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

SPIN libname when set to yes - primarily for MXG support

-Checks added to ensure that PDB SPIN TREND WEEK WTD MONTH

and MTD libnames are allocated as needed by the other

parameters used. In the case of WTD/MTD if they are not

found but WEEK/MONTH are a warning is issued and the

libnames that were found are substitute
Change 34.275 Support for ASG/TMON for DB2 IDAA SA and SB records

EXTMD2SA creates three new datasets:

EXTMD2SB DDDDDD DATASET DESCRIPTION

EXTMD2SX TMD2SA TMD2SA IDAA SA Summary

IMACTMD2 TMD2SB TMD2SB IDAA SB Summary

VMACTMD2 TMD2SX TMD2SBD IDAA SB Detail

VMXGINIT

Dec 12 2016

Thanks to Daniel Hamiel, NedBank, SOUTH AFRICA.

Thanks to Mike Lotter, NedBank, SOUTH AFRICA.

Change 34.274 Support for CICS/TS 5.4 Beta 6 INCOMPAT, new CICSTRAN

EXCICMQR fields inserted.

FORMATS -New variables added to CICSTRAN:

IMAC110 ASFREECT='EXEC CICS*FREE CHILD*COUNT'

UTILEXCL ASFTCHCT='EXEC CICS*FETCH*COMMANDS'

VMAC110 ASFTCHCN='ASYNC API*FETCH*WAIT*COUNT'

VMXGINIT ASFTCHTM='ASYNC API*FETCH*WAIT*DURATION'

Dec 15, 2016 ASNATCN='ASYNC API*RUN DELAYEDCOUNT'

ASRMATTM='ASYNC API*RUN DELAYED*DURATION'

ASRUNCT ='EXEC CICS*RUN*TRANSID*COUNT'

ASTOTCT ='ASYNC API*COMMANDS*COUNT'

PTCOUNT ='PREVIOUS*TRANSACTION*COUNT'

PTSTART ='PREVIOUS*TRANSACTION*START*DATETIME'

PTTRAN ='PREVIOUS*TRANSACTION*TRANSID'

PTTRANNO='PREVIOUS*TRANSACTION*sequence*number'

-New CICMQR MQ Monitor statistics dataset from STID=148

is created.
Change 34.273 Internal restructure of ASMRMFV for possible performance

ADOCRMFV improvements and better design for maintenance.

ASMRMFV -Mitigate Store In Instruction Stream (SIIS) conditions

Dec 6, 2016 and other improvements.

-Updating data imbedded in an instruction stream or

modifying instructions results in additional CPU overhead

maintaining the data and instruction caches.

-ASMRMFV is changed to isolate and align all data used in

subroutines on 256 byte cache boundary lines. This

increases the size of the ASMRMFV load module about 7% to

about 272K.

-IBM Service Call macros for OPEN, CLOSE, RDJFCB, and so

on are split into Execute and List forms because standard

macro expansions update parameters in the instruction

stream.

-Limited volume testing showed about a 1% CPU Time

reduction that may vary in actual production use.

-The MODCB Service Call function is no longer used to

alter the VSAM ACB and RPL control blocks. ASMRMFV only

makes trivial changes to these during processing and

Execute and List forms of MODCB generated a lot of

instructions.

-The MODCBERR subroutine used to process MODCB errors

is deleted.

Change 34.272 Support for the MQ SMF 115 Subtype 231 DSP/ADP/SSL/DNS

EXTY115A segments, each of which creates new dataset:

EXTY115D DDDDDD DATASET DESCRIPTION

EXTY115L TY115D MQMDSP MQM DISPATCHER

EXTY115N TY115A MQMADP MQM ADAPTER

IMAC115 TY115L MQMSSL MQM SSL

VMAC115 TY115N MQMDNS MQM DNS

VMXGINIT Their unique variables that were previously incorrectly

Dec 2, 2016 kept in dataset MQMCHIN have been dropped.

-Each segment contains QCTCPTM and QCTELPT, CPU & Elapsed

time; many observations have CPU Time slightly larger

than Elapsed time (largest 15 with QCTCPTM 364 seconds).

