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

devices (66,816 rows):

PROC ACCESS convert SYBASE to SAS: 265 seconds

or

BCP dump SYBASE to ASCII file 471 seconds

INFILE/INPUT read BCP dump 21 seconds

Thanks to Bernie Davidovich, Predictive Systems, USA.

Change 14.217 DB2 4.1 variables QTGA and QBGA were trashed, because the

VMACDB2 offsets were input from the wrong location. After test

Sep 10, 1996 IF QWHSNSDA GE 11 THEN ... the INPUTs with @81,85,87,89,

93, and 95, should be 97,101,103,105,109,111.

Thanks to Bertolotti Emmanno, Banca Commerciale Italiano, ITALY

Thanks to Burani Daniela, Banca Commerciale Italiano, ITALY

Change 14.216 Three new utilities are now available for Beta testing:

UTILCONT UTILCONT produces a PROC CONTENTS-like output with the

UTILFATD size (MegaBytes and observations) of each data

UTILSIZE set in the library pointed to by the PDB DDna.

Sep 10, 1996 UTILFATD reads the directory of all of your FAT disks

connected to this PC and calculates how much

space the data would take on hard disks of

varying cluster size (512 to 32K).

UTILSIZE reads an MVS PDS to calculate how much space

is needed to put that PDB library on a FAT

disk on your PC for all valid Cluster Sizes.

Large Cluster sizes may waste large amounts of

space when you have many little files. This

is useful to predict how much free space you

will need on your PC to store the MXG source

library.
Thanks to Chuck Hopf, MBNA, USA.

Change 14.215 This is the Beta Version of the new MXG DASD Allocation

ASMDALO Monitor, which creates an SMF record for each of these

EXDALOAL DADSM events:

EXDALOEX DALOALOC ALLOCATE

EXDALOPA DALOSCRA SCRATCH

EXDALOPP DALORENA RENAME

EXDALORE DALOEXTN EXTEND

EXDALOSC DALOPART PARTIAL RELEASE

EXDALOVS DALOPPRL PARTREL PARTIAL RELEASE

IMACDALO DALOVSAM VSAM EXTEND WITHOUT A DEB

TYPEDALO and MXGDALO can be enabled for selected devices and/or

VMACDALO selected datasets. This monitor can be used to identify

Sep 9, 1996 jobs that are living on DASD extents (so you can increase

their primary allocation to avoid exposure to B37/E37

ABENDS. This monitor can be used to track (and charge?)

usage of temporary DASD space. Other uses are expected.

Read the comments in ASMDALO before you proceed.
Change 14.214 Variable SERVUNIT (total service units) was stored in 4

BUILDPDB bytes, whick keeps only 7 significant digits; a large

BUILDPD3 value in SERVUNIT (714 million units) was truncated by

BUIL3005 1000 units because 5 bytes storage is required to exactly

BUILD005 represent a numeric field that was input from four binary

VMAC30 bytes. SERVUNIT is now stored in 5 bytes. But the real

VMAC434 error here was the site's MSOCOEFF=3.0, way too large.

VMACTSOM An MSOCOEFF=.0001 is MXG's recommendation, so that the

Sep 10, 1996 MSO units do NOT contribute to SERVUNIT - only CPU time

Mar 10, 2003 and I/O operations are real computer work - and then the

value of SERVUNIT can be used to measure capacity. The

CPU+I/O service was 31 million units, showing the large

MSOCOEFF made the SERVUNIT field meaningless!

Mar 10, 2003: Original text has .000001 but .0001 is the

minimum value, other than zero.

Feb 21, 2005: See MXG Technical Note in Newsletter 47;

IBM Discovered that the smallest value is actually .0122.

Any value less than that value will be reported in the

MSOCOEFF in TYPE72GO, but only 0.0122 will be used!
Change 14.213 Support for NETVIEW 3.1 compatible changes to type 37

VMAC37 record added two fields, BRFEDCP1 and BRFPCCP1 to the

Sep 8, 1996 generic event report. While these fields were added

compatibly, MXG Version 13.13 failed with NETVIEW 3.1

because of MXG logic error that was fixed by MXG 14.090.

