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

T102S255 BUFFER REFRESH CROSS-INVALIDATED'

T102S256 ALTER GROUPBUFFERPOOL COMMAND'

T102S257 INTER-DB2 NOTIFY MESSAGE SENDING'

T102S258 DB2 DATASET EXTEND SPACE'

T102S259 P-LOCK STATE CONSTANTS'

T102S260 END MVS XES REQUEST'

T102S261 GROUP BUFFER POOL'

T102S262 GBPOOLT CASTOUT THRESHOLD'

T102S263 PAGESET AND PARTITION CASTOUT'

T102S265 SCA ACCESS REQUEST BEGIN'

T102S266 SCA ACCESS REQUEST END'

T102S267 START CF STRUCT REBUILD/EXPAND'

T102S268 END CF STRUCTURE REBUILD/EXPAND'

T102S272 ASSOCIATE LOCATORS STMT EXEC'

T102S273 ALLOCATE CURSOR STATEMENT EXEC'

T102S305 TABLE CHECK CONSTRAINT'

T102S306 LOG RECORDS'

T102S311 GLOBAL TEMPORARY TABLE USAGE'

T102S312 DCE SECURITY AUDIT TRAIL'

T102S313 CHECKPOINT LONG RUNNING UR-S'

T102S314 ACCESS CONTROL AUTH EXIT PARMS'

Change 16.317 Protection for NTSMF 2.2.1, which does not write a 0.0

VMACNTSM record for each interval. Change the test

Dec 28, 1998 IF VERSION LT : '2.1' THEN to IF VERSION LT : '2.1' or

VERSION EQ '2.2:1' THEN ....

If you have specified COMPRESS, then the expected header

version value of 2.2.2 is created and MXG has no error.

Thanks to Wayne Holtz, Reynolds Metal, USA.
Change 16.316 First MXG 16.08 only. CICINTRV failed in building CICDS

VMXGCICI because the member VMXGCICI from CICS TS 1.3 testing was

Dec 23, 1998 inadvertently shipped. The VMXGCICI from MXG 16.07 has

been put back in this MXG 16.08.

Change 16.315 Year 2000 dates cause this summarization example to fail.

ASUMHSM The DATEJUL() function does not support the Century Byte

Dec 23, 1998 (which is an MVS-only Packed-Decimal-implementation) and

this example used DATEJUL() in an INCODE= on a DSRDATE

value of 0100001 (01Jan2000 in 0cyyddd format). The MXG

use of DATEJUL() was replaced by the YEAR2000 member's

algorithm to protect this example. See Change 16.330.

Thanks to John McCray, Huntington Services Company, USA.

======Changes thru 16.315 were in MXG 16.08 dated Dec 23, 1998======
Change 16.314 New function enhancement to VMXGSUM. Named XMXGSUM for

XMXGSUM testing; it is not yet the default. Copy it into your

Dec 21, 1998 USERID.SOURCLIB, rename to VMXGSUM, and check it out.

Feb 16, 1999 This rewrite exploits some new features added by SAS in

Version 7 (if it finds it is running under V7), and adds

new parameters for additional descriptive statistics

under either Version 6 or Version 7 of SAS.

New Parameters and defaults:

(Percentile values only in version 7+.)

P1=, List of variables to calculate Pctile 1

P5=, List of variables to calculate Pctile 5

P10=, List of variables to calculate Pctile 10

P25=, List of variables to calculate Pctile 25

P50=, List of variables to calculate Pctile 50

P75=, List of variables to calculate Pctile 75

P90=, List of variables to calculate Pctile 90

P95=, List of variables to calculate Pctile 95

P99=, List of variables to calculate Pctile 99

STD=, List of variables for Standard Deviation

STDERR=, List of variables for Standard Error

CV=, List of variables for coeff of variation

T=, List of variables for T value

VAR=, List of variables for variance

KURTOSUS= List of variables for kurtosis

AUTONAME=NO, Use the AUTONAME feature

In prior releases of VMXGSUM, you always had to ensure

that any variable you specified in one of the operands

actually existed or the output variable would not be in

the output dataset. Thus, if you wanted the MAX and the

SUM of the variable CPUTM, you would have to specify:

%VMXGSUM(...

SUM=CPUTM,

MAX=MAXCPU,

INCODE=MAXCPU=CPUTM;,...

