* copyright (C) 1984-2019 merrill consultants dallas texas usa
Yüklə 28,67 Mb.
|
|
Summary report when an SPG table Internal Error was detected. -Except for the DSIG3 table any RMF III table id mismatch errors now result in an Abend U0998 Reason Code 2. A table id mismatch is a serious non-recoverable data error. A DSIG3 error can be recovered by processing the next RMF III data set but is still very undesirable.
documentation to explain where in the input stream it should appear. -Numerous documentation updates in: Section 4 "RMF III Table Selection Parameters" Section 5 "Input Data Selection Parameters" Section 6 "Report Control Parameters" Section 7 "Output Data Control Parameters" Section 8 "Error Handling Parameters" Section 12 "Messages" Section 26 "ASMRMFV and MXG PDB Data Relationships"
EXMGCR34 subtype 3 and 4 replacing old records but keeping all of IMACMGCR the original fields and adding these new variables: VMACMGCR MGCRFUNC='FUNCTION*E ENCRYPT*D DECRYPT ?' VMXGINIT MGCRALGO='ALGORITHM*USED' Dec 8, 2018 MGCDSSUF='ADRDSSU*DUMP*RESTORE*COPYDUMP' MGCRDSN1='INPUT*DSN' MGCRVOL1='INPUT*VOLSER' MGCRSTY1='INPUT*FILE*TYPE' MGCRDSN2='OUTPUT*DSN' MGCRVOL2='OUTPUT*VOLSER' MGCRSTY2='OUTPUT*FILE*TYPE' Thanks to Jennifer D. Ayers, West Virginia State Goverment, USA. Change 36.239 Enhanced to display all used parameters, only on the VMXG2DTE first execution in a SAS session. On subsequent execute Dec 7, 2018 DDIN DDOUT PDB DATASET and INITIT are displayed only if Dec 23, 2018 they are different and others only if present. Thanks to Tom MacCabe, Dominion Energy, USA Change 36.238 IMS 14.1 IMS56FA INPUT EXCEEDED RECORD ABEND due to an VMACIMS invalid offset value of 2500 in DLRRXTOF/TPCEXTOF in a Dec 7, 2018 record that has only 572 bytes. Previous 14.1 56FA data Dec 23, 2018 didn't populate that offset to that new TPCX segment, but the 64-byte segment was present but not input. Since the segment contains nothing of value, rather than protecting for an invalid offset, the code for the TPCX segment is bypassed awaiting data from a site that wants the TPCX data. This is not worth a PMR at this time. -You can use this logic in your IMS job's SYSIN safely since this segment is at the end of the IMS LOG record: //SYSIN DD * MACRO STOPOVER MISSOVER % %INCLUDE SOURCLIB(TYPSIMST); but you should remove that MACRO statement when you update MXG. -New message if 56FA record IMSVERSN value is not equal to your site's _IMSVERS old-style macro value that you set in the job's SYSIN - see TYPEIMST/TYPSIMST JCL examples. But _IMSVERS is NOT USED in the 56FA records; it is used only in old IMS log records hex 07 08 0A 31 35 36 40 59. Dec 23, 2018: Variable TPCXLEN re-kept in IMS56FA always a missing value until IBM populates the segment.
VMACVMXA PROBABLE DATA LOSS MESSAGE exposed incorrect new SKIP Dec 6, 2018 logic for 6.4 that failed with 6.3 in MTRSYS, 1.04.. SKIP logic revised to protect old versions. Thanks to Giuseppe Giacomodonato, EPVTECH, ITALY.
