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

TAABPMIC='APPLICATION*BASE*PROGRAM*MICRO*VER NR'

TAITKANM='INITIAL TASK*APPL CNTX*APPLICATION*NAME'

TAITKPNM='INITIAL TASK*APPL CNTX*PLATFORM*NAME'

TAITKONM='INITIAL TASK*APPL CNTX*OPERATION*NAME'

TAITKMAJ='INITIAL TASK*APPL CNTX*MAJOR VER NR'

TAITKMIN='INITIAL TASK*APPL CNTX*MINOR VER NR'

TAITKMIC='INITIAL TASK*APPL CNTX*MICRO VER NR'

TACTKANM='CURRENT TASK*APPL CNTX*APPLICATION*NAME'

TACTKPNM='CURRENT TASK*APPL CNTX*PLATFORM*NAME'

TACTKONM='CURRENT TASK*APPL CNTX*OPERATION*NAME'

TACTKMAJ='CURRENT TASK*APPL CNTX*MAJOR VER NR'

TACTKMIN='CURRENT TASK*APPL CNTX*MINOR VER NR'

TACTKMIC='CURRENT TASK*APPL CNTX*MICRO VER NR'

-New dataset MONIPSB from 'TA' PSB SEGMENTs.

TADDBCCT='TOTAL*DL/I*DATA BASE*CALLS'

TADDLETC='DATA BASE*DLET CALLS*ISSUED'

TADEBCT ='DEDB*CALLS'

TADEDBRC='DEDB*READ*OPERATIONS'

TADEDQCT='EXCLUSIVE*DEQUEUES'

TADENQCT='EXCLUSIVE*ENQUEUES'

TADGHNCT='DATA BASE*GHN CALLS*ISSUED'

TADGHNPC='DATA BASE*GHNP CALLS*ISSUED'

TADGHUCT='DATA BASE*GHU CALLS*ISSUED'

TADGNPCT='DATA BASE*GNP CALLS*ISSUED'

TADISRTC='DATA BASE*ISRT CALLS*ISSUED'

TADLDBIO='DATABASE*I/O'

TADLGNCT='DATA BASE*GN CALLS*ISSUED'

TADLGUCT='DATA BASE*GU CALLS*ISSUED'

TADLIPSB='PSB*NAME'

TADLUSSN='USSN*NUMBER'

TADOBUCT='OVERFLOW*BUFFERS*USED'

TADREPLC='DATA BASE*REPL CALLS*ISSUED'

TADTDQCT='TEST*DEQUEUES'

TADTNQCT='TEST*ENQUEUES'

TADUDQCT='UPDATE*DEQUEUES'

TADUNQCT='UPDATE*ENQUEUES'

TADUOWCC='UOW*CONTENTIONS'

TADWEDBC='WAITS FOR*DEDB*BUFFER'

TADWENQC='WAITS ON*EXCLUSIVE*ENQUEU'

TADWTNQC='WAITS ON*TEST*ENQUEUES'

TADWUNQC='WAITS ON*UPDATE*AND ENQUE'

TAPSBICT='TAPSB*SEGMENT*COUNT'

WKDLDBTM='ELAPSED*TIME FOR*DATABASE I/O'

WKDLICTM='THREAD*TCB*CPUTIME'

WKDLINWT='ELAPSED*WAIT TIME*INTENT*CONFLICT'

WKDLLKTM='ELAPSED*TIME FOR*PI LOCKING'

WKDLPLWT='ELAPSED*WAIT TIME*POOL*SPACE'

WKDSCETM='SCHEDULE*COMPLETED'

WKDSCSTM='SCHEDULE*STARTED'

WKSCHDTM='ELAPSED*TIME FOR*SCHEDULE PROCESS'

-New variables in MONITR 'TR' record dataset.