Thanks to Herb G. Strozinsky, Burlington Northern Sante Fe, USA.

Change 14.212 CICS/ESA 4.1+ GMT offset MCTMNTAD is not exact hours from

VMAC110 GMT (examples were off by 3.15, 0.9, and 0.6 seconds).

Sep 6, 1996 IBM lists APARs OW02410,OW10435,OW14651,OW14350,OW15036

and (most likely) OW15111 as the corrections, but the IBM

error is now circumvented by using the equation:

MCTMNTAD=3600*FLOOR((MCTMNTAD+900)/3600);

to exactly set the GMT offset to the nearest hour.

With CICS 4.1 and later, MXG uses MCTMNTAD to convert the

CICSTRAN variables STRTTIME and ENDTIME to local time,

so the overall impact was minimal, unless you tried to

compare CICS transaction start/stop times with TPX logon

times; the CICS error had transactions starting before

the terminal had logged on to CICS!

Thanks to Pat Quinnette, Principal Mutual Life Insurance, USA.

Thanks to Wolfgang Vierling, Vereinte Versicherungen, GERMANY.
Change 14.211 CICS/ESA 4.1+ does not create obs in STID=4 statistics

CICINTRV dataset CICTSR, causing variable A02TTA in CICEODRV and

Sep 6, 1996 CICINTRV to be missing. However, the transaction counts

are in variable XMRAC in CICXMR, which was previously not

summarized in CICINTRV, but with this change now is, and

if A02TTA is missing, the value of XMRAC will be stored

into A02TTA so your old reports will be protected. Both

A02TTA and XMRAC are kept in CICINTRV/CICEODRV.

Thanks to Tracy Blackstone, Kaiser Permanente, USA.
Change 14.210 Datasets PDB.RMFINTRV, PDB.JOBS, PDB.STEPS, PDB.PRINT

BUILDPDB PDB.SMFINTRV, PDB.SPUNJOBS and (JES2 only) PDB.NJEPURGE

BUILDPD3 are built in sort order, but because they are built in a

BUILD002 DATA step (rather than directly by PROC SORT, as are most

BUILD003 other PDB datasets), SAS does not know they are sorted.

BUILD004 By specifying the dataset option SORTEDBY= on the DATA

BUILD005 statement, SAS now knows the sort order, so that it can

BUIL3005 bypass unnecessary sorting. In making this change, I

IMACPDB also finally removed the archaic _PROTECT token from all

RMFINTRV of the BUILDPDB members; defined in IMACPDB, PROTECT was

SPUNJOBS a SAS version 5 facility that is no longer needed, (see

Sep 5, 1996 comments in member IMACPDB), that defaults to null with

SAS version 6, and to have preserved it and to have added

SORTEDBY= would have required an INCOMPATIBLE change in

member IMACPDB. While the references to _PROTECT in the

BUILxxxx members was removed, the macro itself is still

defined in member IMACPDB to avoid a different

INCOMPATIBILITY (in case you have actually modified

BUILDPDB itself), but the comments in IMACPDB were

revised to clarify its non-use!

Thanks to Carl Sommer, SAS Institute Cary, USA.
====Changes thru 14.209 were printed in MXG Newsletter THIRTY==========

====and were in the early MXG 14.07 tapes dates September 10, 1996=====

Change 14.209 Support for CICS/ESA 5.1.0, now known as CICS Transaction

EXCICLGR Server for OS/390 Release 1, which INCOMPATIBLY inserted

EXCICLGS 19 new fields in the CICSTRAN Transaction Accounting data

EXCICNQG (had the fields been appended instead of inserted, this

EXCICRMG would be compatible!), added 2 new fields to CICSEXCE,

EXCICTQR added new fields in 8 existing Statistic Datasets, and

FORMATS five new Statistics datasets are created from new STIDs.

IMACCICS STATUS: IBM documentation of STID=82 (MNR) record is

IMACEXCL wrong; that STID is skipped without message in this code.

VMAC110 This version has been tested with real data records!

Sep 10, 1996 This change also supports the YYYY versus YY value in the

Statistics subtype's header, and the expanded TRANTYPE

(two bytes, more meanings decoded).