IBM Explains:

CPU start and end times are taken directly from the

TCB to minimize performance impact. This field is only

updated when the TCB is undispatched, so if the TCB

has been dispatched for a while when the TCB CPU time

is taken at the start of the request, this value

could be a bit low, which could mean that the TCB CPU

interval calculation returns a value which is slightly

high. This is as-designed, and the data is still

useful. When elapsed time and CPU time are similar,

or when CPU time appears greater than elapsed time,

the task is getting all the CPU it needs, and you can

interpret elapsed time as an approximation for CPU

time. When elapsed time is significantly larger than

CPU time, then the task is having to wait for CPU or

for some internal wait, and that difference may be of

interest.

-APAR PI46585 is required to correct negative or invalid

values in QCTWTTM, QCTLSTM, and QCXTLGTM in these new

ADP/DSP/SSL/DNP datasets.

-The subtype 215 record replaced the subtype 2 record when

OPMODE(NEWFUNC) is specified; the buffer manager data

that was output in dataset MQMBUFER is now instead output

in dataset TY115215.

Thanks to Carol Arnold, Brown Brothers Harriman, USA.

Thanks to Kevin Colish, Brown Brothers Harriman, USA

Thanks to Richard Harran, IBM MQ Support, ENGLAND.
Change 34.271 New ANALDCO1 provides simple DCOLLECT reporting, using

ANALDCO1 the datasets created by JCLDAYDS. See examples in the

Nov 30, 2016 comments.
Change 34.270 -With PDB=RAWDATA and PDBOUT=WORK, a dataset not found was

VMXGDSN created when summarizing data that had been cleaned up

Nov 30,2016 prior to running the code to read tape data.

-Enhanced to allow you to suppress TAPEDATE by using new

TAPEDATA=null string.
====== Changes thru 34.269 were in this MXG 34.08 dated Nov 25, 2016====
Change 34.269 Support for INCOMPATIBLE TPX PTF R085512 and R085513 that

VMACTPX increased Port Number from 4 to 5 digits.

Nov 23, 2016

Thanks to Johanne Goulet, Government of Quebec, CANADA.

Thanks to Christian Roy, Government of Quebec, CANADA.
Change 34.268 Truncated POEX record with only 54 bytes caused STOPOVER.

VMACPOEX Now, the OFFSET just read is compared with LENGTH and the

Nov 22, 2016 first bad record is reported in the log and all deleted.

Thanks to Scott Wiig, USBank, USA.

Change 34.267 Support for RACF OIMID Token creates TOKOIMIC variable

VMAC80A and support for RACF LTL Token creates TOKMLTL variable

Nov 21, 2016 in TYPE80TK dataset.

Thanks to Mark Tomlinson, Lloydsbanking, ENGLAND.

Change 34.266 WPS Only. %MACRO VMXGPRNT invocation with text on col 72

ANAL113 caused error with WPS 3.00 (03.00.02.00.29316 but was

Nov 20, 2016 parsed correctly with WPS 3.02 (03.00.03.00.016221).

The error had ASIC03 for what should have been BASIC03.

Shortening the line did not eliminate the error.
Change 34.265 ANALDB2R could require //PDB DD because updates to READB2

ANALDB2R for DB2SBP and DB2NET had incorrect tokens that should be

READDB2 WORK. MXG code is now corrected, but circumvent with:

Nov 20, 2016 //PDB DD UNIT=SYSDA,SPACE=(CYL,(500,500),DISP=(,PASS)

on temp DASD, since they were not intended to be kept.

Thanks to John Ordman, Wipro, USA.

Change 34.264 -The section of code creating the prior TREND database was

VMXGALOC incorrect and failed to create the directory on the first

Nov 23, 2016 VMXGALOC execution resulting in an error.

-A FORCEDAY test did not have both side's UPCASEd.

Thanks to Job Varkey, VERISK, USA.

Thanks to Patricia J. Jones, DST, USA.

Change 34.263 SMF 6 CA/Dispatch records were increased to LENGTH=371,

IMACCADI adding two new fields CADIDES2 CADICHAR INCOMPATIBLY due

VMAC6 to MXG tests for LENGTH=347 to detect the V10 vs V11 data

Nov 18, 2016 records; those tests for LENGTH and SMF6LEN are unneeded

now and are removed.

Thanks to Glen Bowman, Wakefern, USA.

Change 34.262 -Support for z/OS 2.1+ PARMDD= EXEC statement JCL

ADOCRMFV parameter, enhanced SYSIN DD support, other improvements

ASMRMFV and fixes.

