* copyright (C) 1984-2019 merrill consultants dallas texas usa
Yüklə 28,67 Mb.
|
|
Number of Private discarded pages eligible for Page Steal. Currently backed frames which are still charged to DB2, minus this count, is the true REAL Storage usage at this time. I can not find the "currently backed frames" fields and have asked IBM for help; this note will be updated when the correct fields are known so the usage variable can be created. IBM DB2 Support response Nov 30: All IFCID225 fields are described in the dataset 'DSNB10.SDSNIVPD(DSNWMSGS)'. It does not look like we capture that statistic. The values are captured from RMF and are included with DB2 statistics records to assist with reporting. Anything not contained will still be available in RMF. -Dec 6: UNINIT PRISTGDPAGE and correct spell as _PAGE. Change 35.276 Support for CICSTRAN User field RFSEMP01/RFSDATA creates IMACICWX seven variables, RFSEMP01F1-RMSEMP01F7. IMACAAAA UTILEXCL
VMAC110 Nov 27, 2017 Change 35.275 Addition of TYPE70TR dataset required protection in the ANALRMFR PDB=SMF part of the program to prevent the Nov 26, 2017 ERROR: No dataset open to look up variables. when the _STY70TR was executed without prior build. Change 35.274 Support for APAR OA53110 adds new variables: VMAC42 S42DSRRU='AVG RESPONSE*RANDOM*READ*CACHE' Nov 22, 2017 S42DSRSU='AVG SERVICE*RANDOM*READ*CACHE' SMF42IFW='AVERAGE*FAST-WRITE WAITS*PER MINUTE' SMF42IHR='AVERAGE*HIT*RATIO'
VMAC74 resource is Virtual Flash Memory. Nov 22, 2017
VMXGSUM subsequent PROC steps would automatically find the output Nov 22, 2017 of VMXGSUM as the last dataset created, but if you added any dataset options like (KEEP or (INDEX then while the dataset was correctly created, an error message was generated that either told you the dataset name was invalid or that it exceeded 42 bytes depending on the SAS version you were running. NOTE: there must be a space between the dataset name and any options you choose to specify.
Thanks to Robert Gilbert, BNP Paribas Fortis, BELGIUM. Change 35.271 Support for Container Pricing in SMF 89 records creates EXTY89R1 New variables in TYPE89 and TYPE892: EXTY89R2 SMF89COREMODECP='CPUS*ACTIVE*ON CP*CORE' EXTY89TI SMF89COREMODEZAAP='CPUS*ACTIVE*ON ZAAP*CORE' IMAC89 SMF89COREMODEZIIP='CPUS*ACTIVE*ON ZIIP*CORE' VMAC89 New variables in TYPE892: VMXGINIT SMF89CURREGS ='INSTANCES*OF CURRENT*REGISTRATIONS' Nov 27, 2017 SMF89TRGREGS ='INSTANCES*OF CURRENT TRG*REGISTRATIONS' SMF89DELTAREGS='INTERVAL*DELTA*CURRENT*REGISTRATIONS' SMF89DELTATRG ='INTERVAL*DELTA*TRG*REGISTRATIONS' New dataset TYPE89TI 'INTERSECTION TENANT RESOURCE GROUP' PRODOWNR= 'SMF89TCPO PRODUCT*OWNER' PRODNAME= 'SMF89TCPN PRODUCT*NAME' PRODVERS= 'SMF89TCPV PRODUCT*VERSION' PRODQUAL= 'SMF89TCPQ PRODUCT*QUALIFIER' PRODID = 'SMF89TCPI PRODUCT*ID' SMF89TIPO 'INTERSECTING*PRODUCT*OWNER' SMF89TIPN 'INTERSECTING*PRODUCT*NAME' SMF89TIPV 'INTERSECTING*PRODUCT*VERSION' SMF89TIPQ 'INTERSECTING*PRODUCT*QUALIFIER' SMF89TIPI 'INTERSECTING*PRODUCT*ID' SMF89_TRG 'TENANT*RESOURCE*GROUP' SMF89TCFG 'TENANT*USAGE*ENTRY*FLAGS' SMF89TCCT 'TENANT*PRODUCT*INTERSECT*CP TCB TIME' SMF89TCZT 