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

datasets. But then you had to update your IMACPDB each

time when I changed something, because your IMACPDB will

(still) override my defaults. But now, with the 16.04+

design, you can instead use the &ADDxxxx macro variables,

with %LET ADDxxxx= var1 var2 ... ; syntax, to add your

variables. Your %LET can be in member IMACKEEP or can

be in the SYSIN stream using the "MACKEEP=". So if you

have an IMACPDB member, compare it with the old member in

ZMACPDB to see what your site added, and use the %LETs to

add those variables and eliminate your IMACPDB.

The new macro variable names and their function are:

MACRO Variable For Dataset/Function

ADD6 TYPE6

ADD25 TYPE25 (JES3 only)

ADD26J2 TYPE26J2 (JES2 only)

ADD26J3 TYPE26J3 (JES3 only)

ADD30U1 TYPE30_1

ADD30U4 TYPE30_4

ADD30U5 TYPE30_5
ADDACT2 Account fields back-merged

from JOBS to STEPS/PRINT

(JES2 only)

ADDACT3 Account fields back-merged

from JOBS to STEPS/PRINT

(JES3 only)

The macros formerly defined in IMACPDB member are now

defined in members BUILD005 (JES2) or BUIL3005 (JES3).

Thanks to Mike A. Geiger, A. C. Nielsen, USA.
Change 16.340 Some variables in ENDEAVOR dataset ENDEVRSE were wrong

VMACENDV because the end-of-comment was missing after the INPUT

Jan 29, 1999 of variable EDVIFUNC, causing out-of-alignment.

Thanks to Kenneth D. Jones, SHL Systemhouse, CANADA.

Change 16.339 Type 42 subtypes 15-19 (VSAM RLS) now exist and test data

VMAC42 exposed IBM errors: their length field in the triplet has

Jan 29, 1999 total length, instead of the length of the segment, which

caused MXG to DELETE all the records with a WRONG LENGTH

message on the log because of this invalid length value.

MXG circumvents by constructing the segment length from

the total length and number of segments. However, IBM

has changed some fields after the early documentation,

and real data corrected several typos & logic errors so

that now each repeated segment will be output. Also, all

TYPE42xx datasets now have their _BTY42xx "By" and their

_STY42xx "Sort" macros implemented.

Thanks to Ed McCarthy, Health Insurance Commission, AUSTRALIA.
Change 16.338 The old &PDB70,&PDB71,...&PDB78 macro variable names were

ASUM70PR still used as the source DDNAME of the TYPE7xxx datasets

RMFINTRV that are read into RMFINTRV, but in the new 16.04+ design

Jan 28, 1999 those names should have been &PTY70,&PTY71... PTY78.

There was no execution error, because both sets of macros

variable names have default values of "PDB", but if you

tried to use the new 16.04+ tailoring to split your PDB

into multiple destinations, DATASET NOT FOUND errors

would result. All of the &PDB7xxx's are now &PTY7xxx's.

Note that &PDBRMFI is still the name for RMFINTRV.

-Same error in ASUM70PR, but &PDB70PR is now &PTY70PR.

Thanks to Linda Carroll, IBM Global Services, USA.

Change 16.337 -Support for these new NT objects:

EXNTBENC -DB2 NT DATABASE MANAGER object creates dataset DB2.

EXNTDB2 with one instance variable and 15 counter variables.

EXNTFAXG -FAX SR. GATEWAY MONITOR object creates dataset FAXGATEW,

EXNTMSCC with one Instance variable and 8 counter variables.

EXNTPCHB -VPN ADAPTER object creates VPNADAPT dataset with two

EXNTRADI Instance variables and 15 counter variables.

EXNTUSER -Terminal Server object USER creates dataset USER with

EXNTVINE one instance variable (USER) and the other fields in the

EXNTVPN PROCESS object (except for IDUSER/IDLOGON!).

FORMATS -Radius Server object creates dataset RADIUS, but only

IMACNTSM discovery records have been read.

VMXGINIT -Benchmark Factory object creates dataset BENCHMRK, but

VMACNTSM only discovery records have been read.