The INCODE created the variable MAXCPU as being the same

as CPUTM and the MAX=MAXCPU calculated the maximum value

found. With version 7, there is a new option in PROC

MEANS called 'AUTONAME.' By using this feature, you can

create 'long' variable names and avoid the need to create

the variables via INCODE processing that you want. The

variable name is constructed by adding an underscore

followed by the statistic name such as MEAN or MAX or

SUM. Using the example above, the specification would

be:
%VMXGSUM(...

SUM=CPUTM,

MAX=CPUTM,

AUTONAME=YES,...
In the output dataset you would then have the variables

CPUTM_SUM and CPUTM_MAX. The format, length, and labels

are all inherited from the initial variable so all of

these have exactly the same labels and formats.

It is NOT my intention to use this feature; I do NOT

intend to create MXG variables with long names (at least

not in the near future!). MXG-provided invocations of

VMXGSUM will NOT exploit this capability; they continue

to use the current design of creating 8-character names

in the INCODE= operand.

But adding this feature to VMXGSUM may be useful in your

own invocations of VMXGSUM.

-Both Versions.

Number of obs were counted with SUM(MISS,NMISS) but now

_FREQ_ variable is used because it's better and avoids an

exposure in V7:

If INHERIT is used, the format of output variables

MISS/NMISS will have a DATETIME format if the first

variable in the VARIABLE statement happens to be a

DATETIME variable. Quite confusing, but not wrong!

Feb 16, 1999: Enhanced code after 16.08 version.
Change 16.313 IMS 6.1 support did not correctly decode type 08 records

VMACIMS for IMS usage via APPC, CPIC driven, due to a modified

VMACIMSA record layout. The quick circumvention is to replace the

Dec 19, 1998 line ELSE IF LINTSUB='.1......'B THEN DO;

with ELSE DO;

to ensure TRANSACT is always input. CPIC records caused

"INVALID DATA FOR YYYYDDD in line xx 65-68" messages.

Thanks to Johannes M. Laveuve, dvg Hannover, GERMANY.

Change 16.312 Support for DOS/VS Power V6.3.0 COMPATIBLY added fields

TYPEDOS LSTEPGN and LSTPGN for RECID='L' (dataset DOSLIST) and

Dec 19, 1998 INCOMPATIBLY inserted DESTUID in the RECID='M' and 'V'

(dataset DOSXRC) shifting LOCLNODE/ADJCNODE/NACACCT/

NACINCR/NACDLR. This change assumes you have V6.3.0;

if you have an earlier version, you will need to find

the line with comment "DELETE IF PRE 6.3.0" and delete!

Thanks to Trevor Ede, EDS (NZ) Ltd, NEW ZEALAND.

Change 16.311 A type in a LABEL for variable ASIESF had CSTORE, but the

VMACRMFV variable contains Average ESTORE for Swapped In users.

Dec 19, 1998 The label for ASIESF is now ESTORE, and variables ASIFMCT

and ASIFMCTI contain 'CSTORE' to be consistent and clear.

Thanks to Tim VanderHoek, Fidelity Investments, USA

Thanks to Juergen Holtz, IBM RMF Development, GERMANY

Change 16.310 Support for VM/ESA 2.3 (COMPATIBLE) changes to MTRSYS and

VMACVMXA USEACT segments adds four new variables to VXMTRSYS data

Dec 19, 1998 set and two new variables to VXUSEACT dataset.

Thanks to Steve Clark, California Federal Bank, USA.

Change 16.309 New variable MKDELDAO contains the original julian value

VMACEDGS for the MKDELDAT value (which becomes the data part of

Dec 19, 1998 datetime variable MKLCTIME). For special values that MXG

changed to 2099 in MKLCTIME, you will have the exact,

original value as a numeric in MKDELDAO.

Thanks to Sam Bass, McLane Co., USA.

Change 16.308 -Revision to RMM EDGR support creates numeric SAS date and

VMACEDGR SAT time variables for CREATE/LAST CHANGED dates/times so

Dec 19, 1998 normal SAS data manipulation can be done. Previously the

fields were carried as character variables. This change

is INCOMPATIBLE if you join old and new EDGR datasets

together, but this finalizes the support as it should be.

-Variables RVRETDAT, RVRELIXD and RVRELSI are now input.

Thanks to Don Friesen, ISM-BC, CANADA.