'TENANT*PRODUCT*INTERSECT*ZIIP TIME' New dataset TYPE89R1 'TENANT RESOURCE GROUP DATA' SMF89TIPO='INTERSECTING*PRODUCT*OWNER' SMF89TIPN='INTERSECTING*PRODUCT*NAME' SMF89TIPV='INTERSECTING*PRODUCT*VERSION' SMF89TIPQ='INTERSECTING*PRODUCT*QUALIFIER' SMF89TIPI='INTERSECTING*PRODUCT*ID' SMF89_TRG='TENANT*RESOURCE*GROUP' SMF89TCFG='TENANT*USAGE*ENTRY*FLAGS' SMF89TCCT='TENANT*PRODUCT*INTERSECT*CP TCB TIME' SMF89TCZT='TENANT*PRODUCT*INTERSECT*ZIIP TIME' New dataset TYPE89R2 'TENANT RESOURCE GROUP DATA' SMF89TIPO='INTERSECTING*PRODUCT*OWNER' SMF89TIPN='INTERSECTING*PRODUCT*NAME' SMF89TIPV='INTERSECTING*PRODUCT*VERSION' SMF89TIPQ='INTERSECTING*PRODUCT*QUALIFIER' SMF89TIPI='INTERSECTING*PRODUCT*ID' SMF89_TRG='TENANT*RESOURCE*GROUP' SMF89NRTRG'CURRENT*TRG*REGISTRATIONS'
EXTY70TR new TYPE70TR dataset with these Tennant Resource Group VMAC7072 variables in APAR OA52694: VMXGINIT TRG_NAME ='TENANT*RESOURCE*GROUP*NAME' Nov 22, 2017 TRG_DESC ='TENANT*RESOURCE*GROUP*DESCRIPTION' TRG_TNTID ='TENANT*IDENTIFIER' TRG_TNTNAME='TENANT*NAME' TRG_SBID ='TENANT*SOLUTION*ID' TRG_SUCP ='TENANT*CP*MSU*UNITS' TRG_SUIFA ='TENANT*ZAAP*MSU*UNITS' TRG_SUSUP ='TENANT*ZIIP*MSU**UNITS' TRG_SULAC ='TENANT*CP 4HR*AVERAGE*MSU' Change 35.269 Support for CICS User field USERPRC1/WANLUPRC. IMACAAAA
IMACICWV PRODTEST
UTILEXCL VMAC110
Nov 21, 2017 Change 35.268 SAS 9.1.3 SP4 (SAS (R) 9.1 (TS01.01M3P02022006) failed SAS* with ERROR: OBTAIN FAILED FOR FILE SMF, RC=24. because Nov 19, 2017 the new parm EATTR=OPT was enabled for non-VSAM datasets, so they can reside on EAV volumes, but that was not supported until SAS 9.2 Thanks to Jeffery Kirsch, Compuware, USA. Change 35.267 -DB2 Netezza IDAA variables Q8STDSKB and Q8STDSKU were VMACDB2 both wrong; Q8STDSKU was incorrectly multiplied and DSKB Nov 17, 2017 was missing that multiplication. -Variables Q8STINSC/UPDC/DELC/DRPC/CRTC/CMTC/RBKC/OPNC WERE ALL WRONGLY SET EQUAL TO Q8STACPU. Thanks to Giuseppe Giacomodonato, EPVTECH, ITALY. Change 35.266 SMF ID=2 SYSTEM=DUMY "SMF Signature Enabled" records that VMACSMF are only 14 bytes were still created after APAR OA50483 Nov 17, 2017 was installed, causing INPUT STATEMENT EXCEEDED error, as MXG expected these records to contain additional data in either a subtype 1 or 2 record. Now, MXG tests for the length=14 and deletes these 'DUMY' records, silently. The APAR reported the exposure was only when Logstream data is read, and infrequently. In IBM's unstated defense, the "subtypes are valid" bit wasn't turned on, justifying the original 14 byte length, but MXG had keyed off that unique system name of DUMY. If you see the DUMY in a hex dump, you can circumvent with MACRO STOPOVER MISSOVER % as your first //SYSIN. to prevent the ABEND until you have this VMACSMF update. Thanks to Paul Volpi, UHC, USA. Thanks to Brian D. Peterson, UHC, USA. Thanks to Donald R. Striegel, UCH, USA.
VMACDB2 line 4057 (BPHITRAT=.;) should have been deleted. You Nov 14, 2017 can recalculate in your reporting using IF QBSTGET GT 0 THEN BPHITRAT= (QBSTGET-(QBSTRIO+QBSTSPP+QBSTDPP+QBSTLPP))/QBSTGET; ELSE BPHITRAT=.; Thanks to Rick Southby, Insurance Australia Group, AUSTRALIA.