New variables added to CICSTRAN (transaction) dataset:

FCTYNAME='TRANSACTION*FACILITY*NAME'

LOGWRTCT='CICS*LOG STREAM*WRITE REQUESTS'

PERRECNT='PER RECORDS*WRITTEN*FOR TASK'

RMUOWID ='RECOVERY*UNIT OF WORK*ID'

RPT

SC24FSHR='SHARED BYTES*FREEMAINED*BELOW 16MB'

SC24GSHR='SHARED BYTES*GETMAINED*BELOW 16MB'

SC24SGCT='SHARED*GETMAINS*BELOW 16MB'

SC31FSHR='SHARED BYTES*FREEMAINED*ABOVE 16MB'

SC31GSHR='SHARED BYTES*GETMAINED*ABOVE 16MB'

SC31SGCT='SHARED*GETMAINS*ABOVE 16MB'

SRV

TERMCNNM='TERMINAL*SESSION*CONNECTION*NAME'

TERMINFO='TERMINAL*OR SESSION*INFO FLAGS'

TRANFLAG='TRANSACTION*FLAGS'

WTRLIOCM='RLS*IO WAITS'

WTRLIOTM='RLS*IO WAIT*DURATION'

WTSYIOCN='SYNCPOINT*IO WAITS'

WTSYIOTM='SYNCPOINT*IO WAIT*DURATION'

New variables added to CICSEXCE (exception) dataset:

EXCMNTCN='TRANSACTION*CLASS*NAME'

EXCMNSRV='SERVICE*CLASS*NAME'

EXCMNRPT='REPORT*CLASS*NAME'

New variables added to CICDS (Dispatcher, by TCB Stats)

for ONC/RPC Mode (5th) TCB and File Owing Mode (6th) TCB

(STID=56):

DSnACT DSnPERCT DSnSCT DSnSTART DSnSWT DSnSYSW

DSnTCBF1 DSnTCBF2 DSnTCBF3 DSnTCBF4 DSnTCBNM DSnTCT

DSnTDT DSnTWT DSnLSTRT

New variables added to CICXMR (Transaction Manager):

(STID=11):

XMRAMISM='ACTION*MISMATCHES'

XMRFAIT ='FORCED*ACTION*INDOUBT*TIMEOUT'

XMRFANW ='FORCED*ACTION*NO WAIT ABILITY'

XMRFAOP ='FORCED*ACTION*OPERATOR'

XMRFAOT ='FORCED*ACTION*OTHER'

XMRFATXN='FORCED*ACTION*TRANDEF'

XMRIACTN='INDOUBT*ACTION*COMMIT OR*BACKOUT?'

XMRITOV ='INDOUBT*TIMEOUT*VALUE*MINUTES'

XMRIWAIT='INDOUBT*WAITS'

XMRIWTOP='INDOUBT*WAIT*OPTION?'

New variables added to CICCONSR (ISC/IRC stats):

(STID=52):

A14ACCM ='ACCESS*METHOD'

A14AICT ='LOCAL*CONNECTION*CREATE*TIME'

A14AIDT ='LOCAL*CONNECTION*DELETE*TIME'

A14EFLGS='PROTOCOL'

A14EPRMN='RECEIVE*SESSION*COUNT'

A14ESECN='SEND*SESSION*COUNT'

A14ESID ='CONNECTION*NETNAME'

A14E1RY ='PRIMARIES*CURRENTLY*USED'

A14E2RY ='SECONDARIES*CURRENTLY*USED'

A14GACT ='GMT*CONNECTION*CREATE*TIME'

A14GADT ='GMT*CONNECTION*DELETE*TIME'

New variables added to CICCONMR (ISC/IRC stats)

(STID=76):

A20ECONL='CURRENT*CNOS*CONTENTION*LOSERS'

A20ECONW='CURRENT*CNOS*CONTENTION*WINNERS'

A20ELMAX='MAX*SESSION*COUNT'

A20EMAXS='CURRENT*MAX*SESSION*COUNT'

A20EMCON='MAX*CONTENTION*WINNERS*ACCEPTABLE'

A20E1RY ='PRIMARIES*CURRENTLY*USED'