Nov 19, 2016 -PARMDD= is a new JCL parameter available with z/OS 2.1

and up. PARMDD= specifies a ddname of a file containing

parameters to be passed to the invoked program coded with

PGM=.

-If the ddname does not exist within the JCL step a

JCL error occurs with the message:
IEF689I JOB jobname FAILED PARMDD DID NOT OPEN
-The ddname may reference a physical sequential file,

PDS/PDSE member, or subsystem DD * or DD DATA data set.

-Unlike the 100 character PARM= parameter limit the

PARMDD= file may contain up to 32760 characters after

data assembly. If this limit is exceeded a JCL error

occurs.

-During data assembly of the PARMDD=ddname file trailing

blanks are stripped from each record and entirely blank

records are discarded by z/OS.

-Also during data assembly PARMDD=ddname fixed records are

checked for sequence numbers and also stripped if found.

There is no sequence number checking for variable

records.

-Any data stripped from a PARMDD=ddname file does NOT

count towards the 32760 character limit.

-The PARM= and PARMDD= parameters on the JCL EXEC

statement are mutually exclusive. If both are coded,

a JCL error occurs with message:

IEFC009I KEYWORD PARMDD IS MUTUALLY EXCLUSIVE WITH

KEYWORD PARM ON THE EXEC STATEMENT

-The PARMDD= parameter supports any RECFM of F, FB, V, or

VB. RECFM=U and Spanned records are not supported.

-The PARMDD=ddname file may have an LRECL up to 32760 for

fixed length records or 32756 for variable length

records.

-The PARMDD=ddname file may be concatenated with other DDs

in accord with usual concatenation rules.

-These are all valid examples of PARMDD=ddname usage:

//stepname EXEC PGM=ASMRMFV,PARMDD=ddname

//ddname DD DISP=SHR,DSNAME=dsname

//stepname EXEC PGM=ASMRMFV,PARMDD=ddname

//ddname DD DISP=SHR,DSNAME=dsname(member)

//stepname EXEC PGM=ASMRMFV,PARMDD=ddname

//ddname DD *

//stepname EXEC PGM=ASMRMFV,PARMDD=ddname

//ddname DD DATA

-For further details see Section 29 "PARMDD=ddname

Support" in the ASMRMFV source or ADOCRMFV members.

-Similar to PARMDD=ddname usage the ASMRMFV SYSIN DD (or

alternate ddname) now supports RECFM FB, F, VB, or V.

RECFM=U and Spanned records are not supported. Prior to

this change only RECFM=FB or RECFM=F was allowed.

-The ASMRMFV SYSIN DD (or alternate ddname) LRECL may

range up to 32760 for fixed length records or 32756 for

variable length records. Prior to this change only

LRECL=80 was allowed.

-A new built-in alternate ddname for SYSIN named SYSINA

may be provided in JCL and will be used instead of SYSIN

if found. It is not necessary to code the SYSIN=SYSINA

ASMRMFV parameter to use this alternate. However, any

other alternate ddnames require SYSIN= in the PARM= field

or PARMDD=ddname file.

-The order of ddname selection precedence for SYSIN is::
1) The SYSIN=ddname parameter in either the JCL PARM=

field or PARMDD=ddname file if present.

2) The //SYSINA DD in JCL if present.

3) The //SYSIN DD in JCL if present.

-A new built-in alternate ddname for SYSPRINT named

SYSPRINA may be provided in JCL and will be used instead

of SYSPRINT if found. If both SYSPRINA and SYSPRINT are

present SYSPRINT is ignored. There is NO SYSPRINT=ddname

parameter because the ASMRMFV log must be opened well

before any parm processing.

-When processing a PARMDD=ddname or SYSIN (or alternate)

files with data exceeding 100 characters in length, the

data is displayed in the existing RMFV002I message in 100

character sections.

-The first and last RMFV002I sections are always shown,

but any intermediate blank sections are not displayed.

The rightmost column for each section display shows the

number of characters remaining to be shown.

-SYSTSIN and SYSPRINA are added reserved ddnames when

SYSIN=ddname is specified.

-RECFM and LRECL are validated for all SYSIN (or

alternate) data sets.

-After MXG Change 34.226 the SYSIN OPEN subroutine

incorrectly attempts to obtain the DSCB for a //SYSIN DD

DUMMY statement. This results in Abend U0998 Reason Code

0018 and has been corrected.