TRASACTV='CURRENT*ADDRESS*SPACE*ADDRESS'

TRASHWM ='ADDRESS*SPACE*HWM'

TRASPMHW='HWM*AUX*SLOTS*PMO'

TRASPMO ='AUX SLOTS*TO BACK*64 BIT PMO'

TRBAPMO ='BYTES*ALLOCATED TO*PRIVATE MEMORY'

TRBHPMO ='BYTES*HIDDEN IN*PRIVATE MEMORY'

TRBPMOHW='HWM BYTES*USABLE IN*PRIVATE MEM'

TRCDSAL ='CURRENT*DSA*LIMIT'

TRCDSAT ='CURRENT*DSA*TOTAL'

TRCEDSAL='CURRENT*EDSA*LIMIT'

TRCEDSAT='CURRENT*EDSA*TOTAL'

TRCMCSU ='CUMULATIVE*COMMON*SUBSPACE*USER'

TRCMSUHW='HWM*COMMON*SUBSPACE*USERS'

TRCMUSU ='CUMULATIVE*UNIQUE*SUBSPACE*USER'

TRCRCSU ='CURRENT*COMMON*SUBSPACE*USERS'

TRCRUSU ='CURRENT*UNIQUE*SUBSPACE*USERS'

TRDSAHWM='HWM*DSA*TOTAL'

TREDSAHW='HWM*EDSA*TOTAL'

TRFGFAIL='NO FROM*GUARD*FAILURES'

TRFGFSZ ='FROMGUARD*FAILURE*SIZE'

TRGDSAAC='CURRENT*GDSA*ACTIVE'

TRGDSAAL='CURRENT*GDSA*ALLOCATED'

TRGDSAHA='HWM*GDSA*ACTIVE'

TRGDSAHW='HWM*GDSA*ALLOCATED'

TRGETSSZ='GETSTOR*REQUEST*SIZE'

TRMEMLMT='MEMLIMIT*SIZE'

TRMLIMTS='MEMLIMIT*SOURCE'

TRNOLMO ='LARGE*MEMORY*OBJECTS'

TRNOSHMO='SHARED*MEMORY*OBJECTS'

TRNUMPMO='PRIVATE*MEMORY*OBJECTS'

TRPGPOOL='PAGEPOOLS'

TRRFPMHW='HWM*REAL*FRAMES*PMO'

TRRFPMO ='REAL*FRAMES*64-B*PMO'

TRRNTPGM='STATE*OF*RENTPGM'

TRSBFLMO='SHARED*BYTES*FROM*LARGE MEMORY'

TRSBLMHW='HWM*SHARED BYTES*IN LARGE MEMOR'

TRSTGPRO='STATE OF*STORAGE*PROTECT'

TRTRNISO='STATE OF*TRANISO'

TRUSUHWM='HWM*UNIQUE*SUBSPACE*USERS'

Jan 6: DEBUG 8 _N_= messages eliminated.

-Divide by 4096 for all durations was added.

Change 32.308 These new-in-XCOM-11.6 variables are now INPUT and some

FORMATS are decoded by new $MGXCMxx formats:

VMACXCOM XCORELEASE XCONXFER_AK XCOPDSMN_AK1 XCODATE_AK1

Dec 30, 2014 XCOTIME_AK1 XCONAME_AK XCOPDSMN_AK2 XCODATE_AK2

XCOTIME_AK2 XCOLUSER_AK XCOPDSMN_AK3 XCODATE_AK3

XCOTIME_AK3 XCOID_AK XCOPDSMN_AK4 XCODATE_AK4 XCOTIME_A

XCOTNAME_AK XCOPDSMN_AK5 XCODATE_AK5 XCOTIME_AK5 XCOPDS

XCOTOTCPU XCOTCBCPU XCOSRBCPU XCOZIIPZCPU XCOZIIPCCPU

XCOZIIPELIG XCONRECS2 XCONSEND2 XCONRECV2 XCONPUT2

XCONGET2 XCOALPRI2 XCOALSEC2 XCOALDIR2 XCOALRUNIT

XCOEATTR XCOLCIPH_LIST XCOCIPHER XCOPROTOCOL XCOLCHAR

XCORCHAR XCOLCCSID XCORCCSID XCOSCCSID XCOTCCSID XCOLTN

XCORTNQ XCOMIERR XCOMCERR XCOLDELIM_ENCODE

XCORDELIM_ENCODE XCOMIREPL XCOMCREPL XCOLDELIMITERS

XCORDELIMITERS XCOMIREPL_CNT XCOMCREPL_CNT XCOXMITF
Change 32.307 TYPE50 with VERSN50=2 ATTCHTYP=4 and LENGTH=254 caused an

VMAC50 INPUT STATEMENT EXCEEDED error because MXG expected 262

Dec 29, 2014 bytes. Now the 8-byte TY50RDQN Read-Queue-Name is INPUT

only when there are 8-bytes left. I presume that field

was added by an APAR.