A20E2RY ='SECONDARIES*CURRENTLY*USED'

New variables added to CICFCR (File Control Statistics)

(STID=67):

A17DSBRU='BROWSE*FOR*UPDATES'

A17RLSWT='RLS*REQUEST*WAIT*TIMEOUTS'

New variables added to CICTCR (Term Control Statistics)

(STID=34):

A06GOFTM='AUTOINSTALL*LOGOFF*GMT*TIME'

A06GONTM='AUTOINSTALL*LOGON*GMT*TIME'

A06OFFTM='AUTOINSTALL*LOGOFF*LOCAL*TIME'

A06ONTM ='AUTOINSTALL*LOGON*LOCAL*TIME'

A06SYSID='OWNING*SYSID OF*TERMINAL*OR SESSION'

New variables added to CICDQG (TDQUQUE Global Stats)

(STID=45):

A11ACNAL='CURRENT*CONCURRENT*BUFFER*ACCESS'

A11ACNWT='CURRENT*BUFFER*WAITS'

A11ACNIU='CURRENT*BUFFERS*CONTAINING*DATA'

A11SCNAL='CURRENT*CONCURRENT*STRING*ACCESS'

A11SCNWT='CURRENT*STRING*WAITS'

A11ACTCI='CONTROL*INTERVALS*IN*USE'

New variables added to CICTSQ (TSQUQUE Global Stats)

(STID=48):

A12SHPDF='SHARED*POOLS*DEFINED'

A12SHPCN='SHARED*POOLS*CONNECTED TO'

A12SHRDS='SHARED*READ*REQUESTS'

A12SHWTS='SHARED*WRITE*REQUESTS'

New dataset CICRMG (Recovery Manager Global) (STID=99):

RMGCSHIN='CURRENT UOWS*SHUNTED FOR*INDOUBT'

RMGCSHRO='CURRENT UOWS*SHUNTED FOR*RO COMMIT'

RMGCSHTI='CURRENT DURATION*SHUNTED FOR*INDOUBT'

RMGCSHTR='CURRENT DURATION*SHUNTED FOR*RO COMMIT'

RMGIAFNW='TOTAL FORCED*INDOUBT*ACTIONS*NOWAIT'

RMGIAFOP='TOTAL FORCED*INDOUBT*ACTIONS*OPERATOR'

RMGIAFOT='TOTAL FORCED*INDOUBT*ACTIONS*OTHER'

RMGIAFTI='TOTAL FORCED*INDOUBT*ACTIONS*TIMEOUT'

RMGIAFTR='TOTAL FORCED*INDOUBT*ACTIONS*TRANDEF'

RMGIAMIS='TOTAL*INDOUBT*ACTION*MISMATCHES'

RMGNWMRO='TOTAL*FORCED*NO WAITING*IN MRO'

RMGNWOTH='TOTAL*FORCED*NO WAITING*IN OTHER'

RMGNWRMI='TOTAL*FORCED*NO WAITING*IN RMI'

RMGNWTD ='TOTAL*FORCED*NO WAITING*IN TD'

RMGNW61 ='TOTAL*FORCED*NO WAITING*IN LU61'

RMGRESYN='RESYNCH*RONISATIONS'

RMGSYBWD='SYNCPOINTS*BACKWARD'

RMGSYFWD='SYNCPOINTS*FORWARD'

RMGTSHIN='TOTAL UOWS*SHUNTED FOR*INDOUBT'

RMGTSHRO='TOTAL UOWS*SHUNTED FOR*RO COMMIT FAIL'

RMGTSHTI='TOTAL DURATION*SHUNTED FOR*INDOUBT'

RMGTSHTR='TOT DURATM*SHUNTED FOR*RO COMMIT FAIL'

New dataset CICLGR (Log Manager Journal Stat) (STID=93)

LGRBFLSH='BUFFER*FLUSH*REQUESTS'

LGRBYTES='BYTES*WRITTEN'

LGRJNLNM='JOURNAL*NAME'

LGRJTYPE='JOURNAL*TYPE'

LGRSTREM='LOG*STREAM*NAME'

LGRWRITS='JOURNAL*WRITES'