TYPETMO2 there were no changes to the performance records. Nov 28, 2018
VMAC110 as is ALWAYS the case for CICS because they insert UTILEXCL fields in existing records. You can use the VMAC110 Nov 30, 2018 with this Change (36.235), if you have records, with no excluded fields and no optional segments, but if you have either (the existence of a tailored IMACEXCL member in your "USERID.SOURCLIB" tailoring proves you do), you will need to use UTILEXCL from this Change to create a new IMACEXCL that knows about these new variables that were INSERTED in Subtype 1 records: NJSAPPNM='NODE.NJS.APPPLICATION*NAME' SOCONMSG='FIRST*MESSAGE*PROCESSED' WBURIOTM='URI*WEBB OPEN*WAIT*TIME' WBURIOCN='URI*WEBB OPEN*WAIT*COUNT' WBURIRTM='URI*WEBB RECV*WAIT*TIME' WBURIRCN='URI*WEBB RECV*WAIT*COUNT' WBURISTM='URI*WEBB SEND*WAIT*TIME' WBIRISCN='URI*WEBB SEND*WAIT*COUNT' WBURVITM='URI*INVOKE*SERVICE*WAIT*TIME' WBURVICN='URI*IINVOKE*SERVICE*WAIT*COUNT' Yes, you need MXG 36.11 for CICS/TS 5.5 because fields were inserted into SMF 110 CICSTRAN records and using old MXG will have trashed values due to the misalignment, but MXG could run and only print error messages, which might be false positives, or could execute with no errors nor log messages, especially if you have a tailored IMACEXCL, but your CICSTRAN dataset will still be invalid.
VMACMVCI CICS/TS 5.5 COMPATIBLY added these variables to DMRDETL: Nov 30, 2018 T6E72XCT='72 EXTENSIONS' T6ESOCNM='FIRST*MSG*PROCESSED' T6EUROPT='URI*WEB OPEN*WAIT TIME' T6EUROPF='URI*WEB OPEN*WAIT FLAG' T6EUROPC='URI*WEB OPEN*WAIT COUNT' T6EURRPT='URI*WEB RECV*WAIT TIME' T6EURRPF='URI*WEB RECV*WAIT FLAG' T6EURRPC='URI*WEB RECV*WAIT COUNT' T6EURSPT='URI*WEB SEND*WAIT TIME' T6EURSPF='URI*WEB SEND*WAIT FLAG' T6EURSPC='URI*WEB SEND*WAIT COUNT' T6EWSIVT='URI*INVOKE*SERVICE**WAIT TIME' T6EWSIVF='URI*INVOKE*SERVICE**WAIT FLAG' T6EWSIVC='URI*INVOKE*SERVICE**WAIT COUNT' T6ENJAPN='NODE.JS*APPLICATION*NAME'
VMAC74 an IBM document that is now confirmed as misleading and Nov 27, 2018 the multiplier of 1024 is restored for these variables: R7491BPC R7491BPS R749DCTBYTR R749FPGBYTR R749FPGBYTS R749FPGCOBS R749FPGDCBS R749ICTBYTR R749PCIBYTR R749PCIBYTT R749PCIDMAR R749PCIDMAW Change 36.232 The SMF 99 Subtype 14 Processor Topology Report provided ANAL9914 by RAY back in Change 33.139 has a new version by Jim Nov 29, 2018 and the report program is now a %MACRO ANAL9914 so you can select CECTYPE and SYSTEM and REPORT with %ANAL9914(CECTYPE=Z14,SYSTEM=SYS1 SYS2,REPORT=JIM); REPORT=JIM uses PROC SGPANEL (SAS only) for prettiness! Thanks to Jim Horne, Lowes, USA. Thanks to Raymond J. Smith, OPTUM, USA. Change 36.231 Truncated POEX record caused INPUT STATEMENT EXCEEDED now VMACPOEX the short record is detected and reported on the log. the Nov 26, 2018 record had POEXNUM=35 File Segments expected, but the record only had room for 15. Thanks to Jack Hyde, UHC, USA.