Thanks to Steven Womer, OCLC, USA.

Change 32.306 If BLDSMPDB was used with RUNDAY=NO, the PDBAUDIT report

BLDSMPDB was incorrectly invoked, causing DDNAME NOT FOUND error.

Dec 29, 2014 Now, PDBAUDIT is not run when RUNDAY=NO is specified.

Thanks to Jim Horne, Lowe's, USA.

Change 32.305 DOCUMENTATION.

FORMATS -Format $MGPROD maps every member in MXG to a product.

Dec 27, 2014 This is work in progress for Spring, 2015.
====== Changes thru 32.304 were in MXG 32.12 dated Dec 23, 2014=========
Change 32.304 Support for RMF APAR OA45421 adds new function to the SMF

VMAC74 74 subtype 4 record, new variables in TYPE74ST dataset:

Dec 19, 2014 R744SISC='INDEX TO*SCM*DATA*SECTION'

R744SNSC='STORAGE*CLASS*MEMORY*DATA*SECTIONS'

R744SSAC='SCM AR*CONDITION*REQUIRED*RESTART'

R744SOSA='SCM AR*CONDITION*SUCCESSFUL*OP'

Change 32.303 Support for APAR OA44798, which adds two variables to the

VMAC22 TYPE22 Subtype 10 record, in dataset TYPE22_A:

Dec 19, 2014 SMF22SMT='MULTI-TARGET*PPRC*STATUS'

SMF22PMT='PREVIOUIS*MULTI-TARGET*PPRC*STATUS'

Change 32.302 MXG 32.06-32.11. The QAWPS program %INCLUDE of BUIL3005

QAWPS text was incorrectly changed to BUIL3206 in 32.06 and

Dec 19, 2014 then was BUILVVNN for each version instead of BUIL3005.

Thanks to Declan Vibert, World Programming, ENGLAND.

Change 32.301 MXG 32.11, WPS ONLY. A letter F left in macro _VARWECR

VMACWECR caused ERROR: DATA SET "WORK.WEBSCRAU" NOT FOUND".

Dec 22, 2014 SAS did not fail; it added the F to the dataset LABEL.

Thanks to Declan Vibert, World Programming, ENGLAND.

Change 32.300 MXG 32.11, WPS on ASCII ONLY, INVALID SMF RECFM. A test

VMACSMF left from Change 32.258 ("OR %SYSPROD(WPS EQ 1") caused

Dec 19, 2014 ERROR: UNRECOGNIZED RECORD FORMAT VBS on ASCII (because

WPS requires RECFM=S370VBS on ASCII). Test is removed.

Thanks to Declan Vibert, World Programming, ENGLAND.
Change 32.299 The default invocation of PDBAUDIT (contents of today's

BLDSMPDB "PDB's", comparison with yesterdays) is at the end of the

Dec 18, 2014 default BUILDPDB (PDB.SPUNJOBS), but that is prior to any

ASUMxxxx or other members that you added, so those other

PDB datasets would not be reported.

-This change adds the PRINTAUDIT parameter to BLDSMPDB

PRINTAUDIT=&MXGPRINTAUDIT

to automatically defer the %PDBAUDIT invocation until

after all of the INCLAFTR programs have executed.

-If you use BUILDPDB and your own includes, you can use

%LET MXGPRINTAUDIT=NO; /* temp replace default */

%INCLUDE SOURCLIB(BUILDPDB);

%INCLUDE - all of your stuff - ;

%PDBAUDIT(PRINTAUDIT=YES);

Change 32.298 Multiple UTILBLDP executions in a single job could get

UTILBLDP errors with missing parens and other nastiness if any

Dec 18, 2014 of the EXPDB*** parameters were used. EPDBINC EPDBCDE

EPDBVAR EPDBOUT are now all nulled at the end.

