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

Mar 17, 2016 were not printed. Superfluous code was removed.

Thanks to Chris Weston, SAS ITRM, USA.

Change 34.061 Support for BMC MAINVIEW FOR IP, creates these three

EXMVIP2C datasets of primary interest:

EXMVIP2F DDDDDD MXG MXG

EXMVIP03 DATASET DATASET DATASET RECORD

FORMATS SUFFIX NAME LABEL SUBTYPE

IMACMVIP

VMACMVIP MVIP2C TAC9I490 TN3270PERF 2C

VMXGINIT MVIP2F TAC9I350 SAWDATA 2F

Feb 26, 2016 MVIP03 TAC9I820 TACCONS 03

Mar 15, 2016

May 7, 2016 Note that IMACMVIP is tailored to only create these

May 30, 2016 three datasets.

Jun 1, 2016 Labels added Jun 1.

Aug 2, 2016 The MXG support for Mainview for IP requires the BMC

utility program BBM9MD73 to "dump" the BMC VSAM file

to a valid VB file that MXG can process.

Aug 2: SWSTOPTX corrected to local time.
CHANGE 34.060 ITRM ONLY, MXG 34.01 ONLY, CRITICAL ERROR because &PDB

VMXG70PR was used instead of &PDBMXG in two places; &PDBMXG has

Mar 15, 2016 always been the intended default macro variable for the

default "PDB" destination, and is required by ITRM.

Thanks to Chris Weston, SAS ITRM, USA.
CHANGE 34.059 Short type 119 subtype 41 with only one triplet populated

VMAC119 caused INPUT STATEMENT EXCEEDED. The record does not

Mar 14, 2016 contain any subtype 41 data. The first three instances

print a DELETED message on log.

==== Changes thru 34.058 were in THIRD MXG 34.01 dated Mar 14, 2016=====
CHANGE 34.058 MXG 34.01: CRITICAL: TYPE72GO only PERIOD 1 was output.

VMAC7072 Change 34.010 added MSU72 but used DO _I_= inside a DO

Mar 14, 2016 that already used DO _I_, which terminated the first DO.

Thanks to Randy Shumate, Reed Elsevier, USA.

=== Changes thru 34.057 were in SECOND MXG 34.01 dated Mar 14, 2016=====
CHANGE 34.057 Documentation only. Member JCLINSTT example has steps to

JCLINSTL FTP download, Unterse, and create USERID and FORMATS; new

JCLINSTT JCLINSTL example has only USERID and FORMATS create, if

VMXGCNFG you have already downloaded and untersed the new version.

Mar 11, 2016 Example DSNAMES are MXG.MXGVVNN.SOURCLIB/FORMATS in these

members and in the ftp instruction email text.

The install instructions stress that if you now depend on

the SAS NLSCOMPATMODE option to handle local characters

(British Pound, French accents, umlauts, etc.) you will

have to change your JCL for MXG to use your site's SAS

JCL procedure, with the CONFIMXG option, because SAS has

stated their intention to remove that option in a future

version.

See examples and comments in member VMXGCNFG.

//MXGSTEP EXEC SAS,CONFIG=MXG.SOURCLIB(CONFIMXG)

//MXGNAMES DD DSN=MXG.USERID.SOURCLIB(MXGNAMES),DISP=SHR

Thanks to Tom MacCabe, Dominion Resources Services, Inc., USA.
CHANGE 34.056 TYPE 60 dataset variable SMF60ELP was misaligned, VVRKEY

VMAC60 was not converted to EBCDIC text, and these address space

Mar 11, 2016 size/address variables are formatted MGBYTES

VVRDSHA VVRDSHU VVRHARBA VVRHURBA

VVRXEBA1-VVRXEBA5 VVRAMASP

Thanks to Michael Friske, FMR, USA.