Jan 30, 1999 -Microsoft Exchange object MSExchangeMSCC created MSEXCHCC

with counts and rates of mail activity between Exchange

and Lotus CC:MAIL. Only Discovery have been read.

-Vines Communications object creates dataset VINES with

33 counters.

=Enhancement/correction to existing NT objects:

-PROCESS dataset variables IDLOGON and IDUSER had missing

values, because MXG had a test for NRDATA=20 instead of

NRDATA=21. The WIDE2MBE variable was renamed to CREATPID

when Demand Technology was able to discover what was put

in that undocumented field (Creation Process ID).

-WINPROXY dataset has two variables now removed from the

record (but MXG still creates them, with missing value,

so your reports won't die), and five new fields were

added (incompatibly, of course!).

Thanks to Jim Quigley, Consolidated Edison, USA.

Change 16.336 IBM Documentation for NPM 2.4 is wrong, causing STOPOVER

VMAC28 error with NPMSUBTY='DD'x record. The VTT section has a

Jan 26, 1999 documented length of 196 bytes, but records with only 176

bytes (thru VTTNNCDS) are being created by IBM. To avoid

the error, two lines were inserted in the VTT input:

VTTNNCDS &PIB.4. original line

@; inserted line

IF LENAOF GE 196 THEN INPUT inserted line

VTTNRRGC &PIB.4. original line

Mar 10, 2000: APAR OW37758 fixed the 2.4 error, but that

APAR was not applied to the 2.5 release, so APAR AW43578

exists for 2.5, with no PTF. See Change 18.032.

Thanks to Brian Crow, British Telecom, ENGLAND.
Change 16.335 Variables H15TCSDT and H15TCEDT in TYPERDSn RDS datasets

VMACRDS have never been correct, because MXG used JULDATE() where

Jan 22, 1999 it should have had DATEJUL(). No one noticed/complained,

probably because variable SMFTIME was valid, but now the

dates use DATEJUL() and its protection algorithm.

Thanks to Forrest Nielson, State of Utah, USA.

======Changes thru 16.334 were in MXG 16.09 dated Jan 20, 1999======
Change 16.334 Variable INNODE in TYPETPMX was changed to INNODEC, from

VMACTPMX numeric to character because INNODE was already defined

Jan 19, 1999 as numeric in other datasets created from SMF records.
Change 16.333 ADSM data transfer units were true Kilobytes so their

VMAC42 multiplier is changed from 1000 to 1024 for variables

Jan 17, 1999 ARCBYTCS,BKPBYTCS,DATBYTCS,ARCBYTRE and BKPBYTRE in

dataset TYPE42AD. I guessed the 1000 because most of

storage in IBM "KB" means 1000 rather than 1024, but

this guess was wrong.

Thanks to Kristyann Manske, University of Wisconsin-Milwaukee, USA.
Change 16.332 Variables DPRTY and IODP are both now formatted as HEX2.

VMAC99 to be consistent with historic priority ranges that are

Jan 17, 1999 commonly in hex (00-FF) rather than decimal.

Thanks to Chuck Hopf, MBNA, USA.

Change 16.331 "Cosmetic Audit" of duplicate variables in LABEL, LENGTH

DOC and FORMAT statements, incorrect dataset name in comments

Jan 17, 1999 and _L macro names where _W macro names belong, and other

similar typos were corrected in these members:

VMAC102 VMAC110 VMAC28 VMAC30 VMAC40 VMACAPAF

VMACCIMS VMACCIMS VMACCMF VMACCMFV VMACDCOL VMACHSM

VMACICE VMACILKA VMACIMS VMACIMSA VMACMWAI VMACMWUX

VMACNTSM VMACPW VMACRMFV VMACVMON VMACVMXA

Thanks to Freddie Arie, Lone Star Gas, USA.
Change 16.330 Status of MXG Year 2000 Support as of Jan 20, 1999.

ASUMHSM

TYPE80A MXG 16.09 is now required for full Year 2000 Support of

TYPEBETA all fields in all records from all supported products.

TYPEDCOL This replaces all prior statements of MXG Y2K Support.