VMAC110 dataset, have the sum of DSGTDT+DSGTWT, DISP+WAIT time VMXGCICI that is greater than the interval DURATM, with DSGTWT as Nov 14, 2017 much as 5900 seconds for a 3600 second interval (and the DSGTDT in those segments are milliseconds or less); that sum was used calculate STARTIME and DURATDS. Now, DURATM is stored in DSGTWT and used for calculations, and then DSGTWT=DURATM-DSGTDT recalculates the possible wait time. See Change 36.076. -The message "ERROR: IF YOU USE CICINTRV..." when MXG detected the condition (DURATM GT INTERVAL REQUESTED) is change to "MXGWARN:..." as few actually use CICINTRV. -This data has only been seen from ancient CICS/TS 4.2. Thanks to Ed Wieszczek, Zions Bank, USA. Change 35.263 If you didn't specify a SORTBY= parameter for the ACCOUNT ANALDB2R report, it could fail trying to resolve a macro variable Nov 10, 2017 that did not exist. Thanks to Harald Seifert, HUK-COBURG, GERMANY. Change 35.262 New DB2 zPARMS added include FORMATS QWP4MUDI='MORE*UNION*DISTRIBUTION' which is decoded by VMAC102 a new format Nov 22, 2017 QWPRSTACS='STATCKGSRT' QWP4BSACP='ALTERNATE*CP' QWP4UDBSG='UTIL*DBBSG' QWP4ULBSG='UTIL*LGBSG' QWP4CYFR ='COPY*FAST*REPLICATION' QWP4DDLM ='DDL*MATERIAL*IMMED*PEND' QWP4CDSTL='CACHDYN*BOTH*CAPTURE*LOAD*NONE' QWP4ZHYPL='ZHYPERLINK*ENABLE*DISABLE*BDATABASE'
VMACTPMX the 700 WHEN clauses into 13 subgroups of first letter of Nov 8, 2017 the WHEN argument text value. Thanks to Kurt Gramling, TSYS, USA. Change 35.260 BVIR History file updated for formats for 3.3 media codes FORMATS and BVIR302 fields were corrected. VMACBVIR Nov 9, 2017 Thanks to Spain.
VMACRMFV with these counts of Million Service Units Nov 8, 2017 ZRBASICPUMSU='ZRBASI*CPU MSU*COUNT' Nov 12, 2017 ZRBCPUCPUMSU='ZRBCPU*CPU MSU*COUNT' Nov 18, 2017 ZRBCPUZIPMSU='ZRBCPU*ZIP MSU*COUNT' Nov 24, 2017 ZRBLCPCPUMSU='ZRBLCP*CPU MSU*COUNT' Nov 30, 2017 where the Software MSU Coefficient CPUPHYAD is used by ZRBCPUCPUMSU=CPUPHYSI*CPUPHYAD/1000000; So, an LPAR in a CEC with CPUPHYAD=20000, with the CPU Partition Dispatch time of 15 seconds in an interval, would have an MSU Count = (15*20000)/1000000 =0.3 MSU. If the interval duration was one minute, the IBM ACT "Actual" MSU on the RMF III CPC report, a projection of this interval's value to an hourly total as if all intervals were this value, would be 60*0.3=18 MSU per hour, which is the value in this new variable: ZRBLCPCPUMSUHR='ZRBLCP*IBM ACT MSU*PROJECTED*HRLYMSU' -Dataset ZRBLCP contains data on ALL LPARS in a SYSPLEX, reading data from only one SYSTEM in that SYSPLEX, but it has an obs for each LCPUADDR in each LPAR. This change creates new ZRBLCPLPAR dataset when ZRBLCP is sorted, with the LPAR totals for each LCPUPRTY engine for each interval, but there is no 4HR AVG MSU variable in ZRBLCP. To create ZRBLCPLPAR the CPUG3, CPCDB and SSHG3 tables are needed. -The actual IBM 4HR MSU (CPUAVG4H) is in ZRBCPU dataset at one minute or even 30 second intervals, but you have to read the data from every system to populate ZRBCPU for all LPARs. -MXG 35.10, the deaccumulation of the six ZRBASI variables (actually added by z/OS 2.2) was not sufficiently tested and could have incorrect values in these variables: ASICPUTA_LF ASITCBTA_LF ASIIOCNT_S ASITRCA_S ASITET ASITRT for jobs that have duplicate names with different ASID-NR and only if you used TYPSRMFV or _SZRBASI to sort ZRBASI. The new-in-35.10 MXG deaccumulation of CPU_LF fields by _SZRBASI failed to include the ASID number, PERIOD, and JCTJOBID to deaccumulate those six variables. The ZRBASI dataset created by TYPERMFV was not in error. -As documented in Change 35.249, the value in ASICPUTA is is larger with 35.10 because the higher resolution CPU in ASICPUTA_LF is stored in ASICPUTA, with ASICPUTA_ORIG keeping the original lower value. -Nov 30: LPARNAME in ZRBCPU is now always populated. Change 35.258 35.09-35.10, Macro Language error, missing double periods ANALID in line ANALID: &PDBMXG..SMFRECNT DOES NOT EXIST; Nov 7, 2017
VMACPOEX the record should have 21 POEX segments but has only 3, Nov 7, 2017 and the last segment is only 90 versus 95 bytes. This change will be updated when the vendor's records are correct. This was from 9.6.1, but I've recently read that version's records with no errors. Thanks to Tracey Davidson, USBank, USA.