CHANGE 34.055 TYPE 120 ST 9 new TYP1209R (REQUEST) and TYP1209N (ASYNC,

EXT1209N Non-Request) datasets completely replace four existing

EXT1209R TYP1209x datasets. The subtype 9 is either a REQUEST or

IMAC120 ASYNC event, and the two new datasets contain only the

VMAC120 variables that are appropriate for that event. These two

VMXGINIT new datasets eliminate the need to populate the A/C/E/S

Mar 11, 2016 datasets, which can be created with zero observations

Mar 16, 2016 by tailoring each _Edddddd "dataset output macro" to

Mar 21, 2016 replace the OUTPUT statement with a blank, either in your

Apr 11, 2016 //SYSIN for jobs that create TYPE120 datasets with

%LET MACKEEP=

MACRO _ET1209A % MACRO _ET1209C %

MACRO _ET1209E % MACRO _ET1209S %

;

or you can put the two macro lines in your IMACKEEP

member in your USERID.SOURCLIB tailoring PDS/directory.

Mar 21:

-The TYP1209U CPU detail dataset contains up to 20 obs

per event, and variable REQASYNC identifies the event,

and added variables identify the source of that event.

The CPU detail dataset metrics are summarized into the

TYP1209N or TYPE1290R dataset.

-The 1209C and 1209S segments have three or 12 obs per

event, so new variables SM1209ES1-SM1209ES3 and

SM1209EO1-SM1209E12 are created with those IDENTITY and

CLASSIFICATION values, eliminating any need for

TYP1209C/TYP1209S datasets. And new variable REC1209SEGS

identifies which segments were populated in each

TYP1209R/TYP1209N/TYP1209U observation.

-The order of segment processing was revised.

-The value of SM1209CI can be negative: That means that

the request didn't finish. The TCB CPU time at the start

is held, and the TCB CPU time at the end is subtracted to

get the SM1209CI value. If the servant abended or some

other bad thing happened and the request never finished,

there was no 'end time', so you get a negative value.

It's actually an indicator that something went wrong!

Apr 11: Labels for the CPU variables are clarified:

SM1209DA='ENCLAVE*TOTAL*CPUTIME'

SM1209DB='ENCLAVE*ZAAP*CPUTIME'

SM1209DC='ENCLAVE*ZAAP*ELIGIBLE*ON CP'

SM1209DD='ENCLAVE*ZIIP*ELIGIBLE*ON CP'

SM1209DE='ALWAYS*ZERO*QUALIFIED*CPU TIME'

SM1209DF='ENCLAVE*ZIIP*TIME*ON ZIIP'

SM1209HG='ENCLAVE*TOTAL CPUTIME'

SM1209HH='ENCLAVE*ZAAP*CPUTIME'

SM1209HI='ENCLAVE*ZAAP*ELIGIBLE*ON CP'

SM1209HJ='ENCLAVE*ZIIP*ELIGIBLE*ON CP'

SM1209HK='ALWAYS*ZERO*QUALIFIED*TIME'

SM1209HL='ENCLAVE*ZIIP TIME*ON ZIIP'

Thanks to Joesph Faska, DTCC, USA.

Thanks to Betty Wong, Bank of America, USA.
CHANGE 34.054 Variable SMF42LAN was not converted to EBCDIC after the

VMAC42 INPUT SMF42LAN $VARYING64. causing unprintable text.

Mar 10, 2016
CHANGE 34.053 BMC APPTUNE FIX BPU8604 caused INPUT STATEMENT EXCEEDED

VMAC102 error for subtype 8005x because the R8 triplet has R8N=1,

Mar 10, 2016 R8O=1512 with R8L=0 in a record that is only 1511 bytes

long. That INPUT is now skipped when the R8L is zero,

pending a correction from BMC.

Thanks to Rudi Claes, KBC, BELGIUM.

CHANGE 34.052 WPS Only, First MXG 34.01 Only. A typo RUN: with colon

VMXGINIT in line 3667 of VMXGINIT must be deleted as it caused

Mar 10, 2016 WPS to fail to initialize. I would have normally caught

this in my QA with that ABEND, but my SETINIT expired and

I had a condition code rather than an ABEND overlooked.
CHANGE 34.051 Change 33.240 updated MACRO _IO30TM but had replicated

IMAC30IO IOTM3390 causing WARINING: VARIABLE IOTM3390 EXISTS.

Mar 10, 2016 Delete the second IOTM3390.

Thanks to Randy Shumate, Reed Elsevier, USA.