TYPEHSM

TYPEMON8 MXG 16.01 and later do support almost all products, but

TYPEMOND five common products now require MXG 16.09 or later:

TYPEOPC HSM SMF Records

TYPETMON CA-1 aka TMS Tape Management Records

TYPETMS5 BETA93 SMF Records

TYPETMV2 Landmark Monitor for MVS Version 2

TYPETMVS VM Accounting records

TYPEVLFC Additionally, six other products that have variables

TYPEVM that were not Y2K compliant, but those variables were

YEAR2000 unlikely to have been used in your reports:

Jan 17, 1999 DCOLLECT - one julian date variable, UCCOLDT

Jan 20, 1999 OPC/A - three obscure julian date variables

Landmark - CICS: TYPEMON8,TYPETMON,TYPEMOND

MVS: TYPETMV2,TYPETMVS

one internal datetime TMMDCCLK.

ASTEX - one julian date SDATE

RACF SMF - two variables REVOKEDTE, RESUMEDTE

VLF - VLF Catalog records from SYSLOG.

Products were found with julian dates that were kept as

0cyyddd value (0100001 for 01Jan2000), which is a valid

Y2K format, but which is one not supported by SAS's

DATEJUL() function, so while MXG correctly created valid

Y2K values, your reports or exit logic would fail if they

used DATEJUL() on those 0cyyddd values. No ABEND, but

missing values or invalid argument messages on your log.
For example, TMS records have julian date values like the

EXPDT that are kept as 0cyyddd that could cause problems.

This DATEJUL() problem itself is not new. MXG reported

it and has provided an algorithm (described in member

YEAR2000, implemented by Change 15.050) that protects

julian dates as they are converted into SAS DATETIME

or DATE variables, which are Y2K compliant. But that

algorithm did not change the raw julian date value.

This change revises the protection algorithm so that it

now replaces the original 0cyyddd value with the valid

yyyyddd value in the raw julian date variable, unless the

yyyyddd value is missing (i.e., an invalid "date" like

99000), when the original raw value is not overwritten.
ASUMHSM testing with Y2K HSM records exposed the error,

and I realized that TMS and all products that keep julian

date variables were exposed, so every MXG member that had

a julian date field was examined, and I found four more

with the kept-julian-date-problem. I also found six

products (nine variables) that had been overlooked by MXG

Change 15.050 that were having unprotected sex with the

DATEJUL() function. All are now algorithm-protected.

Members with kept-julian-date-fields and the variables

that are now converted to yyyyddd:

TYPEDCOL UCCOLDT

TYPEHSM DSRDATE,VSRDATE,FSRBDATE,FSRDATR,FSRDLM,

FSRDLU

ASUMHSM DSRDATE

TYPEOPC TRLEVDAT,MT0CPED,MT0DAT

TYPETMS5 BTHDATE,CRTDT,DSADT,EXPDT,TMVADATE,LDATE,

TMAUDATE

TYPETMV2 JDRDATE

TYPETMVS (archaic) JDRDATE
Members overlooked in Change 15.050 and the variables

that are now protected:

TYPEBETA BETASTRT, BETAEND, BETASTSE, BETAENSE

TYPEDMON SDATE

TYPEHSM WFSRDATR, WFSRDATS, WFSRDATE

TYPETMV2 LMRKDATE, ENDTIME, STRTTIME

TYPETMVS (archaic) LMRKCARK

TYPE80A REVOKDTE, RESUMDTE

TYPEMOND TMMDCCLK

TYPEMON8 (archaic) TMMDCCLK

TYPETMON (archaic) TMMDCCLK

TYPEVLFC DATETIME

TYPEVM STARTIME, ENDTIME
In addition, thirty-five other members were changed

to streamline the algorithm where the creation of the

DATEYYYY variable was not required. These members

were and still are Y2K compliant, but are now more

robustly protected and their non-kept julian dates

are now converted to yyyyddd format:

TYPE90,TYPE84,TYPE434,TYPE37,TYPE1415,TYPE110,

TYPECTLT,TYPEFILA,TYPEIDMJ,TYPEIDMS,TYPEIMS,

TYPEIMSA,TYPEIPAC,TYPEITRF,TYPENDM,TYPENETP

TYPENSPY,TYPEOMVT,TYPEPDSM,TYPEROSC,TYPESIM,

TYPETRMS,TYPEWSF,TYPEXCOM,TYPEXSTR,TYPESUPR,

TYPESLRI,TYPERTEJ,TYPEDMS,IDMSLO57,IDMSJRNL,

ANALTMS,ANALSNAP,ANALRACF,ANALCM29

Thanks to John McCray, Huntington Services Company, USA.

