger, Fiducia IT AG, GERMANY.

Thanks to Sieghart Seith, Fiducia IT AG, GERMANY.

Thanks to Dennis Gaertner, Fiducia IT AG, GERMANY.

Change 31.217 Support for CICS "multi journal records" adds examples in

EXCICJRN the EXCICJRN member to decode multiple journals to create

Oct 10, 2013 multiple datasets from different journal records. You

would copy the example EXCICJRN member into your "USERID"

MXG tailoring library, then comment out its default code

block, un-comment and modify its "multi" example code to

create and output your variables into CICSJOUR, then copy

the MACRO _SCICJRN and MACRO _KCICJRN definitions into

IMACKEEP in your tailoring library, to define which of

the WORK.CICSJOUR variables are output into which of your

new datasets. Then your SYSIN program would be:

%LET MACFILE=

%QUOTE( IF ID=110 AND SUBTYPE=0; );

%INCLUDE SOURCLIB(TYPE110);

_SCICTRN

RUN;

if you ONLY want to process the Journal Records.

Thanks to Paul Bennett, Euroclear, BELGIUM.

Change 31.216 INVALID DATA FOR S42CSYNC because it is not an SMFSTAMP8

VMAC42 value, instead containing a nonstandard 2013281F0916483F

Oct 10, 2013 format that is now "hand" decoded to 08OCT2013:9:16:48.3.

Thanks to Karl Lasecki, Chemical Abstracts Service, USA.

Change 31.215 New variables added to TYPETLMS are now supported:

VMACTLMS BACODLVL='ENHANCED*CODE*LEVEL*03X'

Oct 7, 2013 BAKEYIDX='ENCRYPTION*KEY*INDEX'

BAPCTCMP='PCT*OF COMPACT*FOR FILE'

BAPCTFIL='PCT*VOLUME*USED*BY FILE'

BAUSQFIL='USAGE*SEQUENCE*FOR FILE'

Thanks to Scott Barry, SBBWorks Inc, USA.

Thanks to Yves Cinq-Mars, IBM Global Services, CANADA.

Change 31.214 CICS dataset CICSJS variables SJSMAJCP and SJSMINCP in

VMAC110 statistics dataset CICSJS were incorrectly documented and

Oct 3, 2013 labeled as CPU time so MXG input them as such, but IBM

now confirms that they are elapsed time in milliseconds

in the record, so both are now correctly input, labeled

(TOTAL*ELAPSED*TIME IN MAJOR/MINOR GARBAGE COLLECTION)

and formatted TIME13.3 to display the millisecond as the

final digit in the decimal place.

Thanks to Mark Wittie, Fidelity Institutional, USA.

Thanks to Warren Cravey, Fidelity Institutional, USA.

Thanks to Paul Albrecht, Fidelity Institutional, USA.
Change 31.213 MXG 31.06-31.07 only, and only if syntax of 102.xxx was

UTILBLDP used to select ID=102 records by IFCID/SUBTYPE. Errors

Oct 4, 2013 for old-style MACRO _V102xxx were printed on the log.

Introduced in Change 31.184.

Change 31.212 New (internal) parameters SPINTAPE and NOEXIMSG added.

ANALDB2R -SPINTAPE=YES will cause VGETOBS to read an input SAS data

VGETOBS library when it is on tape/sequential format, building

Oct 18, 2013 the MXGTABLES dataset that list all SAS datasets on the

tape. Used now in ANALDB2R but available to other MXG

utilities. Normally, the MXGTABLES is used and then

discarded by VGETOBS, but with this parameter it is kept.

Default for SPINTAPE is NO in VGETOBS, set by caller.

-Option DATASET=_ALL_ will also cause MXGTABLES to be kept

for use in subsequent internal utilities.

-NOEXIMSG=NO (default) added to suppress messages when the

input datasets are on tape, after the first time that DD

is encountered.

-ANALDB2R exploits this new VGETOBS to snapshot a list of

all of the datasets found in the first libname in PDB=,

whether it is on tape or disk, and then builds a MACRO

variable with that list of available datasets, avoiding

the earlier design that required many executions of the

VGETOBS macro, potentially eliminating many tape mounts.
Change 31.211 ANALHSM exploited the fact that SAS will always look for

ANALHSM the last dataset created when doing a SORT or PRINT. But

Sep 26, 2013 the latest changes to VGETOBS now create a temporary

dataset that intervenes and disappears so the explicit

dataset to be sorted/printed had to be specified.

Thanks to Karl Lasecki, Chemical Abstracts Service, USA.

Change 31.210 Variables R744SATM,R744SDTM,R744SQTM,R744SSTM are in

VMAC74 seconds per event, but that was not obvious. Their label

Sep 25, 2013 now contain *SECONDS*PER....' and are formatted 8.6.

So a value of 0.000094 is 94 microseconds per ....

Thanks to Cesar V Cocco, CitiCorp, USA.
Change 31.209 -If you specified PDB=YES and used the LDB2xxxx parameters

READDB2 to reroute the output datasets, there was a failure when

CLEARDB2 looking for the YES DDNAME in VGETOBS. Should have been

Sep 27, 2013 pointing to the PDB2xxxx specific MACRO VARIABLE rather

than PDBOUT.

-Specifying STATS/STATISTICS but then Suppressing one of

the statistics datasets caused DATASET NOT FOUND errors;

now if a required dataset is not present, DB2STATS won't

be created but the READDB2 won't terminate with an error.

Not all old style macros defined in VMACDB2 were cleared.

This could cause many strange errors, but ONLY if there

were multiple executions of READDB2 in the same session,

e.g., the MXG QA Test stream. This change was found when

many additional %READDB2 QA tests of previously reported

errors were added to the QA stream.

-%TRIM and %QCMPRES %MACRO functions were replaced by

these %SYSFUNC calls to use COMPBL as those %MACRO

functions did not always work correctly and caused

errors where macro variables exceeded the 64K limit.

In addition, elimination of %TRIM and %QCMPRES removes

the past issues with incorrect SASAUTOS allocations.

%LET IDREADS=%SYSFUNC(COMPBL(&IDREADS 102));

%LET IDREADS=%QCMPRES(&IDREADS 102);

%LET IDREADS=%SYSFUNC(COMPBL(&IDREADS 102));

%LET IDREADS=%QCMPRES(&IDREADS 102);
Change 31.208 -ASUM113 variables LPARBUSY and MIPSEXEC were wrong for

ASUM113 the zEC12 processor; they should be divided by DELTASUM

Sep 24, 2013 and not DELTATM in those equations for the processor

Thanks to Andrew Hebden, Barclays, ENGLAND.