CHANGE 34.050 Variables SM1209CM, SM1209CR, SM1209CS were incorrectly

VMAC120 kept in datasets TY1209C, TY1209S, and TY1209U, and were

Mar 9, 2016 used incorrectly in the _ST1209C and _ST1209U BY lists,

so combining multiple PDBs build with and without this

change could fail with a NOTSORTED error on either.

This Change was included in 34.055, above.

Thanks to Joesph Faska, DTCC, USA.
CHANGE 34.049 Support for ASG/TMON Version 4.0 for CICS, REQUIRED.

VMACTMO2 Version changed all duration fields from microsecond to

Mar 9, 2016 todstamp units, but these 15 variables were not divided

by 4096 (the other 399 were), so these variables will

have values larger by that 4096 factor.

CICOVHTM FILEIOTM TAARQRTM TAARQWTM TAAWTTTM TAAWTWRT

TADSPCPU TADSPDTM TADSPSTM TADSPWRT TATCBSTM TATCBSTM

TATCBSTM TIIWTWRT TMCGADT

Fortunately, none of these duration variables are in the

primary MONITASK dataset.

Thanks to Miguel Machin, CAREFIRST, USA

Thanks to Alan Gray, CAREFIRST, USA.

=== Changes thru 34.048 were in FIRST MXG 34.01 dated Mar 7, 2016======
CHANGE 34.048 Support for BE93 Version 6.1.0 (INCOMPATIBLE, header was

VMACBETA relocated). No new variables nor datasets.

Mar 7, 2016

Thanks to Rudolf Sauer, T-SYSTEMS INTERNATIONAL GmbH, GERMANY.

CHANGE 34.047 Support for z/OS 2.2 RMF III data records.

VMACRMFV -No change is needed for the ASMRMFV program that reads

Mar 6, 2016 the Compressed VSAM file to create the RMFBSAM data file.

-New variables in ZRBASI dataset:

ASICPUTA_LF ='CPU*TIME'

ASIDCTIA_S ='CHANNEL*CONNECT*TIME'

ASIDP ='DISPATCHING*PRIORITY'

ASIFRXA_LF ='FIXED*FRAMES*ABOVE'

ASIFRXB_LF ='FIXED*FRAMES*BELOW'

ASIFRXH_LF ='FIXED*FRAMES*HIGH'

ASIIOCNT_S ='EXCPS'

ASIQSCANRES ='QSCAN*RESOURCES*RETURNED'

ASIQSCANRESSQ1 ='QSCAN*ASIQ*SCANRES*SSQ1'

ASIQSCANRESSQ2 ='QSCAN*ASIQ*SCANRES*SSQ2'

ASIQSCANSPECREQ='QSCAN*SPECIFIC*REQUESTS'

ASIQSCANTIME ='QSCAN*REQUESTS*ISSUED'

ASIQSCANTIMESQ1='QSCAN*ASIQ*SCANTIME*SSQ1'

ASIQSCANTIMESQ2='QSCAN*ASIQ*SCANTIME*SSQ2'

ASITCBTA_LF ='TCB*TIME'

ASITRCA_S ='TRANSACTIONS'

ASITRT ='TRANSACTION*RESIDENT*TIME'

ASI_FREEMAINEDFRAMES='FREEMAINED*FRAMES'

ASI_HVSHRPAGEVALIDATIONS='PAGE*VALIDATONS*HI SHARE'

ASI_LVSHR1MGBYTES='HWM*HIGH*VIRTUAL*SHARED'

ASI_LVSHR1MNMOMB='SHARED*1M*MEMORY*OBJECTS'

ASI_LVSHR4KB ='SHARED*BYTES*HI VIRT'

-New variables in ZRBCPU dataset:

CPC_ATD_AAP ='AVERAGE*THREAD*DENSITY*AAP'

CPC_ATD_CP ='AVERAGE*THREAD*DENSITY*CP '

CPC_ATD_IIP ='AVERAGE*THREAD*DENSITY*IIP'

CPC_CAPF_AAP ='MT CORE*CAPACITY*FACTOR*AAP'

CPC_CAPF_CP ='MT CORE*CAPACITY*FACTOR*CP '