Change 32.297 New argument LIBNAMES to select which LIBNAMEs are used,

PDBAUDIT and new options added to PRINTAUDIT= argument:

Dec 18, 2014 LIBNAME=_ALL_ Default, search all open LIBNAMES:

A LIBNAME is open if:

-zOS it has been touched by a DATA or PROC step

or there was a LIBNAME statement used

-ASCII there MUST have been a LIBNAME statement

PRINTAUDIT=

YES - default - datasets and reports generated

NO - PDBBAUDIT becomes a null statement

DATAONLY - only builds datasets and does not print

any reports

PRINTONLYCHANGE - builds datasets and prints only

the report of differences

PRINTONLYCONTENTS - prints only the CONTENTS report
Change 32.296 New macro variable &MACJESN is added in VGETJESN so you

VGETJESN can delete records by their TYPETASK values. For example,

VMXGINIT the large number of SMF 30s written for OMVS tasks can be

Dec 18, 2014 can be deleted from your BUILDPDB datasets, using

//SYSIN DD *

%LET MACJESN=

%QUOTE( IF TYPETASK EQ 'OMVS' THEN DELETE; ) ;

%INCLUDE SOURCLIB(BUILDPDB);

Note that using MACJESN "instream" in your //SYSIN only

impacts this job, so you could separately run TYPE30

program and see all those OMVS task records.

Thanks to Richard Stuchell, Visa, USA.

Change 32.295 Variables XMLSYSTEM/TEMSSEQ in T112TEMS dataset created

UTILEXCL by SUBSTR(XMLRECORD) with INPUT XMLRECORD $VARYING32000,

VMAC112 so SAS defaults their length to 32000, when there is no

Dec 17, 2014 LENGTH statement. That is a problem ONLY because both

are in the BY list for PROC SORT, which then caused

ERROR: HOST SORT CAN NOT BE USED (SORTPGM=HOST/SORT)

WARNING: HOST SORT CAN NOT BE USED (SORTPGM=BEST)

because DFSORT/SYNCSORT don't allow a BY list over 32760

bytes long. Both are now shortened in a LENGTH statement.

-An ERROR occurs with the site option SORTPGM=HOST/SORT.

Instead, if SORTPGM=BEST is used, SAS issues the WARNING

and proceeds to use its internal sort. Knowing this now,

I recommend SORTPGM=BEST in your site's CONFIG.

-The UTILEXCL program has a long BY list and has noted in

comments that you must use OPTIONS SORTPGM=SAS.

-But how did this slip thru my QA? Well, it turns out that

SAS does NOT validate the BY list length if the dataset

has zero observations; I don't always have obs for every

MXG dataset. I think that lack of validation is a defect,

but now that I'm aware SAS may not find these errors for

me, I've revised the UTILVREF QA program to now calculate

the length of the BY list for every dataset and report

any new exposures so they can be corrected/documented.

-Note: There are other HOST SORT CAN NOT BE USED causes.

Thanks to Gaetan Martel, Intact Corportation Financiere, CANADA.
Change 32.294 This analysis example to compare two WEEK's TYPE72GO data

ANALCPU for each Service Class, matching intervals from midnight

Dec 17, 2014 had &PDBMXG..TYPE72GO instead of WEEK.TYPE72GO and so it

failed with LIBREF PDB IS NOT ASSIGNED.

Thanks to Jerry Schmidt, Northeast Utilities, USA.
Change 32.293 Warning added to ASUMDB2A when "Rollups" are detected.

ASUMDB2A Created when ACCUMACC is specified, rollups summarize

Dec 17, 2014 DB2 events, leaving no "detail" event data in DB2ACCT, so

you need to be aware Rollups impact ANALDB2R reports and

makes any summarization of the already summarized data of

questionable utility. The new log messages print:

MXGWARN: DB2 ROLLUPS DETECTED. SUMMARIZED VALUES CANNOT

MXGWARN: BE USED FOR DETAIL ANALYSIS. TOTAL AND MAX

MXGWARN: VALUES WILL BE CORRECT BUT AVERAGE VALUES WILL

MXGWARN: BE INCORRECT. USE WITH CAUTION.