-MXG Change 34.226 incorrectly altered VSAM TESTCB macro

results test for an VSAM RRDS type data set causing a

VSAM KSDS to be accepted as valid as an RMF III data set.

This caused an I/O error on the first read with Abend

U0998 Reason Code 0029 and has been corrected.

-RMFV008I DATASET LAST OPEN and RMFV009I ORIGIN messages

can be missing from ASMRMFV log for some RMF III data

sets after MXG Change 34.133 and this is also corrected.

-Message RMFV056S is now issued when PATTERR=ABEND instead

of RMFV056E as this is considered a severe error.

-SYSIN=ddname processing now correctly issues message

RMFV004E instead of RMFV056E.

-Incorrect test for '*/' end of imbedded comment string

fixed.

-Expand RMFV005E message to contain first 100 characters

of a bad parameter up from 80 as maximum that will fit

within the 126 character WTO text limit. If the

parameter in error exceeds 100 characters only the first

100 characters are shown.

-Correct SRST search handling for parameter strings

exceeding 256 bytes in length.

-Add short problem description text to RMFV005E message

if displayed parameter length will allow.

-Change '=' character search in keyword parameter

processing in PARMS subroutine to use SRST instruction

for better performance.

-FINAL subroutine setting Return Code 0016 when only

warnings for RED Invalid Processor and SPG Internal

error exist is fixed to issue Return Code 0008.

-RMFV018S SYNAD I/O error message loop can result after a

subsystem DD * data set for SYSIN (or an alternate)

specifies an LRECL other than 80 in JCL. z/OS apparently

continues to call the SYNAD routine for the same error

repeatedly with a WRONG LENGTH RECORD indication. The

problem has been circumvented.

-RMFV007S message was not always showing Reason Code of

blanks when the Reason Code is not available for the

service in error.

-Documentation Section 17 is retitled to "U0998 Abend

Reason Codes".

-Documentation Section 19 "Output LRECL" is retitled as

"Input and Output" LRECL.

-Former documentation Section 29 Summary is now Section

30.

-Former documentation Section 30 Bibliography is now

Section 31.

-New documentation Section Section 29 "PARMDD=ddname

Support" added.

-Updated following documentation sections for alternate

SYSIN/SYSPRINT and PARMDD=ddname support:

Section 3 "Execution JCL"

Section 5 "Input Data Selection Parameters"

Section 6 "Report Control Parameters"

Section 9 "JCL and SYSIN Parameter Usage"

Section 10 "Parameter Syntax Rules"

Section 11 "Parameter Coding Examples"

Section 12 "Messages"

Section 15 "Program and IBM Limitations"

Change 34.261 Support for Thales Security Record Version x.y INCOMPAT.

VMACTHAL This update supports records with and without subtypes.

Nov 17, 2016 For mapping by record ID, you must define these macros

with YOUR SMF record Ids, either in your IMACKEEP member

or in a %LET MACKEEP= argument in your //SYSIN:

%LET MACKEEP=

%QUOTE(

MACRO _IDTHALS 195 % /*SMF ID FOR SUMMARY RECORD*/

MACRO _IDTHALN 196 % /*SMF ID FOR SNAPSHOT RECORD*/

MACRO _IDTHALC 199 % /*SMF ID FOR CDS RECORD*/

MACRO _IDTHALE 198 % /*SMF ID FOR EXCEPTION RECORD*/

MACRO _IDTHALV 197 % /*SMF ID FOR SECURITY RECORD*/

MACRO _IDTHALR 194 % /*SMF ID FOR RESPONSE*/

);
To process records with SUBTYPES, you need these macros

either in IMACKEEP or with %LET MACKEEP= in //SYSIN:
%LET MACKEEP=

%QUOTE(

MACRO _IDTHALX 200 % /*SMF RECORD ID FOR NO SUBTYPES*/

MACRO _SUBTHAL

IF SUBTYPE GT . THEN DO;

IF SUBTYPE=0 THEN ID=_IDTHALC;

ELSE IF SUBTYPE=4 THEN ID=_IDTHALE;

ELSE IF SUBTYPE=8 THEN ID=_IDTHALV;

ELSE IF SUBTYPE=12 THEN ID=_IDTHALS;

ELSE IF SUBTYPE=16 THEN ID=_IDTHALN;

ELSE IF SUBTYPE=32 THEN ID=_IDTHALR;

END;

%

);
Thanks to Randy Schlueter, FirstData, USA.