CPC_CAPF_IIP ='MT CORE*CAPACITY*FACTOR*IIP'

CPC_MAXCAPF_AAP='MT CORE*MAXIMUM*CAPACITYAAP'

CPC_MAXCAPF_CP ='MT CORE*MAXIMUM*CAPACITYCP '

CPC_MAXCAPF_IIP='MT CORE*MAXIMUM*CAPACITYIIP'

CPC_MODE_AAP ='MT CORE*MODE*AAP'

CPC_MODE_CP ='MT CORE*MODE*CP '

CPC_MODE_IIP ='MT CORE*MODE*IIP'

CPC_PROD_AAP ='MT CORE*PRODUCTIVITY*AAP'

CPC_PROD_CP ='MT CORE*PRODUCTIVITY*CP '

CPC_PROD_IIP ='MT CORE*PRODUCTIVITY*IIP'

CPU_PARK_CP ='PARKED*TIME*CP'

CPU_PARK_IFA ='PARKED*TIME*IFA'

CPU_PARK_ZIP ='PARKED*TIME*ZIP'

CPU_ONLINE_CP ='ONLINED*TIME*CP'

CPU_ONLINE_IFA ='ONLINE*TIME*IFA'

CPU_ONLINE_ZIP ='ONLINE*TIME*ZIP'

-New variables in ZRBGEI dataset:

GEILSMO ='MEM OBJ*HI VERT*BACKED IN*1MB FRAMES'

GEIRFREM='FREEMAINED*FRAMES*ALL ASIDS'
CHANGE 34.046 ASUMCACH now works without RMF III data and supports tape

ASUMCACH or disk. When the PDB is on tape, TYPE74 is copied to

Mar 5, 2016 //WORK to prevent having two open tape datasets.
Change 34.045 JES2 SMF 26 z/OS 1.13 TRIPLET segment before PRINT caused

VMAC26J2 INPUT STATEMENT EXCEEDED ERROR on z/OS, or a FLOATING

Mar 5, 2016 POINT EXCEPTION on ASCII. The unexpected order

misaligned the input of offset variable SMF26OJC to have

a value of 3,806,577,725, which then caused the error

when INPUT @SMF26OJC was executed with that large value.

The SMF manual has always shown all 7 segments at offset

50, so it's the order in the SMF manual that has

previously defined their order in the record. This

change heuristically detects the order of those two

segments.

-Change 33.046 added support for the SMF26JCR field in MXG

33.02 last year, and that new code failed on the reversed

records at this one site, where all records on some

systems were reversed, and all records on other systems

had the correct segment order.

Thanks to Rich Kuehn, Global eXchange Services, Inc., USA.
Change 34.044 New Capacity Group reports of CEC resources consumed by

ANALGRCA LPARS within a capacity group by RMFINTRV workloads,

GRAFWRKC reporting percent CPU, total CPU time, estimated hourly

Mar 6, 2016 software MSU, and memory consumption by workloads and

then by LPAR.
ANALGRCA and GRAFWRKC both report on Group Capacity.

A good place to start is gragwrkc - it will take your

RMFINTRV dataset and build a picture by CEC and Capacity

Group of %CPU busy, CPU time, estimated hourly MSU, and

memory with a pair of graphs for each, the first by your

RMFINTRV workloads, the second by LPAR within the

Capacity Group. You can narrow down the squeaky wheel

to a workload and/or lpar, and then you can use ANALGRCA

to fine tune the analysis.
ANALGRCA will do much the same thing but lets you zero-in

on what is pushing you to the cap or to a threshold you

specify. The threshold can be an absolute number of MSU

or a percentage of the total group capacity. It will look

at the LPARS in the group using ASUMCELP, the workloads

using RMFINTRV, and the SMFINTRV dataset to look at

tasks. There are parameters to specify the date to

examine, the interval to use (but it must be the same as

the cecintrv in asumcelp.) and for workload and job level

data only those intervals that exceeded the threshold are

used in reporting.
CHANGE 34.043 MXG 33.33. Change 33.316 missed the four reserved bytes,

VMAC71 causing SMF71CPx, SMF714Kx, & SMF71PLx to be misaligned.