VMAC119 alignment for these last 4 variables: ENCAPMODE/AUTHPROTO Nov 26, 2018 were incorrectly input as $CHAR2. S119SS_IPS_ENCAPMODE S119SS_IPS_AUTHPROTO S119SS_IPS_AUTHALG S119SS_IPS_ENCALG -MXG ZERT subtype 11 dataset TYP11911's GMTOFFTM create is valid with current records, but 2017 records had missing values that caused incorrect values, because SAETIME was often missing and on a different time zone. Thanks to Luis Mendoza, Black Knight, USA ====== CHANGES THRU 36.229 ARE IN MXG 36.10 DATED Nov 21, 2018========== Change 36.229 If you specified a dataset name in ASCII and did not UTILBLDP change the value of ECHO= to NO, a syntax error was Nov 21, 2018 generated when the created code was read back in to be displayed. If you used a filename (preferred) there was no error. Thanks to Giuseppe Giacomodonato, EPVTECH, ITALY. Change 36.228 The MXGWARN:MISSING TYPE70/TYPE70PR could be printed for VMAC7072 ICF-only LPARs, because MXGCIN was set to 'VM' in very VMXG70PR old logic, prior to SMF70CIN being provided by IBM. Nov 29, 2018 Now, MXGCIN is always set to SMF70CIN in VMAC7072, and the ICF LPARs with SMF70STN non-blank are skipped in VMXG70PR. Thanks to Scott Barry, SBBWorks Inc., USA.
IHDRCTLT and macro variable &MACCTTH created for instream use. VMACCTLT Missing value messages for dates eliminated. VMXGINIT
Nov 16, 2018 Thanks to Randal Schlueter, FirstData, USA. Change 36.226 Major revision to CICS RESPONSE TIME SLA MEMBER. ASUMCICR
Nov 20, 2018 Change 36.225 Documentation only. Examples clarified. ANALCAPD
Nov 16, 2018 Change 36.224 Documentation only. Examples added to select specific UTILBLDP subtypes of records to be added to PDB, Example 2 was. Nov 16, 2018 missing quotes.
ANALDB2R for PACKTYPE. Nov 19, 2018 -Variable QB3TDIO was misspelled in PMSTA02 causing another unitialized message. Change 36.222 Numerous STC formats were updated with new values for FORMATS HSC 7.3. Calculation of MSZ with CTP test, and new VMACSTC $MGSTCRR format added, so please update FORMATS. Nov 21, 2018 Thanks to Randy Hewitt, HPE, USA.
VMACVMXA was skipped when the number of IFL engines was changed Nov 15, 2018 from 5 to 3. MXG's attempt to detect wrap was invoked causing large values in LCUCACTM LCUCLPTM LCXCMTIT. The logic was revised to detect and delete the defective interval in VXSYTCUP. A PRM will be raised with IBM. Thanks to Giuseppe Giacomodonato, EPVTECH, ITALY.
UTILEXCL is included in RMISIOTM, and -SUM(WTDISPTM,WTSYIOTM); VMAC110 are overlapped with the sum of all of the other waits. Nov 20, 2018 -The WTOTIOTM created in UTILEXCL (when you have EXCLUDED fields) did not include these wait variables: DSAPTHTM DSCHMDTM DSMMSCTM DSTCBMTM FCVSWTTM GNQDELTM ISALWTTM JVMSUSTM MAXJTDTM MAXOTDTM MAXSTDTM MAXXTDTM PTPWAITM RQPWAITM RQRWAITM RRMSWATM RUNTRWTM SRVSYWTM SYNCDLTM TCALWTTM TDELWTTM TDILWTTM but only one, DSCHMDTM, is populated. -This is the new code for WTOTIOTM WTOTIOTM=SUM( DSPDIOTM,ENQDIOTM,GNQDELTM,WTICIOTM,WTLMIOTM, WTWCIOTM,RUNTRWTM,SRVSYWTM,RQRWAITM, RQPWAITM,SYNCDLTM,MAXOTDTM,MAXJTDTM,MAXSTDTM, MAXXTDTM,RRMSWATM,PTPWAITM,RMISIOTM,JVMSUSTM, DSTCBMTM,DSMMSCTM,WTDWIOTM,DSCHMDTM,FCVSWTTM, ISALWTTM,TCALWTTM,TDELWTTM,TDILWTTM,DSAPTHTM); IF (WTOTIOTM GT SUM(WTDISPTM,WTSYIOTM)) AND (WTOTIOTM-SUM(WTDISPTM,WTSYIOTM)) GT 0 THEN WTOTIOTM=WTOTIOTM-SUM(WTDISPTM,WTSYIOTM); Thanks to Jim Franklin, DXC Technology, USA.