Thanks to Bob Lembo, American Stores, USA.
Change 16.307 -DB2 GTF-format records only. Using TYPEDB2G or PDB=GTF

VMACDB2 to read GTF records created zero observations in DB2ACCT

UDB2GTF and DB2ACCTB because SUBTYPE was missing for the ID=101

Dec 19, 1998 record. The GTF record doesn't have ID/SUBTYPE and MXG

has to create, but this GTF section was not updated when

the ID=101 record was split into subtypes. Most use of

GTF-format for DB2 is trace records, so this had little

impact for most sites.

-It turns out that not all segments for a single record

occur together in the GTF file. For example, the final

segment for record 160 came after the first segment of

record 161. This requires that the GTF file be SORTed

before it is input to UDB2GTF to reconstruct full length

records that MXG can read. The SORT FIELDS=(37,4,BI,A)

will do the trick. Comments added to UDB2GTF.

Thanks to Ted Blank, IBM, USA.

Change 16.306 -SAS Version 7 only. The new INHERIT parameter causes the

VMXGSUM attributes (LABEL, LENGTH, FORMAT, etc.) of same-named

Dec 19, 1998 variables to be propagated from input to output when PROC

MEANS/SUMMARY is used. Designed for MXG, INHERIT is now

invoked when VMXGSUM is run under V7; with it, VMXGSUM

can frequently bypass the last step.

Change 16.305 The APAF variables kept depended on the __NRLPS macro,

VMACAPAF but even if it was set to 10, only 8 engine's data was

Dec 18, 1998 kept. The logic was revised and the &__NRLPS macro was

eliminated; sixteen CPU variables will always exist now,

with online CPUs populating their set of variables.

Thanks to Steve Donahue, Blue Cross and Blue Shield of Texas, USA.

Change 16.304 MXG 16.04-16.07 only. Observations for RACFEVNT=19

VMAC80A (PERMIT COMMAND) were output into dataset TYPE8018

Dec 18, 1998 instead of dataset TYPE8019 due to a typo. The line

_ETY8018 after ELSE IF RACFEVNT=19 should be _ETY8019.

Thanks to Michael A. Yaffee, Duke Power, USA.
Change 16.303 Archaic members TYPEMONI and TYPETMON still had _L macro

TYPEMONI names instead of _W macro names, but since TYPEMONI is

TYPETMON for now-defunct version 7 and was replaced by TYPEMON8,

Dec 17, 1998 and since TYPETMON was replaced by TYPETMO2 for V2, this

slipped thru testing.

Thanks to Chris Weston, SAS Institute Cary, USA.

Change 16.302 The MACRO _LCIMSPGM CIMSPGM % statement should be

VMACCIMS MACRO _LCIMSPGM CIMSPROG %

Dec 17, 1998 The value was misspelled starting in 16.04.

Thanks to Chris Weston, SAS Institute Cary, USA.

Change 16.301 An extended example of AS/400 processing using FTP to

ADOCQAPM send and manage the multiplicity of OS/400 files that

JCLQAPMX must be moved to the SAS execution platform for TYPEQAPM.

Dec 16, 1998

Thanks to Clark Jennings, Reynolds Metals Company, USA.
Change 16.300 Unused Change Number.

======Changes thru 16.299 were in MXG 16.07 dated Dec 5, 1998======

Change 16.299 Variable MKDELDAT contained '14MAY2051'D for YY=99 and

VMACEDGS DDD GE 365; the statement YYYYDDD='20DEC2099'D; should be

Dec 5, 1998 YYYYDDD=2099365;, which is the julian date, because later

code converts the julian date to a SAS date.

Thanks to Sam Bass, McLane Co., USA.
Change 16.298 Variable DSGTAMXT was removed from PDB.CICINTRV because

VMXGCICI it only existed in CICS 3.3; variables STAMATHD,STAMITHD,

Dec 5, 1998 and STAHIWAT were put back in PDB.CICINTRV; they had been

inadvertently dropped.

Thanks to Chris Weston, SAS Institute ITSV, USA.
Change 16.297 MXG 16.07 Dec 4 only; end-comment */ was missing after

VMAC28 variable VRTSTATE, causing UNINITIALIZED variable and

Dec 4, 1998 incorrect data in NPMLANOD dataset.

Thanks to Freddie Arie, Lone Star Gas, USA.

======Changes thru 16.296 were in MXG 16.07 dated Dec 4, 1998======
Change 16.295 Enhancement to MXG RMF reporting replicates IBM XCF

ANALRMFR report. Specify REPORT=XCF,RPTOPT=USAGE MEMBER PATH.