Mar 2, 2016

Thanks to Rick Southby, Insurance Australia Group, AUSTRALIA.

CHANGE 34.042 Support for DCOLLECT FLAG4 bits that indicate each

VMACDCOL size variables are now in MegaBytes, previously in

Mar 7, 2016 KiloBytes, in DCOLBKUP and DCOLMIGS datasets.

Mar 8, 2016 -Mar 8. UBALLSP and UMALLSP 024* changed to 1024*.

Mar 22, 2016 -Mar 22. All ten UBFLAG and UMFLAG bit tests corrected.

May 8, 2016 -May 8: Change 34.112 corrected UBDSIZE high by 1024.

Thanks to Scott Barry, SBBWorks, Inc., USA.
CHANGE 34.041 Support for ATF V531 Enhanced Summarization Phase 2

FORMATS inserted these new variables:

VMACATF ATFPGMSW ='PROGRAM*SWITCHES'

Mar 6, 2016 ATFXSNOTN='OTHER*ITEMS**VAR'

Mar 9, 2016 ATFXSNOTL='OTHER*ITEMS**LENGTH'

Apr 5, 2016 ATFXSUOW ='TRANSACTION*UOW'

and the DLI-DB/DLI-TM/DBD/DB2/MQ/OTHERA/OTHERB segments

have new GROUP BUCKET NUMBERs and/or ITEM CODES that are

decoded in new FORMATS.

-Mar 9: (after 34.01) Short 8-byte DBD supported, DLI DB

and DLI TM and OTHER-A segments have been validated.

-Apr 5: Variables ATFXSRSP ATFXSACP ATFXSIQT were wrongly

divided by 4096 twice.
CHANGE 34.040 TYP11921 variable NTHOSTTN is increased from $8 to $64 to

VMAC119 support host names that are fully qualified TCP/IP domain

Mar 2, 2016 name.

Thanks to Stephen S. Hoar, Lloyds Banking, ENGLAND.

Thanks to Gary Nash, Lloyds Banking, ENGLAND.
CHANGE 34.039 Support for new SMF Type 29 IMS JAVA CPU and Garbage Coll

EXTY29GC creates new datasets

EXTY29JA DDDDDD MXG MXG

IMAC29 DATASET DATASET DATASET

TYPE29 SUFFIX NAME LABEL SUBTYPE

TYPS29

VMAC29 TY29GC TY29GC IMS JVM GARBAGE COLLECTION 2

VMXGINIT TY29JA TY29JAVA IMS JVM CPU USAGE 2

Mar 2, 2016 See Change 34.221 which revised and validated with data.

CHANGE 34.038 New variables created in TYPE1209 dataset:

VMAC120 SM1209HE='ENCLAVE*JOINED OR*CREATED?'

Mar 1, 2016 SM1209HF='ENCLAVE*SCHEDULED?'

Thanks to Joseph Faska, DTCC, USA.
CHANGE 34.037 Using %LET MXGREADSMF=BOTH caused ERROR 181-185 VARIABLE

VMACSMF SMFINFILE already exists. Code revised to use LOGINFILE

Feb 28, 2016 variable for the INFILE LOGGER.

Thanks to Chris Weston, SAS ITRM, USA.

CHANGE 34.036 TYPE30_5 dataset can have ABEND='SYSTEM' CONDCODE=0000 if

FORMATS a step had a SYSTEM or USER ABEND, but the last step did

VMAC30 not ABEND (e.g., a FLUSH step followed the ABEND). Since

Mar 1, 2016 the type of ABEND is unknown in the TYPE30_5 JOB record,

MXG now sets ABEND='ABEND' instead of ABEND='SYSTEM' in

TYPE30_5. However, in PDB.JOBS, MXG populates both ABEND

and CONDCODE from the LAST step that ABENDed, so you will

not see ABEND='ABEND' except in TYPE30_5. And, since it

is really STEPS that ABEND, and not JOBS, you should use

the PDB.STEPS or TYPE30_4 for ABEND analysis.

Thanks to Linda S. Berkley, USPS, USA.
CHANGE 34.035 These SYNCSORT variables are now kept, formatted $HEX2:

VMACSYNC SYNRETRY='RETRY*FLAG'

Feb 25, 2016 SYNMISCF='SMFFLAG3*MISC*FLAG'

Thanks to Bruce Bordonaro, Pershing, USA.