Thanks to Xiaobo Zhang, Insurance Service Office, USA.
CHANGE 16.329 Support for OS/390 R 2.7 (COMPATIBLE) adds new subtypes,

EXTY746B records, and segments:

EXTY746F -RMF type 74 record has new subtype 6 which creates three

EXTY746G new MXG datasets describing Hierarchical File Systems.

FORMATS While storage size fields in the raw record are in pages

VMAC42 or MB, MXG converts them all to bytes and formats with

VMAC7072 MGBYTES so the storage variables have the same, standard

VMAC73 units that display as MegaBytes, GigaBytes, etc.

VMAC74 -TYPE746G - HFS Global Statistics - one per interval

VMAC79 R746G1C ='FIRST PAGE*FINDS*IN CACHE'

VMXGINIT R746G1NC='FIRST PAGE*NOT FOUND*IN CACHE'

VMAC109 R746GLRC='RETURN CODE*BPX1PCT*BUFFLIM'

EXTY109 R746GLRS='REASON CODE*BPX1PCT*BUFFLIM'

IMAC109 R746GMC ='METADATA*FINDS*IN CACHE'

TYPE109 R746GMNC='METADATA*NOT FOUND*IN CACHE'

TYPS109 R746GMNF='VALUE OF FIXED(MIN)'

Jan 16, 1999 R746GMXV='VALUE OF*VIRTUAL(MAX)'

R746GSFL='STATUS FLAGS'

R746GSRC='RETURN CODE*BPX1PCT*GLOBSTAT'

R746GSRS='REASON CODE*BPX1PCT*GLOBSTAT'

R746GUSF='PERMANENT*FIXED*STORAGE*IN USE'

R746GUSV='VIRTUAL*STORAGE*IN USE'

-TYPE746B - HFS Buffer Statistics - one per buffer pool

R746GBF ='BUFFER*WAS FIXED*PRIOR TO*IO REQUEST'

R746GBNF='BUFFER*WAS NOT FIXED*PRIOR TO*IO REQUEST'

R746GNDS='NUMBER OF*DATA SPACES*FOR BUFFER POOL'

R746GSB ='SIZE OF*BUFFERS*IN BUFFER POOL'

R746GSBN='BUFFER*POOL*NUMBER'

R746GSBF='SIZE OF*PERMANENT*FIXED*BUFFERS'

R746GSBP='SIZE OF*BUFFER*POOL'

-TYPE746F - HFS File System Detail Statistics

R742PSNM='FILE*SYSTEM*NAME*(DSN)'

R746F1C ='FIRST PAGE*WAS FOUND*IN CACHE'

R746F1NC='FIRST PAGE*NOT FOUND*IN CACHE'

R746FCTM='CURRENT*TIME*STAMP'

R746FIJ ='INDEX JOINS'

R746FINT='INDEX NEW TOPS'

R746FIRH='INDEX PAGE READ HITS'

R746FIRM='INDEX PAGE READ MISSES'

R746FIS ='INDEX SPLITS'

R746FIWH='INDEX PAGE WRITE HITS'

R746FIWM='INDEX PAGE WRITE MISSES'

R746FMC ='METADATA WAS FOUND IN CACHE'

R746FMNC='METADATA NOT FOUND IN CACHE'

R746FMTM='MOUNT*TIME*STAMP'

R746FPC ='DATA BUFFER BYTES CACHED BY THIS FILESYS'

R746FPD ='BYTES USED FOR ATTRIBUTE*DIRECTORY'

R746FPF ='BYTES*INTERNALLY*USED*BY HFS'

R746FRFI='RANDOM FILE DATA I/O REQUESTS'