VGETDDS "concatenated" PDB data libraries. VMXGSET // EXEC MXGSASV9 Nov 7, 2017 //PDB1 DD //PDB2 DD %VGETDDS(DDNAMES=PDB: ); DATA MYVIEW/VIEW=MYVIEW; %VMXGSET(DATASET=MNTHJOB); PROC SQL; SELECT YEAR(DATEPART(JINITIME)) AS MYYEAR LABEL= 'YEAR', MONTH(DATEPART(JINITIME)) AS MYMONTH LABEL = 'MONTH', SYSTEM, JOB, TYPETASK , ACCOUNT1, SUM(NORMCPU) AS TOTALNORMCPU , JOB FROM MYVIEW WHERE JOB LIKE 'MYJOB%' AND MONTH(DATEPART(JINITIME)) = 09 GROUP BY JOB, SYSTEM ; Thanks to Paul W Schreiber, AT&T, USA. ====== Changes thru 35.255 are in this MXG 35.10 dated Nov 6, 2017===== Change 35.255 -MXG 35.10, Change 35.240 DIVIDE BY ZERO error when the VMAC42 IOCCOUNT=0 in TYPE42SR, now protected, but had no impact Nov 6, 2017 on the TYPE42SR dataset, DCMEPCT still missing. Nov 14, 2017 -Subtype 5 INPUT STATEMENT EXCEEDED when 68 bytes were added for the SYNC segment, but SYNC Offset was zero. Thanks to Jim Horne, Lowe's Companies, Inc., USA. Thanks to Stan Adriaensen, AXA-Tech, BELGIUM.
VMACDB2 incorrectly multiplied by 4096. Nov 5, 2017 Thanks to Rick Southby, IAG, AUSTRALIA. Change 35.253 Some of the macro variables coming back from PROC SQL PDBAUDIT are longer than the 262 character literal length since Nov 5, 2017 the spaces count and could generate spurious messages particularly in the QA stream. Change 35.252 MXGWARN message variable JESNR missing value and blank VGETJESN TYPETEST is suppressed for records with SUBSYS='SMS' and Nov 4, 2017 JCTJOBID='INIT'; variable values are unchanged as those records for tasks in initiation have neither. VGETJESN already suppressed warning for JCTJOBID='MSTR'.
EXCICASG counters to the CICVT (VTAM) dataset, new STID=149 record VMAC110 creates new CICASG Statistics for the AS Domain. VMXGINIT
Nov 3, 2017 Thanks to Perry Lim, Union Bank, USA. Change 35.250 Support for Thruput Manager fields JBAACT JBDEA JLIMT and VMACTPMX REQUIRED. Nov 4, 2017 Thanks to Kurt Gramling, TSYS, USA. Change 35.249 -Support for z/OS 2.3 RMF III CPUG3 record dataset ZRBCPU VMACRMFV changed CPUHOOFF offset value caused some variables to Nov 3, 2017 be wrong, notable the Capacity Group Name and adjacent. -Variable ASICPUTA in ZRBASI dataset will be larger with this change, as IBM has added new accumulated ASICPUTA_LF (long float) field with higher resolution, which is now deaccumulated and REPLACES ASICPUTA's original value. New variable ASICPUTA_ORIG contains the original value. The ASICPUTA_LF was 18% larger than ASICPUTA_ORIG. -These accumulated fields are also now deaccumulated: ASITET ASITRT ASITCBTA_LF ASIIOCNT_S ASITRCA_S Thanks to Kurt Gramling, TSYS, USA. Change 35.248 Variables QWA225SS and QWB225SS are now kept in DB2STATS, VMACDB2 and two new-in-DB2 V12 REAL2G variables are input/kept. Nov 1, 2017 QWA225SS='DBM1*31-BIT*IN-USE*SYSTEM*AGENTS' QWB225SS='DIST*31-BIT*IN-USE*SYSTEM*AGENTS' QWA225HVPAGESINREAL2G='DBM1*HVPAGES*IN*REAL2G' QWB225HVPAGESINREAL2G='DIST*HVPAGES*IN*REAL2G' Thanks to Rick Southby, IAG, AUSTRALIA.