CHANGE 34.034 Reserved Change Number.

Feb 28, 2016

CHANGE 34.033 Change 31.118 added new fields in the EXGRXEXT (Extended

VMACEDGR Record), but those fields are also in the basic dataset

Feb 28, 2016 records EDGRDEXT and the volume records EDGRVEXT, and

this change adds them to those two datasets.

Thanks to Thomas Giordano, Australian Defence Department, AUSTRALIA.
CHANGE 34.032 Support for DB2 Trace IFCIDS 311 and 321.

VMAC102

Feb 24, 2016
CHANGE 34.031 Cosmetic. If your _IMSVERS does not match the version in

VMACIMS the IMS56FA record, the previous MXGNOTE is now MXGWARN.

Feb 24, 2016 and the text is clearer when they do match.
CHANGE 34.030 TYPE42D4 DATASET variables SMFA2GTAA & SMFA2GTAB are now

VMAC42 correctly INPUT and kept, replacing incorrectly spelled

Feb 22, 2016 SMFA2GSA and SMFA2GSB, with the CA and CI Splits.

Thanks to Michael Friske, FMR, USA.

CHANGE 34.029 Variables SMF70GNM and SMF70GMU added to PDB.RMFINTRV to

VMXGRMFI enable reporting of workloads by Capacity Group.

Feb 22, 2016
Change 34.028 MXG 33.33. ASUM70PR Change 33.306 required PDB.TYPE70 to

VMXG70PR exist in the //PDB data library, so the group capacity

Feb 22, 2016 metrics could be created, but it was not documented that

that the TYPE70 dataset was required. Normally TYPE70 is

in the PDB with the TYPE70PR dataset, but if TYPE70 was

not found in the //PDB library, the ASUM70PR failed.

Now VMXG70PR verifies that TYPE70 exists and uses it if

found, or does not read it if not found, which causes the

new Group Variables to not exist in the ASUM output.

Or, with 33.33 you can circumvent the error using this

code ahead of your ASUM70PR include:

%LET MACKEEP=%QUOTE( MACRO _LTY70 NEWPDB.TYPE70 % );

Change 34.027 MXG set the TYPE1131 CPU Speed SM1132SP to 5000 for z13

VMAC113 because early records contained either 5208 for CP or had

Feb 22, 2016 ZERO for the IP speed. But that is now wrong with the new

Feb 29, 2016 sub-capacity z13s (Speed=3173), so the logic now forces

the 5000 value only if the record contains zero or 5208.

Thanks to Andrew Hebden, Barclays, ENGLAND.

Change 34.026 Support for MVMQ (BBMQ) PTF BPL2558 which was to change

VMACBBMQ all duration fields from TODSTAMP to microseconds, but

Feb 22, 2016 BMC now reports that pre and post that PTF, the values

Mar 10, 2016 have always been in seconds with microsecond fractions.

Mar 22, 2016 MXG was dividing by 4096 presuming TODSTAMP, which is

known to have always been wrong and is now corrected.

Apparently, the primary use has been counts and events

and not durations, as no one noticed!

-Mar 10: Cosmetic typo QSCLOSETTIME to QSCLOSTTIME.

-Unrelated, but observed, that all character date/times

were on GMT zone; new datetime numeric variables replace

them and are set to the local time zone.

Change 34.025 Dataset TYPE70EN, variable PCTMVSBY=100 was incorrectly

VMAC7072 calculated when SMF70PAT, Parked Time, was close to the

Feb 12, 2016 DURATM, but did not exactly match. A 1 second delta is

now required to calculate a non-zero PCTMVSBY=0.

Thanks to Giuseppe Giacomodonato, EPVTECH, ITALY.
Change 34.024 Dataset TYPE72GO, variable RESPSTD, Standard Deviation of

VMAC7072 Response time was 2% larger than IBM Value; the 1.04875

Mar 1, 2016 conversion factor is corrected to 1.024.

Thanks to Richard Stuchell, VISA, USA.