Many and many other spelling errors have been corrected. Nov 13, 2018 Thanks to Michael R. Novak, USPS, USA. Change 36.218 Support for APARs OA52915 and OA52950 for SMF 21 counts VMAC21 to use the 4-byte 3590 counters instead of the original Nov 13, 2018 3-byte 3490 counters is already in place since the 4-byte counters were added. A minor non-impacting change to the test that sets the 3-byte counters with FFFFFFx to a missing value was made. Change 36.218 Reserved Change Number. Change 36.217 -Support for new subtype 6 BETA97 record, which creates FORMATS two new datasets: EXTYB976 DDDDDD Dataset Description EXTYB97S TYB976 BETA9706 BETA97 ST 06 IMACBE97 TYB97S BETA9706D BETA97 ST 06 DATASET VMACBE97 -Support for the Relocate data in subtype 22 BETA972REL VMXGINIT now decodes the expiration date fields to SAS dates and Nov 9, 2018 the B97RELTY values are decoded.. Thanks to Andreas Menne, Finanz Informatik, GERMANY.
ASMRMFV interval when a RMF III table id mismatch occurs as Nov 8, 2018 RMFV075E messages. -Extraneous ')' in message RMFV008I when processing a GDG for RMFBSAM. -NOSHOWZERO option not showing any table detail lines except for MXG00. Thanks to Randall Schleuter, First Data Corporation Change 36.215 Variable MSUSOFT, the interval Software MSU in TYPE72GO VMAC7072 for each Service/Reporting Class, can often be a missing Nov 7, 2018 value due to incorrect logic used to populate the CECSUSYSTEM and CECSUVALUE arrays from the preceding type 70 to provide SMF70CPA for the MSU calculation. If there is no preceding 70 from this system, "new" variable R723CPA is used so MSUSOFT will always be populated. And, if it is really the 4 Hour Average value you want, member ASUM4HRS will create that value for any variable in any dataset!
Thanks to Thomas Heitlinger, FIDUCIAGAD, GERMANY. Change 36.214 If OPTIONS OBS=0 has been set by SAS due to an error, VMXGSUM VMXGSUM produced unclear messages and zero obs created. Nov 7, 2018 Now, OBS=0 is detected and VMXGSUM shut down with error message that is clear. Thanks to Glen Bowman, Wakefern, USA.