Change 34.260 Support for SMF 110 Subtype 2 STID=32 creates new dataset

EXCICLDY DDDDDD DATASET DESCRIPTION

FORMATS CICLDY CICLDY CICS LOADER PRIVATE LIBRARY

IMAC110 that is added in CICS/TS 5.4 OPEN BETA.

VMAC110

VMXGINIT

Nov 18, 2016

Thanks to Harald Seifert, HUK-COBURG, GERMANY.

Change 34.259 Three JCL Test examples still had Includes of TYPSQAPM

JCLTEST9 which was replaced by TYPSQACS for the AS/400.

JCLTESS9

JCLTES92

Nov 14, 2016

Thanks to Earl Kline, Luminex, USA.

Change 34.258 Variable SMT_CORE_FLAG='LPARBUSY*VALID?' with value Y/N

VMAC7072 is now kept in TYPE70EN dataset.

Nov 14, 2016

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

Change 34.257 Support for TMON/CICS Version 4.1 (COMPATIBLE, no change

EXMONCSE to the existing MXG datasets) and support for the CS CTG

EXMONCSW records create three new datasets:

EXMONCSX DDDDDD DATASET DESCRIPTION

IMACTMO2 TMOCSE MONICSE LANDMARK CS-CSE SEGMENT

VMACTMO2 TMOCSW MONICSW LANDMARK CS-CSW SEGMENT

VMXGINIT TMOCSX MONICSX LANDMARK CS-CSX SEGMENT

Nov 13, 2016

Change 34.256 -DB2STATS variables QISEDPSC QISEDPSF QISEDPSL QISEDPSM

Many QVASBRPT QVASBRP QVASACEB QVASACEF QJSTDPXN QJSTDPXT are

Nov 11, 2016 kept and labeled and QWHCJOBSTEP is labeled.

-TYPE117 variables SM117NOR,SM117RSQ are labeled.

-TYPEIAM variables IAMACFL0-7,IAMACIN0-1,IAMCRIN0-8 and

IAMBOPCR IAMBUFCR IAMCOREO IAMCOREX IAMDDL IAMDSNL and

IAMRLSFP are labeled.

-TYPE42 variables SMF42FSH/FSI/FSJ/FSK correctly labeled.

-TYPE64 variable SMF64UTY extra asterisk removed in label.

-TYPE71 variables SMF71C3A/CPM/CPX extra asterisk removed.

-TYPE73 variables SMF73HEN and EXTENDSEG are now labeled.

-TYPE74 variables R748RAI is labeled in TYPE748R dataset.

-TYPE99 variables S99EE_CP_CHIPID S99EE_CP-BOOKID labeled.

Thanks to Chris Weston, SAS ITRM, USA.

Change 34.255 The z13 Topology Report had a typo Z!3 instead of Z13 and

ANAL9914 variable SYSTEM was added to the second report.

Nov 10, 2016

Thanks to Trevor Holland, ANZ, AUSTRALIA.

Change 34.254 CICS Dispatcher CICDS dataset variable DSGSRBT, SRB time

CICINTRV is now kept in the CICINTRV dataset.

Nov 9, 2016

Thanks to Randy Schlueter, FirstData, USA.

Change 34.253 Change 34.160 removed UPCASE function for BASExxx path

VMXGALOC names, but that segment was inadvertently deleted and is

Nov 9, 2016 restored. Impacted only Linux due to case sensitivity.
Change 34.252 SMF 119 Subtype 50 Dataset NUM11905 is NOT the count of

VMAC119 KEY segments, but is ALWAYS One, causing MXG to output

Nov 8, 2016 the first KEY. (And LEN11905 is the TOTAL length of all

KEY segments plus the 4 bytes for LEN/KEY itself).

The number of KEY segments is NOT provided, but MXG now

uses LENLEFT to find and INPUT and output to TYP11950 for

each KEY.

Thanks to Andreas von Imhof, Rabobank, THE NETHERLANDS.

Thanks to Ronald Kristel, Rabobank, THE NETHERLANDS.
Change 34.251 SMF 80 Top Secret records, format $MG080TS maps all of

FORMATS RDT table entries for variable TSFLCLAS.

VMAC80A Member VMAC80A has the SAS program in comments to update

Nov 9, 2016 the $MG080TS table.

Nov 15, 2016 Variable TSRESNAME contains the Resource Name.