R746FSF ='BYTE SIZE OF*FILE*SYSTEM'

R746FSFI='SEQUENTIAL FILE DATA I/O REQUESTS'

R746FSFL='STATUS*FLAGS'

R746FSNL='LENGTH OF*FILE*SYSTEM*NAME'

R746FSRC='RETURN CODE*BPX1PCT*FSSTATS'

R746FSRS='REASON CODE*BPX1PCT*FSSTATS'

-RMF type 70 record has new bit values for SMF70PFG that

MXG decodes into two new variables in TYPE70PR:

LPARCHDE='NUMBER OF*DEDICATED*PROCESSORS*CHANGED?'

LPARCHSH='NUMBER OF*SHARED*PROCESSORS*CHANGED?'

-RMF type 73 record has new CPMF (Channel Path Measurement

Facility) mode variable in dataset TYPE73:

SMF73CMI='CPMF*MODE'

which is decoded by the new MG073CP format:

0='0:CPMF IS NOT ACTIVE'

1='1:COMPATIBILITY MODE'

2='2:EXTENDED MODE'

-RMF type 79 record subtype 12 also has a new CPMF*MODE

variable in dataset TYPE79C

R79CCMI='CPMF*MODE'

that is also decoded by the new MG073CP format, above.

-SMF type 42 subtype 5 and 6 contain new field S42DSDAO

that MXG creates in datasets TYPE42DS, TYPE42SR, and

TYPE42VT as new variable named:

AVGDAOMS='AVERAGE*I/O DEV-ACTIVE-ONLY*MS PER SSCH'

-New SMF type 109 record is written by OS/390 Firewall

Server, and contains log messages of firewall activity

that appear to be quite useful to security folks. The

ICAxxxxx and ICACxxxx message structure will be decoded

when I have test data from an interested site; for now,

those messages are just carried as text strings.

These changes have not been tested with 2.7 data yet.
Change 16.328 For Percentile Goal Service Classes, a new MXG variable

VMAC7072 PCTMETGO reports the actual percentage of transactions

Jan 14, 1999 that met the goal. For example, take a Service Class

Sep 17, 2002 that has a goal value of 1 second. Its MAP values are

actually percentages of the goal and the product of the

MAP value and the goal value is the response time value.

If the goal is 90% in 1 second, and your data values are:
MAP .5 .6 .7 .8 .9 1 1.1 1.2 1.3 1.4 1.5 2 4 FF

TRN 8638 108 101 69 59 53 56 43 42 30 1 147 134 22

then 90.9% (8638/9503) of those transactions ended in the

first bucket, in one-half-second. You easily met your

90%-in-one-second goal; in fact 90% of your transactions

were satisfied in less than 50% of your goal value, so

your goal is being met at a level of 50% of the response

time goal, so IBM's PERFINDX is 0.5, "twice as good".

(See why you need to be careful with percentages,

and always include, "of what").

But if we sum the number of transactions in the first six

buckets (i.e., those transactions whose response time was

less than or equal to the goal value), there were 9028 of

the 9503 transactions (95%) that met the one second goal,

and the value of new variable PCTMETGO=95.00.

Thanks to Chuck Hopf, MNBA, USA.

Change 16.327 For Service Classes with Average Goal value that had no

VMAC7072 delay samples (i.e., PCUTSOU=0 and PCTDLTOT=0), VELOCITY

Jan 14, 1999 cannot be calculated so it was missing value, but MXG

then erroneously set PERFINDX to zero, even when it

could be calculated. The original logic:

IF (TRANS EQ 0 AND TRANSEXC GT 0) OR R723CRGF='A' THEN ..

IF R723CVAL LE 0 OR VELOCITY LE 0 THEN PERFINDX=0;

ELSE PERFINDX=AVGELPTM/R723CVAL;

END;

now has the test for VELOCITY removed:

IF (TRANS EQ 0 AND TRANSEXC GT 0) OR R723CRGF='A' THEN ..