New dataset CICNQG (Enque Manager Global) (STID=96)

NQGCNQRT='CURR DUR*ENQUEUES*RETAINED'

NQGCNQSR='CURRENT*ENQUEUES*RETAINED'

NQGCNQSW='CURRENT*ENQUEUES*WAITING'

NQGCNQWT='CURR DUR*ENQUEUES*WAITING TIME'

NQGPOOL ='ENQ*POOL*ID'

NQGTIRJB='IMMEDIATE*REJECTED*ENQBUSY'

NQGTIRJR='IMMEDIATE*REJECTED*ENQ RETAINED'

NQGTNQRT='TOT DUR*ENQUEUES*RETAINED*WAITED'

NQGTNQSI='TOTAL*ENQUEUES*ISSUED'

NQGTNQSR='TOTAL*ENQUEUES*THAT WERE*RETAINED'

NQGTNQSW='TOTAL*ENQUEUES*WAITED'

NQGTNQWT='TOT DUR*ENQUEUES*WAITED'

NQGTWPOP='WAITING*ENQS*PURGED BY*OPERATOR'

NQGTWPTO='WAITING*ENQS*PURGED BY*TIMEOUT'

NQGTWRJR='WAITING*ENQS*REJECTED*RETAINED'

New dataset CICLGS (Log Manager Logstream) (STID=94)

LGSBRWRD='LOG*BROWSE*READS'

LGSBRWST='LOG*BROWSE*STARTS'

LGSBUFAP='BUFFER*APPEND*REQUESTS'

LGSBUFWT='WAITS*DUE TO*BUFFER*FULL'

LGSBYTES='BYTES*WRITTEN'

LGSCUFWT='CURRENT*FORCE*WAITERS'

LGSDELET='LOG*DELETES'

LGSPKFWT='PEAK*FORCE*WAITERS'

LGSRTRYE='RETRYABLE*ERRORS'

LGSSTREM='LOG*STREAM*NAME'

LGSTFCWT='TOTAL*FORCE*WAITS'

LGSWRITS='LOG*WRITES'
Change 14.208 DB2 datasets DB2GBPST and DB2GBPAT contained repeated obs

VMACDB2 for Global Buffer Pool zero, with no others output; the

Aug 29, 1996 offset was not incremented. After the %INCLUDE of member

EXDB2PST, insert : OFFQBGL=OFFQBGL+LENQBGL; After the

%INCLUDE of EXDB2PAT insert OFFQBGB=OFFQBGB+LENQBGB;

Thanks to Bertolotti Emmanno, Banca Commerciale Italiano, ITALY

Thanks to Burani Daniela, Banca Commerciale Italiano, ITALY
Change 14.207 Support for Boole & Babbage's PRO/SMS replacement for

EXPROSMS STOPX37 SMF record, a complete rewrite but with almost

IMACPROS all of the same variables in dataset TYPEPROS instead

TYPEPROS of TYPEX37.

VMACPROS

Aug 27, 1996

Thanks to Katrina Le, First Virginia Services, Inc, USA.
Change 14.206 Logic for IFCID=231 was wrong. Replace the statements

VMAC102 LENTRY=(QW0231NM-4)/QW0231RN; DO _I_=1 TO QW0231RN;

Aug 27, 1996 with LENTRY=52;NENTRY=(QW0231NM-4)/LENTRY;DO...TO NENTRY;

Caused INVALID DATA messages, and trashed T102S231 data.

Thanks to Ted Blank, IBM, USA.
Change 14.205 The RMF report REPORT=CHAN was always a summary report;

ANALRMFR with this change, a detail report is produced when an

Aug 27, 1996 INTERVAL= is specified, a summary if not.
Change 14.204 The deaccumulattion of DB2STAT0, DB2STAT1 and creation of

DIFFDB2 DB2STATS was move to the end of member DIFFDB2, and the

Aug 27, 1996 code rearranged so that the %INCLUDE of (EXDB2STS) is the

last physical statement in that data step. In this way,

a sophisticated user can add code in EXDB2STS to do their

own summary of DB2 data, after all of the DB2 statistics

datasets have been deaccumulated. In Pierre's case, he

wanted to extract fields from DB2STATR and DB2STATS and

merge with DB2STATS; now he can do that within EXDB2STS.