Change 34.023 Parameter list was alphabetized and the parameters used

UTILBLDP are displayed on the log. A check was added to the list

Feb 12, 2016 of parameters where an = should not exist as that is

normally an indication that a comma was left off of one

of the parameters.
Change 34.022 -DB2STAT2 statistics dataset new variables:

VMACDB2 QDBPFRAM='FRAMESIZE'

Feb 11, 2016 QDBPVPMI='VPSIZEMIN'

Feb 28, 2016 QDBPVPMA='VPSIZEMAX'

Mar 23, 2017 QDBPSPSZ='SIMULATED*BUFFER*POOL*SIZE'

QDBPSPST='SIMULATED*SEQ*THRESHOLD'

-DB2ACCTP Package dataset, new flag variable

QPACINCO='INCOMPATIBLE*FUNCTION?'

-Documentation: In DB2ACCTP, if QPACRUSM='Y', these fields

are listed by IBM as invalid:

QPACCRNT QPACINSP QPACPAC QPACPKNM QPACCOLN QPACPKID

QPACCONT QPACSCB QPACSCE QPACBJST QPACEJST QPACASCH

QPACAANM QPACAAFG QPACINCO

-Added Mar 23, 2017:

Existing Rollup flag variable DB2PARTY='R' is also set

if QPACRUSM, the new Rollup Summary flag variable is set.

When DB2PARTY='R', the four datetime variables above

QPACBJST QPACEJST QPACSCB QPACSCE

are always zero for BJST/EJST and missing for SCB/SCE.

Those were the individual package event datetimes that

are lost with DB2 ROLLUP summarization.

Thanks to Scott Barry, SBBWorks, Inc., USA.

Change 34.021 CICSTRAN variable TRANFLAG creates variables from each

FORMATS byte, some are decoded with new formats:

VMAC110 TRANFLAGTF='TRANFLAG*TRANSACTION*FACILITY*NAME'

Feb 12, 2016 VALUE $MGCICTF

'0'='BIT0:NONE'

'1'='BIT1:TERMINAL'

'2'='BIT2:SURROGATE'

'3'='BIT3:DESTINATION'

'4'='BIT4:3270 BRIDGE'

;

TRANFLAGID='TRANFLAG*TRANSACTION*IDENTIFICATION'

VALUE $MGCICTI

'0'='BIT0:SYSTEM TRANSACTION'

'1'='BIT1:MIRROR TRANSACTION'

'2'='BIT2:MIRROR TRANSACTION DPL'

'3'='BIT3:ONC RPC ALIAS TRANS'

'4'='BIT4:WEB ALIAS TRANSACTION'

;

TRANFLAGWL='WLM*STATUS'

VALUE $MGCICWL

'0'='BIT0:WLM REPORT'

'1'='BIT1:WLM NOTIFY COMPLETION'

'2'='BIT2:WLM NOTIFY'

;

TRANFLAGIN='INDOUBT*STATUS'

VALUE $MGCIC8I

'0'='BIT0:INDOUBT WAIT=NO'

'1'='BIT1:INDOUBT ACTION=COMMIT'

'2'='BIT5:INDOUBT FAILURE'

;

TRANFLAGUO='UOW*STATUS'

VALUE $MGCIC8U

'0'='BIT2:UOW INDOUBT ACTION'

'1'='BIT3:UOW SHUNT'

'2'='BIT4:UOW UNSHUNT'

'3'='BIT5:INDOUBT FAILURE'

;

TASKDATALOC='TASKDATALOC*BELOW?'

TASKDATAKEY='TASKDATAKEY*CICS?'

TASKISOLATE='TASKISOLATE*NO?'

TASKDYNAMIC='TASKDYNAMIC*YES?'

Thanks to Perry Lim, Union Bank, USA.

Change 34.020 PK-ZIP SMF records INPUT STATEMENT EXCEEDED error due to

VMACPKSZ one field expanded from 2 to 4 bytes.

Feb 10, 2016

Thanks to Dorothy Yeung, Toyota, USA.

Change 34.019 -STC Variables STC26MST & STC26MET were on GMT time zone,

FORMATS now corrected to local to match other datetimes.

VMACSTC -STC variables STCxxADR were $HEX8 formatted, but that