ANALDB2R variable message Oct 30, 2017 -If you used INTERVAL=, BEGTIME=, or ENDTIME= and did not specify a SORTBY= the default is BY QWHSSSID QWHCPLAN and QWHCAID, so datetime variable QWACBSC is not carried forward and no accounting report was created; this could also cause duplicate variables in SORTBY list if QPACPKIC PACKTYPE or QWHSTCK were in the tailored SORTBY list. Change 35.246 -SMF 113/HIS formulas for the z14 were updated by IBM with ASUM113 L3P changed, which is also impacted the RNI value which VMAC113 uses L3P. VMACVMXA -The z13 code set the SM1132SP Speed value to 5000 because Oct 27, 2017 the value was wrong, but that code was removed for the z14, as those records contain the expected 5208 MHz. -John's updated formulas are available at http://www.ibm.com/support/techdocs/atsmastr.nsf/WebIndex/TC000066 Thanks to John P. Burg, IBM, USA. Thanks to Nick Varley, SYNCSORT, USA. Change 35.245 SMF 119 Subtype 81 INPUT STATEMENT EXCEEDED because MXG VMAC119 expected full 4096 length for DIRU and DORU fields; now Oct 27, 2017 DILEN and DOLEN length of text are used for $VARYING4096. Thanks to David Campbell, Sun Trust, USA. Change 35.244 New parameters TRNDKEEP and SPINKEEP added to let you VMXGALOC control how many copies of each are retained. Oct 27, 2017
FORMATS CICS Product Number; format MGIBMCI needed a new entries Nov 1, 2017 with 70='5655-Y04' and 71='5655-Y04'. Other related formats were also updated. Thanks to Patrick J. Holloman, Navy Federal Credit Union, USA. Change 35.242 Support for Power Exchange Version 10.1.1. VMACPOEX
Oct 25, 2017 Thanks to Jan Tielemans, KBC, BELGIUM. Change 35.241 -Typo, NEDNC=SMFTIME should be NENDC=SMFTIME. ANAL119 -Uninit variables and incorrect date corrected 12/31. Oct 23, 2017 Dec 31, 2017 Thanks to Jon Whitcomb, Great Lakes Educational Loan Service, USA. Thanks to Earl Kline, Luminex, USA. Change 35.240 Support for APARs OA52132, OA52133, and OA61734, some of VMAC42 which fields were listed in the z/OS 2.3 SMF Manual. Oct 23, 2017 CODE HAS NOT BEEN TESTED WITH NEW RECORDS. See Change 35.288. Change 35.239 MXG 34.09. RMFINTRV fails with macro GOT70PR not resolved VMXGRMFI when the PDB did not contain the expected PDB.TYPE70PR. Oct 22, 2017 Thanks to Tracy Davidson, USBank. USA. Change 35.238 A typo BASEWEEJ should have been BASEWEEK - only affected VMXGALOC the aging off of old WEEK directories. Oct 22, 2017
VGETSORT a spurious message about an invalid DO loop was printed. Oct 20, 2017
DEDUP701 due to the Split 70s duplicates, added SMT merges and the Oct 20, 2017 multiple datasets that create TYPE70 that prevent the use of the normal NODUP in the final sort to remove dupes. Hash logic from ANALDUPE is executed in the IMACFILE exit as SMF is read, examining only SMF 70 subtype 1, deleting duplicates prior to their input and reporting the first three DUPES on the log if any were found. You enable the 70 subtype 1 duplicate removal with //SYSIN DD * %let macfile=%quote(%include sourclib(dedup701);); Thanks to MP Welch, Bank of America, USA. Thanks to Garth Bloomfield, DXC Technology, AUSTRALIA. Thanks to Peter Gray, DXC Technology, AUSTRALIA. Change 35.235 RMF III ZRBCPU is enhanced with CPC_HOMEFLAG decoded: VMACRMFV CPCCAPAVAIL='CAPACITY*VALUES*AVAILABLE?' Oct 20, 2017 CPCVARYCPU ='VARYCPU*OPTION*SET?' CPCLPARMGT ='WLM*LPAR*MANAGEMENT*ENABLED?' CPCMTMETRIC='MULTI*THREADING*METRICS*AVAILABLE?' CPCABSMSU ='ABSMSU*CAPPING*OPTION*SET?' Thanks to MP Welch, Bank of America, USA.