Dec 4, 1998

Thanks to Jiann-Shiun Huang, CIGNA, USA.

Change 16.294 Cosmetic cleanup of variable labels.

VMAC28 -VMAC110. CICRELSE='CICS*RELEASE*IN JOURNAL*SEGMENT'

VMAC110 -VMACIKLA S23TTIME='SESSION*TERMINATE*TIME'

VMACILKA -VMACMEMO ACTMTEDA='ACCMTEDA='ACC NUMBER OF*PICKED UP*EDA'

VMACMEMO ACTMTETR='ACCMTEDA='ACC NUMBER OF*PICKED UP*ETR'

Dec 3, 1998 Variable SENDNCO now kept in MEMODIST.

(If you have the MEMO product, please send me

the current documentation, as there are other

variables that I do not have good labels for).

-VMAC28 VRTNRCPN='NEW*REMOTE*CP*NAME'

VRTNRNID='NEW*REMOTE*NETWORK*ID'

VRTRTCID='REMOTE*TCID'

Thanks to Chris Weston, SAS Institute ITSV, Cary.
Change 16.293 BETA93 Version 3.1.3 INCOMPATIBLY inserted fields in all

VMACBETA subtypes, causing INPUT EXCEEDED RECORD LENGTH errors,

Dec 3, 1998 and many fields previously in the valid SMFSTAMP8 format

Dec 17, 1998 (BETALT,BETASTSE,BETAENSE,BETASTRT,BETAEND) are now in a

never-before-seen-in-SMF time format of packed hhmmss00!

If you use SMFSTAMP8 to input these non-standard times,

there is no SAS error, but incorrect datetime values

are created. Their data value for 10:12:23 on 23NOV98

is '101223000098325F'x, but when INPUT with SMFSTAMP8

the result is a wrong value of 22Dec1998:04:57:20.64

because SMFSTAMP inputs the 10:12:23:00 time value as

PIB4. and gets 2696240.64 seconds (31.2 days) which is

then added to midnight on julian date 98325, 23Nov98!
READTIME fields should be unchanged, since BETA93 copies

rather than creates them. The below table lists the data

that has been verified that it was changed to new format,

and are supported in this version of MXG. The NODATAYET

subtypes will be printed on your SAS log if found in your

data; please email me that log so I can enhance MXG to

support and validate those NODATAYET subtypes.
Subtype datetime fields MXG dataset/status

0 BETALT BETA0 VERIFIED

1 BETABT, BETALT BETA1 NODATAYET

3 BETART BETA3 NODATAYET

5 BETALT BETA5 NODATAYET

6 BETAWST BETA6 NODATAYET

20 BETALT, BETASTRT, BETAEND BETA20 VERIFIED

21 BETALT, BETASTRT, BETAEND BETA21 VERIFIED

40 BETALT BETA40 NODATAYET

41 BETALT BETA41 NODATAYET

42 BETALT BETA42 NODATAYET

Thanks to Jurgen Niegsch, AUGI AG, GERMANY.

======Changes thru 16.292 were in MXG 16.06 dated Dec 2, 1998======
Change 16.292 Support for APAF Versions 4.1 and 4.3 (INCOMPATIBLE):

EXAPAFPO -Version 4.1 adds new variables to APAFSYSD dataset:

VMACAPAF CPUDEDF ='BIT MAP*OF DEDICATED*FENCED*CPUS'

Dec 2, 1998 CPUDEDG ='BIT MAP*OF DEDICATED*G-POOL*CPUS'

Dec 18, 1998 CPUDEDMP='BIT MAP*OF DEDICATED*CPUS'

CPUFENCE='BIT MAP*OF FENCED*C-POOL*CPUS'

CPUINSTL='BIT MAP*OF INSTALLED*CPUS'

CPUSHRMP='BIT MAP*OF SHARED*CPUS'

IOCD1DAT='IOCDS-1*DATE'

IOCD1NAM='IOCDS-1*NAME'

IOCD1NUM='NUMBER OF*THE ACTIVE*IOCDS-1'

IOCD1TIM='IOCDS-1*TIME'

TOTWEIGF='TOTAL*OPER SPECIFIED*WEIGHTS'

There are two elapsed fields, one in the System section

and one in the LPAR CPU Utilization section, and they

are not exactly the same:

LPAR Duration System Duration LPAR minus System

13:24.79 13:24.71 +.08

15:02.08 15:01.96 +.12

14:59.59 14:59.08 +.51

15:00.48 15:00.79 -.31

14:59.43 14:59.79 -.38

While awaiting an explanation from Amdahl, I keep the

value from the System section as variable DURATM. Amdahl

INCOMPATIBLY inserted TIMESLIC in LPAR CPU Util section,

but it already existed in the System Data Section; if it

is possible to have different timeslice values per LPAR,

then I will change MXG, but for now I keep the TIMESLIC

from offset 82 in the System Data Section.