VMAC110 removed. Nov 7, 2018
VMAC102 One record with two offsets (SC_OFF and SQL_OFF) that Nov 6, 2018 were larger than the record LENGTH caused the error. Offset lengths are now compared to LENGTH and if larger, log messages are printed. QW0376SC_OFF=48392 LENGTH=5210 IFCID376 SC ERROR QW0376SQL_OFF=27283 LENGTH=5210 IFCID 376 SQL ERROR These offsets, added in DB2 V11, are after VERSION but Version Length is ZERO in all records, and the first eight bytes of VERSION are nulls, so possibly the offsets are in those first eight bytes, so I've added variable VERSION1ST to display their value. A PMR will be raised with IBM DB2 Support. Thanks to Glen Bowman, Wakefern, USA. Change 36.211 -These AVG and STD values were not created in TYPE749: VMAC74 R7491DISAVG='AVG*INDIVIDUAL*DEFLATE*INPUT*BYTES' Oct 31, 2018 R7491DISSTD='SSQ*INDIVIDUAL*DEFLATE*INPUT*BYTES'' Nov 3, 2018 R7491DOSAVG='AVG*INDIVIDUAL*DEFLATE*OUTPUT*BYTES' Nov 28, 2018 R7491DOSSTD='SSQ*INDIVIDUAL*DEFLATE*OUTPUT*BYTES'' R7491IISAVG='AVG*INDIVIDUAL*INFLATE*INPUT*BYTES' R7491IISSTD='SSQ*INDIVIDUAL*INFLATE*INPUT*BYTES'' R7491IOSAVG='AVG*INDIVIDUAL*INFLATE*INPUT*BYTES' R7491IOSSTD='SSQ*INDIVIDUAL*INFLATE*INPUT*BYTES'' and these variables are now uncommented and kept /* SSQ VARIABLES NOW COMMENTED OUT SINCE STD EXIST R749FSQE R749FSQQ R7491DIS R7491DOS R7491IIS R7491IOS*/ This link https://www.ibm.com/support/knowledgecenter/en/ SSLTBW 2.1.0/com/ibm.zos.v2r1.erbb200/erbb200205.htm with PCIE Function Activity overview calculations show that the "MegaBytes" and "KiloBytes" are 1000 and not the 1024 per KiloByte that IBM uses in other 74 subtypes, so the 1024s in subtype 9 code are changed to 1000. -Nov 28: Change 36.233 reverted back to 1024 as IBM has confirmed the KB are 1024 and not 1000. -R749FPGBPRT is now the in-use buffer percent of memory. -R749BPC is relabeled AVERAGE*IN-USE*BUFFER*SIZE in bytes. -R749CIUTIL was incorrectly calculated. Thanks to Heimir Hauksson, Barclays, ENGLAND Anthony T. Sofia, IBM, USA.
VMAC119 they were created when not wanted when _N119 was used. Oct 31, 2018
TECHNOTE -APAR OA55602 RMF PP WLMGL TOTAL STORAGE INVALID after Oct 31, 2018 APAR OA52694; impacted only the TOTAL STORAGE and the STORAGE SHARED in RMF Report, 1000 times too large. -APAR OA53459 SMF TYPE 65 IS MISSING THE HLQ FOR THE GENERATION DATA SET WHEN DELETE GDG FORCE PURGE IS ISSUED on z/OS 2.2, not on 2.1. 2.3 not mentioned. -OA52950 3490 CHANNEL (PRE-COMPRESSION) BYTES AND DEVICE (POST-COMPRESSION) BYTES BEING REPORTED INCORRECTLY.in RMM (MXG TYPEDGR) and SMF 21 statistics, too high. -OA54992 SMSVSAM SMF TYPE 42 SUBTYPE 15 RECORDS MAY INCORRECLY RECORD ABOVE THE BAR BMF STATISTICS FOR DATA SETS. Occurs with a data set is assigned to a dataclass with above the bar usage set to yes, but the data set is not actually using above the bar buffers, which happens if RLSABOVETHEBARMAXPOOLSIZE is not set or if the change was made and the data set hasn't gone through a CLOSE to refresh the information. LOCAL FIX: Ensure data sets assigned to data class with above the bar usage are using above the bar buffers. -APAR OA56000 SMF RECORD LENGTH FOR SMF TYPE 98 HFTS RECORDS EXCEEDS THE MAX ALLOWABLE X'7FF4' (32,756). Subtype 1 are defined as x'8000' (32,768) and these records have caused CA SMF DIRECTOR product to fail and reject these records as invalid. -APAR PI96628 MQ V9: SMF116 RECORDS CONTAIN INVALID VALUES IN FIELD WTASMSTC, time spent in IXLLSTM call, and WTASSSTC for CICS and CHINIT connections accessing shared queues. -APAR PH99111 WMQ V9.0 SMF116 LATENCY RECORD FIELDS SET TO ZERO FOR SOME QUEUES. Fields MAXLATNT MINLATNT and TOTLATNT are zeroed after migrate to Version 9.0, for some queues. -APAR OA55594 SMF78 SUBTYPE 3 SUPERVPAV VALUES NOT REPORTED DUE TO R783DST BIT 7 FLAG NOT BEING SET R783XIND, causing I/O Queueing Activity Report to be missing the Alias Management Groups and/or the Logical Control Units section is missing the Alias Management Group field. Happens when the first device on a SuperPAV XPAV Control Units is offline. LOCAL FIX: Vary the first device on all SuperPAV XPAV Control Units online. -APAR OA55292 RMF FICON DIRECTOR STATISTICS REPORT HAS AN OCCASIONAL VERY LARGE VALUE FOR A PORT BANDWIDTH, READ OR WRITE. APAR IS OPEN.