IF R723CVAL LE 0 THEN PERFINDX=0;

ELSE PERFINDX=AVGELPTM/R723CVAL;

END;

Thanks to Chuck Hopf, MBNA, USA.
Change 16.326 TYPE72 for Compatibility Mode OS/390 R2.4+ was incomplete

VMAC7072 as new delay variables were input in the subtype 3 SMF

Jan 14, 1999 record for WLM Goal Mode, but were not input for the

subtype 1 Compat Mode record. These variables are new:

SMF72ADT='INELIGIBLE*AFFINITY*TIME'

SMF72CVT='JCL CONVERSION*TIME'

SMF72IOD='TOTAL*I/O*DELAYS*SAMPLES'

SMF72IOT='DASD*IOS QUEUE*TIME'

SMF72IQT='INELIGIBLE*OTHER*RESOURCE*TIME'

SMF72QDT='TOTAL*QUEUE*DELAY*TIME'

SMF72TSA='TOTAL*EXECUTION*SAMPLES'

Thanks to Rick Mansfeldt, GE Capital, USA.

Change 16.325 Documentation only; the instructions implied you needed

BLDNTPDB to run three separate BLDNTDPBs, one for DAY, WEEK, and

Jan 14, 1999 MONTH, but in fact, all you need is to use daily:

%BLDNTPDB(RUNDAY=YES,RUNWEEK=YES,RUNMNTH=YES);

and the daily will always be built, the weekly will be

built on the WEEKSTRT= date (default=MON), and the

monthly will be built on the first day of the month.

Thanks to Wayne Holzback, Reynolds Metal Company, USA.

Change 16.324 Variables R744FTIM, FSQU, FCTM, and FCSQ should have

VMAC74 been divided by 4096.

Jan 14, 1999

Thanks to Bruce Widlund, Merrill Consultants, USA.

Change 16.323 Change 16.271 was not supposed to change VMXGSUM, but it

VMXGSUM did; the comments around &DATETIME=DATETIME; statement

Jan 14, 1999 that were added by 16.271 have now been removed.

Thanks to Freddie Arie, Lone Star Gas, USA

Change 16.322 Support for CICS TS 1.3 (INCOMPATIBLE). IBM inserted

ADOC110 data fields in the middle of the accounting record, so

EXCICCF6 you MUST use MXG 16.09 or later to read CICS TS 1.3 data.

EXCICCF7

EXCICCF8 Lots of new measurements have been added, especially for

EXCICCF9 Web activity (and the first JAVA measurements in MVS!!!).

EXCICNC4 There are also new subtypes 3, 4 and 5, and new stat data

EXCICNC5 records and new fields added to existing statistics, and

EXCICSOR there are now 11 TCBs that CICS can dispatch!

FORMATS The type 110 subtype 1 (CICSTRAN dataset) has these 61

VMAC110 new variables added:

VMXGCICI ACTVTYID='ACTIVITY*ID'

Nov 29, 1998 ACTVTYNM='ACTIVITY*NAME'

Jan 12, 1999 BAACDCCT='ACTIVITY*DATA*CONTAINER*REQUESTS'

BAACQPCT='ACQUIRE*PROCESS*REQUESTS'

BADACTCT='DEFINE*ACTIVITY*REQUESTS'

BADCPACT='DELETE*ACTIVITY*AND CANCEL*REQUESTS'

BADFIECT='DEFINE*INPUT*EVENT*REQUESTS'

BADFTECT='DEFINE*TIMER*EVENT*REQUESTS'

BADPROCT='DEFINE*PROCESS*REQUESTS'

BALKPACT='LINK*PROCESS*ACTIVITY*REQUESTS'

BAPRDCCT='PROCESS*DATA*CONTAINER*REQUESTS'

BARACTCT='RESET*ACTIVITY*REQUESTS'

BARASYCT='RUN*PROCESS*ACTIVITY*ASYNC'

BARATECT='RETRIEVE*REATTACH*EVENT*REQUESTS'

BARMPACT='RESUME*PROCESS*ACTIVITY*REQUESTS'

BARSYNCT='RUN*PROCESS*ACTIVITY*SYNC'

BASUPACT='SUSPEND*PROCESS*ACTIVITY*REQUESTS'

BATOTCCT='TOTAL*DATA*CONTAINER*REQUESTS'