Thanks to Pierre Liu, Deere and Company, USA.

Change 14.203 Cosmetic. Duplicate variables in KEEP= list were removed

several in members TYPETMON, VMAC110, VMAC7072, VMAC79, VMAC80A

Aug 27, 1996 VMACCOMP, VMACCTLG, VMACCTLT, VMACMWUX, VMACNOAM,

VMACNSPY, VMACQAPM, and VMACTPX.

Thanks to Freddie Arie, Lone Star Gas, TEXAS.
Change 14.202 Cosmetic. Duplicate labels in VMACACHE were deleted.

VMACACHE

Aug 27, 1996

Thanks to Wolfgang Vierling, Vereinte Versicherungen, GERMANY.

Change 14.201 Change 14.152 revised IMF processing to allow for use of

IMACCIMS tape for large IMS files, but the _KIMFTRN reference in

TYPECIMS TYPECIMS must be deleted, as changes to CIMSTRAN are now

Aug 27, 1996 controlled by _KIMFTRX, not _KIMFTRN, which is no longer

defined. Comments in IMACCIMS clarify their usage.

Thanks to Wolfgang Vierling, Vereinte Versicherungen, GERMANY.

Change 14.200 The label for TYPE71 variables SHPxxxAV='MAXIMUM should

VMAC71 be SHPxxxAV='AVERAGE for these new shared page group

Aug 27, 1996 counters.

Thanks to Tim Vanderhook, Fidelity Systems, USA.

Change 14.199 READDB2 required a ddname of PDB (and worked fine as long

READDB2 as one was present), because the _LDB2STR macro was not

Aug 26, 1996 redefined in READDB2. Find the two _LDB2STB macro defs,

replicate each, and change STB to STR, STATB to STATR.

Thanks to Chuck Hopf, MBNA, USA.
Change 14.198 Variable AVGMEMSZ was not kept in TYPE72GO dataset. This

VMAC7072 metric, MSOUNITS based, is a poor indicator of memory,

Aug 26, 1996 but it was not my intention to drop it. It is kept now.

Thanks to Clark Jennings, Reynolds Metals Company, USA.

Change 14.197 Support for Landmark's TMON/DB2 Version 3 (INCOMPATIBLE).

EXTMDDF The Statistics Interval data TMDBDA1 has 33 new variables

EXTMDDF inserted, and 50 new variables were inserted into TMDBDAB

IMACTMDB dataset. The Accounting data TMDBDB2 has 36 variables

VMACTMDB inserted, 9 new variables in TMDBDBB, 1 new in TMDBDBF,

Aug 26, 1996 6 new in TMDBDBK, and a new dataset, TMDBDF, is created

for the 'DF' record type. Only the 'DA' and 'DB' records

have been validated with Version 3 data. Test data is on

hand for the 'DC','DD','DF','BB','BH', and 'BI' records,

but DSECTs were provided only for DA and DB records. If

there is a need, additional record types will be decoded.
Change 14.196 MXG 14.05-08 only. INVALID DATA FOR R744FCTM and FCSQ

VMAC74 because those two fields, added by Change 14.165, are not

Aug 23, 1996 always there. After INPUT of R744FSQU, insert

@; IF SMF744FL GE 100 THEN INPUT

so they are only INPUT when present in the record!

See Change 14.273.

Change 14.195 Dataset DB2STATR outputs the first QWHSLOCN segment

VMACDB2 NRQLST times, because the OFFQLST pointer was not

Aug 23, 1996 incremented. Insert: OFFQLST=OFFQLST+LENQLST;

immediately before the line reading:

END; /* END TRACE SECTION FOR TYPE 100 SUBTYPE 0 */

But then DIFFDB2 deleted all but the first obs, so the

number of observations was way off too! This error also

caused remote counts to be wrong in ANALDB2R.

Thanks to Pierre Li, Deere & Company, USA.

Thanks to Carl Stotmeister, Deere & Company, USA.