-Version 4.1 adds new variables to APAFLPAR dataset:

ACTLCPS ='NUMBER OF*ACTIVE*LCPS'

CIDVALUE='CID*VALUE*ZERO*OR ONE'

DEDLCPS ='NUMBER OF*DEDICATED*LCPS'

OPRLCPS ='NUMBER OF*OPERATING*LCPS'

PCPMAP ='BIT MAP*CPUS*EXECUTED ON'

SHRLCPS ='NUMBER OF*SHARED*LCPS'

-Version 4.1 adds new variable to APAFCHN dataset:

CHPSCA ='SYSTEM*CHANNEL*ADDRESS*HEX'

-Version 4.1 creates new APAFLPAC dataset from the

subtype 32x LPAR CPU Utilization Data Subsection.

-Version 4.3 adds a new Pool Data Section, but the test

data was received in time for first release, but the new

APAFPOOL dataset was added Dec 18, from subtype 49 data.

Thanks to Robert Gilbert, National Power, ENGLAND.

Change 16.291 MXG 16.06 dated Nov 30 only. Member RMFINTRV had three

RMFINTRV typos (corrected in testing, but member not updated):

Dec 2, 1998 Line 226, remove SYSPLEX. Lies 361 and 362, add SYSPLEX

to both KEEP= lists.

Change 16.290 NPMLANOD dataset variable TIC_UTIL was missing because

VMAC28 variable DURATM, the denominator of TIC_UTIL equation in

Dec 1, 1998 the AOFTYPE='CSL' logic, was not calculated. The lines

of code inside 'CSL' that created ENDTIME and STARTIME

are moved above the TIC_UTIL calculation, and a new line

of code: DURATM=ENDTIME-STARTIME; was added. Also, the

test IF LENAOF GE 204 in 'CSL' was changed to GE 192.

Variable DURATM was also not calculated in the ETH, LSV,

and SAM segments; it is now calculated there, and is also

added to macros VA28CSL/ETH/LSV/SAM so it will be kept in

the datasets created by those segments.

Variable TIC_UTIL exists in dataset NPMLANOD but has

a missing value. You can create values without having

to re-reading SMF by running this program:

DATA PDB.NPMLANOD; SET PDB.NPMLANOD;

DURATM=ENDTIME-STARTIME;

TIC_UTIL=100*

((((LCSLBYTS*LCSLBTPT)+(LCSLBYTR*LCSLBRPT))/1000000000)+

(((LCSLTFRS*LCSLFTPT)+(LCSLTFRR*LCSLFRPT))/1000000))

/DURATM;
I recommend that you cut and paste those lines of code,

rather than trying to type them!

Thanks to Anke Mineur, dvg Hannover, GERMANY.

======Changes thru 16.289 were in MXG 16.06 dated Nov 30, 1998======
Change 16.289 Final testing of 16.06 left a PUT 'JOURN= ' ... statement

VMAC110 in the (archaic) CICS non-ESA code that could print many

Nov 30, 1998 lines on SAS log, with no other impact. Deleted the line.

Thanks to Ben Coonfield, Browning-Ferris Services, Inc, USA.

Change 16.288 In 16.04+ the sort order for all RMF datasets was changed

ANALRMFI to BY SYSPLEX SYSTEM STARTIME from BY SYSTEM STARTIME,

ASUM70PR but I failed to note the change, and I failed to change

GRAFRMFI the BYs in RMFINTRV until this change. The change is

MONTHBLD required because the same SYSTEM name can be in multiple

RMFINTRV SYSPLEX members. If you have only one SYSPLEX, this

VMAC7072 change has no impact, but if you have multiple SYSPLEXs,

VMAC71 you may have to change your BYs to BY SYSPLEX SYSTEM....

VMAC73 (Otherwise, your reports may fail with NOT SORTED error).