Note that %ANALID reports tabulate which DB2 Subsystems

have enabled ACCUMACC; %ANALID reports are automatically

created by BUILDPDB to tabulate your input SMF data, or

it can be directly executed to read/report on your SMF.

Change 32.292 -ASCII only. The MXG default SORTSIZE=400M is removed so

AUTOEXEC the default SORTSIZE is chosen. No problem was reported,

AUTOEXEU but this archaic value could negatively impact sorts.

AUTOEXEW -Optional ODS and DM commands are now enclosed in comment

Dec 16, 2014 blocks, rather than enabling by default.
Change 32.291 Support for GDPS SMF 105 Record APAR PI26702 (replaced

VMAC105 PI16853) INPUTS eight new two-byte fields into existing

Dec 16, 2014 variables SM105LTV/LOV/LPV/LSV/LUV/LCV/LJV/SN195LFV that

were previously only one-byte fields. When the APAR is

installed, it's flag bit detects its presence and inputs

the new fields transparently. One byte was too small if

a client had an LSS with a full 256 devices defined.

Thanks to Dave Clitherow, IBM GDPS Development, UK.

Change 32.290 "Archaic" IMS log processing programs, TYPEIMS7 to create

TYPEIMS7 IMS07 IMS07D IMA0A7 IMS0708 IMSUMRY datasets or JCLIMSL6

VMACIMS and ASMIMSL6 to create IMSTRAN.IMSTRAN, both now use the

VMACIMSA newer and larger (by about 5%) DLREXTIM for the IMSCPUTM

Dec 12, 2014 value instead of the original CP CPU time field, DLRTIME.

DLREXTIM was added in IMS 10.1 and recommended by IBM IMS

support. DLRTIME is also now kept in those datasets.

NOTE: THE RECOMMENDED, NON-ARCHAIC MXG IMS LOG PROCESSING

in JCLIMSTT that creates the IMS56FA IMS Transaction Data

Set was not changed: IMSCPUTM correctly used TPEXTIME.

-IMSCPUTM/DLRTIME/DLREXTIM contain ONLY the CP CPU time

and DLRAZAAP/TPEZAAP contain ONLY zIIP/zAAP CPU time.

Thanks to David Christianson, State of Wisconsin, USA.
Change 32.289 -RMF III processing performance enhancements, message

ADOCRMFV improvements, fixes, and documentation upgrades.

ASMRMFV -Improved ASMRMFV handling of all Return Codes, Reason

CLRMFV Codes, and Info Codes into messages eliminating some

JCLRMFV instructions.

JCLCRMFV -Severe error messages RMFV003S and RMFV007S will now

JCLDRMFV provide the failing subroutine name for better diagnosis.

Dec 20, 2014 In prior ASMRMFV versions only the general name of the

failing service (OPEN, CLOSE, etc.) was shown.

-Messages RMFV024I, RMFV025I, RMFV028I, RMFV029*,

RMFV031I, RMFV037I, and RMFV999I are modified for better

alignments and legibility.

-Message RMFV028I for Indexes is now a multi-line message.

-Messages RMFV012I with Sample Begin/End Date/Time stamps

always be issued even if NODETAIL is in effect.

-When the NODUPDSN option is in effect message RMFV101I

will include the number of DSNAME COMPARES in both Detail

and Summary Reports as an indicator of the overhead

incurred detecting duplicate data set names.

-Message RMFV105I will now show the table name for the RMF

III Data Set Header as DSI instead of DSH. This was the

only report table id not conforming to the actual RMF III

internal id which is ERBDSIG3.

-Using the RMF III internal table identifications

consistently allows for more efficient table validation.

ASMRMFV documentation and source code will still refer to

this table as the DSH or Data Set Header.