Dec 6, 2016

Change 34.250 SMF 120 Subtype 100 ODM records had the order of two

VMAC120 variables, SM120RULEXFSUM/SM120RULEXCALLS reversed.

Nov 2, 2016

Thanks to Paul Volpi, UHC, USA.

Change 34.249 z/VM Linux Appldata datasets VXAPLSLM,VXAPLSLN,VXAPLSLP

EOAPLSLM had occasional large values; the logic to de-accumulate

EOAPLSLN was not reset for FIRST.VMDUSER. For these datasets that

EOAPLSLP are written for each interval for each user whether or

VMACVMXA not any resources were consumed, MXG only outputs an obs

Nov 1, 2016 when an interval had activity, and now the DURATM will

contain the actual duration since the last interval that

was output.

Thanks to Graham Harris, RBS, ENGLAND.
Change 34.248 DB2 Netezza IDAA Q8STxxxx variables were incorrectly

CLEARDB2 output in DB2STAT1/DB2STATS, which is a one instance per

EXDB2NET interval dataset, but there can be multiple Q8ST segments

EXDB2NET per interval (only the first segment was output).

IMACDB2 Now, ALL Q8STxxxx variables in DB2STATS/DB2STAT1 are set

READDB2 to a missing value, and the new DB2NETZA is created with

VMACDB2 one observation for each Q8ST segment.

VMXGINIT

Nov 1, 2016

Thanks to Erling Andersen, SMT, DENMARK.

Change 34.247 Support for APAR OA43999 RACF Database UNLOAD adds these

VMACRACF new variables to RACF0200 dataset:

Oct 28, 2016 USBD_PWD_ALG='ALGORITHM*USED TO*PROTECT*PASSWORD*/.

USBD_LEG_PWDHIST_CT='LEGACY*PASSWORD*HISTORY*ENTRIES*/

USBD_XPW_PWDHIST_CT='KDFAES*PASSWORD*HISTORY*ENTRIES*/

USBD_PHR_ALG='ALGORITHM*USED TO*PROTECT*PASSPHRASE*/.

USBD_LEG_PHRHIST_CT='LEGACY*PASSPHRASE*HISTRY*ENTRIES*/

USBD_XPW_PHRHIST_CT='KDFAES*PASSPHRASE*HISTRY*ENTRIES*/

and these overlooked RACF0200 variables are now created:

PWDENV_EXISTS='PASSWORD*PKCS#7*ENVELOPE*CREATED?'

PWD_ASIS ='EVALUATE*PASSWORD*ENTERED*CASD?'

PHRDATE ='EVALUATE*PASSWORD*ENTERED*CASD?'

PHRGEN ='PASSPHRASE*GENERATION*NUMBER'

CERT_SEQN ='PASSPHRASE*GENERATION*NUMBER'

Dataset RACF0560 missing values messages eliminated .

Thanks to Karl Lasecki, Chemical Abstracts Service, USA.

Change 34.246 RMF III dataset ZRBASI variable ASIDP did not exist in

VMACRMFV z/OS 1.13 nor z/OS 2.1, but the MXG test for ASIVERG3

Oct 27, 2016 used GE '16'x, causing ASIDP to contain characters. IBM

didn't document that the ASI record version was changed

to '1A'x (discovered only in 2.2 data records). Now,

ASIDP is missing in 1.13 or 2.1 and populated in 2.2.

Thanks to Randy Hewitt, HPE Enterprise Services
Change 34.245 MXG 33.08-34.07. SMF 42 dataset TYPE42L1 had misaligned

VMAC42 fields starting with SMF42HUA, causing these variables

Oct 27, 2016 to be wrong: SMF42HUA-SMF42HUG, SMF42HCA-SMF42HCX and

and SMF42HEH-SMF42HEK.

Thanks to Ann Knapik, Progressive Insurance, USA.

Thanks to David Buckmiller, Progressive Insurance, USA.

Thanks to William Keezer, Progressive Insurance, USA.

Thanks to Chris Weston, SAS ITRM, USA.

Change 34.244 A large number of regions could cause ARRAY EXCEEDED

UTILEXCL errors and/or invalid DCN/DRL test values in IMACEXCL if

Nov 9, 2016 an existing PDB.CICSDICT was appended with dictionary

Dec 14, 2016 records with the same SMF times. The MAX NREC value is

now created from PDB.CICSDICT and used for the new NREC

to separate those identical records.