VMAC74 The member ANALRMFI (four PROC PLOTS) was change to add

VMAC75 SYSPLEX to the BY list. In the other ????RMFI members,

VMAC76 GRAFRMFI/TRNDRMFI/etc., the reporting is SORTED by SYSTEM

VMAC77 without the SYSPLEX variable to preserve your existing

VMAC78 reporting BY SYSTEM. Unless you actually have the same

VMAC79 SYSTEM name in two SYSPLEXes, this MXG choice will keep

WEEKBLD your existing reports BY SYSTEM and hopefully avoid any

WEEKBLDT need to change your reports.

Nov 30, 1998

Thanks to Richard S. Ralston, Whirlpool Corporation, USA.
Change 16.287 MXG 16.03-MXG 16.06. NPM datasets NPMIN25L and NPMIN25P

VMAC28 were misspelled as NPMINB5L and NPMINC5P beginning with

VMAC110 MXG 16.03, but are now spelled correctly. CICS Stats

Nov 30, 1998 datasets from Boole's MainView for CICS were misspelled

as CICBBxx instead of CICSBBxx, but are now correct.

Thanks to Chris Weston, SAS Institute ITSV, USA.

Change 16.286 MXG 16.06 only, CICS TS 1.2 only. INPUT STATEMENT EXCEED

VMAC110 error. Variables DS6SWT and DS6SCT should not have been

Nov 30, 1998 input at all. When I cleaned up the code for the sixth

CICS TCB, I inadvertently INPUT those two variables, but

they existed only back in CICS/ESA 3.3 and do not exist

in the CICS Transaction Servers (and only the CICS TS

releases have the sixth TCB!). Deleting all references

to variables DS6SWT and DS6SCT will correct my error.

Thanks to Thomas Wieland, Phoenix Home Life Mutual Insurance, USA.
======Changes thru 16.285 were in MXG 16.06 dated Nov 21, 1998======
Change 16.285 IBM TCP/IP undocumented fields changed the record lengths

VMACTCP and caused MXG's length-based logic to classify FTPSERVER

Nov 21, 1998 events as TCPSERVER. This revision no longer uses LENGTH

nor subtype; events are now identified based on data in

the record. While there are subtypes that theoretically

could be used, and the initial support for the new STAT

expected subtype 5, using subtypes would require a table

that you would have to update, since IBM did not see fit

to fix subtypes to events, but left them for you to set!

This redesign eliminates the need for subtype tables too.

Thanks to Randy Shumate, LEXIS-NEXIS, USA.
======Changes thru 16.284 were in MXG 16.06 dated Nov 18, 1998======
Change 16.284 Support for NPM Release 2.4 (INCOMPATIBLE)!

EX028VCS The good news is that NPM can capture IP resources from

EX028VMN TCP/IP connections. The bad news is that the NPM coders

EX028VRT INSERTED the new fields in the middle of the SCD segment,

FORMATS so many existing fields are trashed without this change.

IMAC28 -SCD segment inserted LSCDIPNM and LSCDIPPN incompatibly:

VMAC28 LSCDIPNM='IP ADDRESS*IF RESOURCE*IS IP' $15

VMXGINIT LSCDIPPN='IP PORT NUMBER*IF RESOURCE*IS IP' $5

Nov 18, 1998 Previously, the ACD/NCD/RCD/SCD segments used the same

MXG code, because they were congruent, but the new data

was inserted only in the SCD segment, so the MXG code was

restructured to separate SCD processing from ACD/NCD/RCD.

-VCD segment added three new fields compatibly:

LVCDCSMN='CSM*GROUP*NAME' $8

LVCDMNPN='MNPN*NAME' $8

LVCDRTPN='RTP*NAME' $8

-VVR segment added new variable

VVRNRBLK='TIMES WHEN*VR*WAS BLOCKED'

-Format MG028VA values 20x-26x (32-38 in decimal) are

actually 14x-1Ax (20-26 in decimal); format corrected.

-Format MG028NT, NPM Resource Type decoding, adds new

values: 14X='14x:IP'

6CX='6CX:CSM BUFFER POOL'

6DX='6DX:CSM STORAGE'

6EX='6EX:APPN TOPOLOGY'

6FX='6FX:APPN DIRECTORY SERVICES'

70X='70X:RTP DATA'

71X='71X:MNPS DATA'

-New subtype 'DC'x creates NPMVSVCS dataset:

LABEL='028VCS: VTAM CSM COMMON STORAGE MANAGER'

with buffer pool/storage/data space statistics

for 4k/16k/32k/64k areas.

CSM enables host applications to share data with VTAM