-When the POLICY option was specified with the SIZE option

no Service Policy information was produced. The SIZE

option provides a quick inventory of the space usage,

index usage, and attributes of all allocated RMF III data

sets with no RMFBSAM output.

-When the SIZE option is used now any RMF III data set

filters such as NODUPDSN, SYSPLEX=, SYSTEM=, Date/Time,

and DOW= (and their aliases) will be honored to provide

Index and Space usage for only selected data sets. In

prior ASMRMFV versions all data set filters were ignored

when SIZE was specified.

-The SIZE option will now bypass opens and closes for

unneeded non-VSAM data sets RMFBSAM, RMFFILT, and RMFSKIP

for better performance.

-Summary messages RMFV100I, RMFV012I, and RMFV014I will

now also be issued when SIZE is in effect.

-SZ is no longer an alias for the SIZE option, it is now

an alias for SHOWZERO. Please use SI as a SIZE alias

instead.

-New parameters SHOWZERO (alias SZ) and NOSHOWZERO (alias

NOSZ) control display of RMF III table statistics in

message RMFV105I when there are zero occurrences of a

particular table. The default is SHOWZERO providing the

same behavior as prior ASMRMFV versions.

-These parameters replace the ZEROPRT/NOZEROPRT options

and respective aliases since the purpose was unclear from

the names. However, the old parameters are still

accepted without error.

-SHOWZERO/NOSHOWZERO parameter documentation is added to

Section 6 "Report Control Parameters" in ASMRMFV and

ADOCRMFV members. The SIZE option is also updated.

-Updates for revised messages are made to Section 12

"Messages" in ASMRMFV and ADOCRMFV members.

-Section 25 "Summary" is updated to add new parameters.

-REQUIREMENT: In order to implement these features the

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.
Change 32.288 SMF Type 22 log message UNKNOWN SECID=40 printed because

VMAC22 the CPU segment is only 6 bytes, MXG read 7. Obviously,

Dec 6, 2014 this is NOT a frequently used SMF record! Error was

introduced by Change 32.064 in MXG 32.03.

Thanks to Gerard Bosker, Rabobank Nederland, THE NETHERLANDS.
Change 32.287 VMXGGETM was checking the length of NRECORDS when it

VMXGGETM should have been checking for a value of MAX, when the

Dec 6, 2014 NRECORD=MAX option was chose, causing log message

NOTE: Variable MAX is uninitialized.

Change 32.286 A second execution of %VMXGSRCH might do nothing; both

VMXGSRCH generated WARNING: MULTIPLE LENGTHS FOR BY VAR MEMNAME

Dec 3, 2014 but only the first execution executed your search.

Thanks to Rodger Foreman, Trans Union, USA

Change 32.285 A RNAME (Minor Queue Name) can contain hex characters

VMACRMFV so new variable ENTMINNAHEX with $HEX72 format provides

Dec 3, 2014 the hex values when ENTMINNA contains non-printables.

These additional variables will also be updated and this

change text will be revised when completed.

Dataset Variable

CMFVEN ENRERNM

CMFRV RVRERNM

TYPEMIM MIMCMRNM

TYPEPDL RTYPEU

TMVSNQ NQMINOR

TYPE87 SMF87QSCAN_RNAME

TYPE796 TYPE796MIN

TYPE797 TYPE797MIN

Thanks to Scott Barry, SBBWorks Inc., USA.
Change 32.284 -INVALID THIRD ARGUMENT TO FUNCTION SUBSTR in NDM 'PT' SMF

VMACNDM record when LOCNULL=1 is corrected by removal of SUBSTR.

Dec 3, 2014 -Additional NDMRTYPE values of A# C# D# S$ U# UK are now

also output in dataset NDMAE.

-Protection for ancient short records added so that now

every single NDM record I've ever received is tested.

Thanks to Gerard Bosker, Rabobank Nederland, THE NETHERLANDS.

Thanks to Michael Oujesky, DTCC, USA.

====== Changes thru 32.283 were in MXG 32.11 dated Dec 2, 2014=========
Change 32.283 Debugging PUTLOG removed, printed multi-million lines,

VMACRMFV causing a refresh and re-date of 32.11.