Thanks to Ralph Baechle, Deere & Company, USA.
Change 14.194 MXG 14.05-14.06. DB2STATB dataset will have extra obs

DIFFDB2 that should not have been OUTPUT. Change 14.167 added

Aug 23, 1996 protection for IBMs skipping QWHSISEQ, but four lines

IF SEQCHECK NE 1 ... must be IF SEQCHECK GE 1 ...

to protect for change in Buffer Pool ID, which creates a

negative value in variable SEQCHECK and should not have

been output. A more robust algorithm, testing for change

in QBSTPID was added, and dataset DB2STATR had the same

exposure and was revised to protect for change in

QLSTLOCN in the deaccumulation logic.

Thanks to Tom Parker, Hogan Systems, USA.
Change 14.193 The second occurrences of IOERTMPC and IOERPRMC should be

VMACCTLT spelled IOEWTMPC and IOEWPRMC, in both the Label

Aug 20, 1996 statement and in the INPUT statement.

Thanks to Freddie Arie, Lone Star Gas, TEXAS.

Change 14.192 MXG 14.06 test stream only. In JCLTEST6 and JCLUXRE6,

JCLTEST6 step TESTOTHR, DD DUMMY for //TANDCTRL and //TANDLINE

JCLUXRE6 were missing, and in JCLUXRE6, //XPMODS was misspelled

Aug 20, 1996 as //XPUMODS.

Thanks to Freddie Arie, Lone Star Gas, TEXAS.
====Changes thru 14.191 were included in MXG 14.06 dated Aug 20, 1996===
Change 14.191 MXG 14.05 only. ERROR 228-185 INFORMAT $PIB IS UNKNOWN in

VMAC74 BUILDPDB, if you have tailored BUILDPDB to process the

Aug 19, 1996 CRR Cache RMF Reporter SMF records (VMACACHE). Change

14.152 to VMAC74 incorrectly INPUT RTRNCODE $CHAR2.

but that line must be changed to RTRNCODE &PIB.2.

to be consistent with VMACACHE.

Thanks to Damian Stevenson, PMSA, AUSTRALIA.

Thanks to Johnny Allen, Yellow Freight, USA.

Change 14.190 Minor. The count of reports requested was too high by

ANALCISH one.

Aug 17, 1996
Change 14.189 Enhancements to ANALRMFR for IBM-like RMF reports.

ANALRMFR -Shared Device Activity Report is complete ( SDEVC ).

Aug 17, 1996 This report will not contain any Devices that were

OFFLINE or have zero values at time of the interval, as

they are not kept in the MXG PDB.TYPE74 data set.

-Channel Activity Report, updates for MVS/ESA 5.1, the

addition of SMF73CPD adds new channel path descriptions.

Workload Activity Report-Goal Mode, REPORT=WLMGL is

-installed along with a new parameter RPTOPT=WGPER, which

will allow for report selection level. See comments

within ANALRMFR for more details. This report will not

contain any Policies that have all values of zero at end

of interval, as they are not kept in the MXG PDB.TYPE72GO

data set.

Change 14.188 CICINTRV is now restored to the pre-MXG 14.04 logic. In

CICINTRV MXG 14.04 the new "CICINTRZ" logic replaced CICINTRV, but

Aug 16, 1996 CICINTRZ was prematurely released (it creates incorrect

data in some situations, and it stopped creating the

PDB.CICEODRV dataset which is needed for CICS Shutdown

reports). However, the old CICINTRV logic was also wrong

for CICS 4.1 data (because multiple CICLDG observations

were being kept, which caused multiple observations in

datasets it created), so this CICINTRV member was fixed

to keep only the first CICLDG observation, which solved

it error. CICINTRZ ultimately will replace CICINTRV, as

it will create true interval data from the multiplicity

of statistics events (INT,REQ,USS,RVR,EOD) in CICINTRV

and the End of Day in CICEODRV, but until it is fully

validated, it is safer to return to the simpler logic.

So this change negates change 14.132. Additionally

Change 13.154 was removed from this CICINTRV. It added

an INCLUDE of IMACCICS inside CICINTRV (so you did not

have to include IMACCICS before you invoked CICINTRV),

but that change prevented ANALCISH from re-defining the