BATOTECT='TOTAL*EVENT*REQUESTS'

BATOTPCT='TOTAL*PROCESS*ACTIVITY*REQUESTS'

CFCAPICT='OO CLASS*LIBRARY*API*REQUESTS'

CFDTWACN='CF DATA*TABLE*WAIT*COUNT'

CFDTWATM='CF DATA*TABLE*WAIT*TIME'

CHMODECT='CICS*DISPATCHER*MODE*CHANGES'

CLIPADDR='CLIENT*IP*ADDRESS'

DB2CONCN='DB2*CONNECTION*WAIT*WAIT*COUNT'

DB2CONTM='DB2*CONNECTION*WAIT*WAIT*TIME'

DB2RDYCN='DB2*READY*QUEUE*WAIT*COUNT'

DB2RDYTM='DB2*READY*QUEUE*WAIT*TIME'

DB2REQCT='DB2*REQUEST*COUNT'

DB2WAICN='DB2*WAIT*WAIT*COUNT'

DB2WAITM='DB2*WAIT*WAIT*TIME'

DHCRECT ='DOCUMENT*CREATE*REQUESTS'

DHINSCT ='DOCUMENT*INSERT*REQUESTS'

DHRETCT ='DOCUMENT*RETRIEVE*REQUESTS'

DHSETCT ='DOCUMENT*SET*REQUESTS'

DHTOTCT ='TOTAL*DOCUMENT*REQUESTS'

DHTOTDCL='TOTAL*DOCUMENT*CREATED*LENGTH'

GNQDELCN='GLOBAL*ENQUE*DELAY*COUNT'

GNQDELTM='GLOBAL*ENQUE*DELAY*TIME'

IMSREQCT='IMS*REQUEST*COUNT'

IMSWAICN='IMS*WAIT*COUNT'

IMSWAITM='IMS*WAIT*TIME'

J8CPUTCN='USER TASK*J8 MODE*CPU TCB*COUNT'

J8CPUTTM='USER TASK*J8 MODE*CPU TCB*TIME'

JVMSUSCN='JAVA*VM*SUSPEND*COUNT'

JVMSUSTM='JAVA*VM*SUSPEND*TIME'

JVMTIMCN='JAVA*VM*EXECUTION*COUNT'

JVMTIMTM='JAVA*VM*EXECUTION*TIME'

L8CPUTCN='USER TASK*L8 MODE*CPU TCB*COUNT'

L8CPUTTM='USER TASK*L8 MODE*CPU TCB*TIME'

MAXOTDCN='MAX*OPEN*TCB*DELAY*COUNT'

MAXOTDTM='MAX*OPEN*TCB*DELAY*TIME'

MSCPUTCN='USER TASK*OTHER MODE*CPU TCB*COUNT'

MSCPUTTM='USER TASK*OTHER MODE*CPU TCB*TIME'

MSDISPCN='USER TASK*OTHER MODE*DISPATCH*COUNT'

MSDISPTM='USER TASK*OTHER MODE*DISPATCH*TIME'

PCDPLCT ='DPL*PROGRAM*LINKS'

PRCSID ='PROCESS*ID'

PRCSNAME='PROCESS*NAME'

PRCSTYPE='PROCESS*TYPE'

QRCPUTCN='USER TASK*QR MODE*CPU TCB*COUNT'

QRCPUTTM='USER TASK*QR MODE*CPU TCB*TIME'

QRDISPCN='USER TASK*QR MODE*DISPATCH*COUNT'

QRDISPTM='USER TASK*QR MODE*DISPATCH*TIME'

QRMODDCN='QR*MODE*DELAY*COUNT'

QRMODDTM='QR*MODE*DELAY*TIME'

RRMSURID='RRMS/MVS*UNIT OF RECOVERY*ID'

RRMSWACN='RRMS/MVS*WAIT*COUNT'

RRMSWATM='RRMS/MVS*WAIT*TIME'

RUNTRWCN='RUN*TRANSACTION*WAIT*COUNT'

RUNTRWTM='RUN*TRANSACTION*WAIT*TIME'

S8CPUTCN='USER TASK*S8 MODE*CPU TCB*COUNT'