Dec 2, 2014

Thanks to Scott Barry, SBBWorks Inc., USA.

Change 32.282 ASCII only. If the dataset for which you are searching

VGETOBS was a VIEW, it was found, but the DSNAME returned was a

Dec 2, 2014 period and the member type was blank. The period should

have been in the VGETOBS variable and the member type

should have been set to VIEW. The VGETDSN variable was

also incorrectly set to period on z/OS, all fixed now.

Thanks to Karl Olafsson, Advania, ICELAND.

Thanks to Hreinn C. Hreinsson, Advania, ICELAND.

====== Changes thru 32.281 were in MXG 32.11 dated Dec 1, 2014=========
Change 32.281 MXG 32.02-32.10. BUILDPDB fails if //PDB is on tape after

VMAC74 Change 32.031 added code to VMAC74 that caused SAS to try

Nov 28, 2014 to open PDB.TYPE748A and PDB.TYPE748R simultaneously, and

that can't be done when //PDB is a sequential library.

Code revised to create a temporary copy first. BUT:

WE RECOMMEND YOU NEVER USE //PDB ON TAPE for the BUILDPDB

job, not only because you can't have two datasets open,

but because BUILDPDB first writes many datasets to the

PDB, but then has to read them back in (e.g.,RMFINTRV has

to read ALL of the TYPE7xxx RMF datasets to create that

summary PDB.RMFINTRV dataset), and since there is no

dictionary on sequential libraries, SAS has to read every

record to find the first dataset, then rewind back to the

start, read to find the second dataset, etc., causing

massive increase in elapsed time. And consider what

happens when it's a five volume tape dataset that has to

have each volume mounted, read, dismounted, etc., etc.

Instead, make your //PDB DD a temporary DASD file for

the building of the PDB data library, but then add

a //REALPDB DD UNIT=TAPE,DSN=YOUR.REAL.PDB,. . . and a

PROC COPY IN=PDB OUT=REALPDB MEMTYPE=DATA;

to archive the PDB to tape efficiently.

-TYPS74 fails the same way/same reason if PDB is on TAPE.

Thanks to Jerry Schmidt, Northeast Utilities, USA.

Change 32.280 UNUSED Change Number.

Nov 22, 2014

Change 32.279 Mobile Workload support was enhanced by adding variables

MOBWRK02 SMF89IST SMF89EST SMF89ST SMF89UST in MOBWRK02, needed to

MOBWRK05 determine which of the SMF89 starting timestamps should

Nov 26, 2014 be used to set the STARTHR, and MOBWRK05 was updated to

use the STARTHR in all the merges, which is needed for

hours when an LPAR was moved to a different CEC.

Thanks to Graham K. Harris, RBS, ENGLAND.
Change 32.278 TYP11902 dataset variables TTTTLSSP/NC/ST/FP/UI wrong

VMAC119 when IBM inserted 16 bytes after TTDUAKRC and before the

Nov 26, 2014 relocatable OFF11903 segment, but MXG overlooked using

OFF11903 in the INPUT (probably because prior iterations

of this SMF 119 type record happened to have OFF11903 at

the COL after TTDUAKRC.) Both the OFF11903 and OFF11904

segments are now corrected and protected for inserts.

-New variable added to TYP11902 dataset, discovered in the

z/OS 2.1 IP Programmers guide, 2014 edition, which does

not show those 16 bytes after TTDUAKRC.

TTTTLSNC4='ATTLS*NEGOTIAGED*4-BYTE*CIPHER'

Thanks to Michael Creech, Black Knight Financial Services, USA.

Change 32.277 Crypto variable execution times R7023MET and R7023CRT are

VMAC7072 now correctly multiplied by the R7024SF Scaling Factor;

Nov 25, 2014 that needed multiply was overlooked when they were added.

Thanks to Michael Creech, Black Knight Financial Services, USA.

Change 32.276 Variables XCODSN2='REMOTE*DSN*WITH*G000V00' is now KEPT

VMACXCOM in TYPEXCOM; the adjacent XCODSN1 was also overlooked and

Nov 21, 2014 is now kept, but that Local DSNAME value was also in the