BUILD005 all variables in JOBS/STEPS/SMFINTRV/NJEPURGE/PRINT. BUIL3005 Macros _KDBJOBS/_KDBSTEPS/_KDB30UV/_KDBNJEP/_KDBPRIN EXPDBKEP were defined but were not referenced. You put all your Oct 21, 2017 definitions in EXPDBKEP in your USERID.SOURCLIB using MACRO _DBJOBS KEEP= A B C D . . % or MACRO _DBJOBS DROP= A B C D . . % and then you instantiate them in BUILDPDB SYSIN using %LET MACKEEP= %QUOTE( %INC SOURCLIB(EXPDBKEP); ); %INCLUDE SOURCLIB(BUILDPDB); Thanks to Thomas Orlando, UBS, SWITZERLAND. Change 35.233 Protection for truncated SMF 80 Extended Relocate segment VMAC80A with 12 fields expected but only 10.5 fields are in the Oct 19, 2017 SMF record. MXGERROR for the first three instances. Record may have been truncated by ftp processing. Change 35.232 Documentation. The zIIP CPU time for BMC Utilities is TYPE30 not recorded in SMF 30 records for the JOB/ASID of the Oct 19, 2017 Batch Utility job, but is in the 30s for BMC's XBM Started Task. IBM Utilities do record zIIP CPU time in the SMF 30 for the batch job.
VMAC80A using the SMF80DTP/RACFTYPE=44 relocate segment, using VMXGINIT segments with the subkeyword/EV44TXT='ORIGINATED_FROM'. Oct 18, 2017 MXG now populates variables NODE80A USERID and DIRECTED Dec 7, 2017 with values of DIRECTED_BY_AT, DIRECTED_BY_ONLY_AT or DIRECTED_AUTOMATICALLY. You would use this syntax: %LET RACFDIRECTED= %QUOTE( IF NODE80A IN ('NODE1','NODE2') AND USERID IN ('USERID1','USERID2') THEN DELETE; ) ; %INCLUDE SOURCLIB(TYPS80A); -Dec 7: NODE was changed to NODE80A to avoid a conflict if TYPE80A and TYPE6 were used together. The three fields are created for the test but are not kept. Thanks to Kerry J. Sommers, John Deere, USA. Thanks to Joan T. Keemle, John Deere, USA. Thanks to Francois Vancoppenolle, P&V Group, BELGIUM.
GRAFCEC Avg consumption. GRAFCEC now creates all of the charts GRAFLPAR previously produced by GRAFLPAR, plus some new ones, thus Oct 17, 2017 GRAFLPAR obsolete. And GRAFCEC now allows multiple input libnames. Dataset ASUMCELP must exist in the first data library and must have non-zero obs, or GRAFCEC will die with a dataset not found error. -GRAFLPAR error if you specified PDB=PDB PDB1 and using SAS/GRAPH, it failed trying to write the graphics catalog to two libnames. Catalog will only be written to the first of the two or more libnames specified by PDB=. But note GRAFLPAR is now obsolete, replaced by GRAFCEC. Thanks to Daniel Mckinzie, Zions Bank, USA.
VMACDB2 now protects for lost/skipped intervals of input data for Oct 23, 2017 for these datasets: DB2STATB DB2STSBP DB2GBPST DB2NETZA DB2STAT5 DB2STAT0 DB2STAT1 DB2STATR. Large gaps (like missing a day) with repeated values in QWHSISEQ caused large DURATM which caused BEGTIME to be a different date. Since DB2 SMF 100 stat records can only Yüklə 28,67 Mb. Dostları ilə paylaş:
|