VMAC7072 to the TYPE70 dataset: Oct 31, 2018 SMF70LACCR='LONGTERM*MSU*DFSMS*DATASET*ENCRYPTION' SMF70OS_PRTCT='OSPROTECT*VALUE' Change 36.207 Support for SMF 122 Subtype 2 zExplorer IBM Dependency EXTY122B Based Build, DBB, Release 1.0.3, which records usage of IMAC122A the DBB toolkit by z/OS users, creates new dataset. VMAC122A DDDDDD DATASET DESCRIPTION VMXGINIT Oct 30, 2018 TY122B TYPE122B ZEXPLORE DBB DEPEND BASED BUILD Change 36.206 When the input DD is tape or sequential, VMXGSUM and VMXGSUM VGETOBS cannot reasonably detect the presence or absence Oct 28, 2018 of datsets and have always tried to process them but if DSNFERR was on failed with a dataset not found error. But, if DSNFERR was set to NODSNFERR it then tried to process and if the dataset did not exist could generate UNINITIALIZED variable messages and a 0 OBS output. Now if VGETOBS detects a sequential LIBNAME and NODSNFERR MXGNOTEs are produced to let you know what may happen.
ASMRMFV Class, Report Class, and Resource Group Information Oct 26, 2018 Sections. -The SVP version level was raised from X'03' to X'04. -Either or both version levels may exist in z/OS 2.3 systems depending on the software currency of z/OS. -ASMRMFV has been changed to input these Sections using actual rather than documented lengths. -The SVP, ASI, ENC, and RCD RMF III tables are all affected when the SVP version level is X'04'. -MXG Changes 36.191 and 36.201 support new MXG variables for the new data in the VMACRMFV member. -ZEROCPU and ZERODVT parameters were not recognized due to a length specification error in the parameter table. Change 36.204 Plots and or TABULATE of MSU capacity GRAFMSU Reports. GRAFMSU For General purpose CPUs: Nov 16, 2018 CPC Capacity . LPAR Rolling 4 hr Avg Usage LPAR Hourly Usage LPAR Defined Capacity LPAR Capacity based on % SHARE LPAR Capacity based on # CPUs For IIP and IFA/CBP CPUs: CPC Capacity . LPAR Rolling 4hr Avg MSU (Estimated) LPAR Hourly Usage LPAR Capacity based on % SHARE LPAR Capacity based on # CPUs See doc in the member for examples of usage and see change 36.156 for examples of wrapping in ODS statements to create HTML or PDF output.
ANALRMFR number of CPUs*intervals so a CEC with 4 CPUs on a 15 Oct 26, 2018 minute interval showed up with 4 hour intervals. DURATM is now set to the INTERVAL= specified. Change 36.202 Used primarily inside of other macros to detect the VGETOBS existence and other information about SAS datasets. But, Oct 25, 2018 if you happened to pass incomplete text, like .ASUMCELP (where it should have been DDNAME.ASUMCELP) you could get strange errors and an ABEND, due to the period. Now, both sides are validated, or if only one is, WORK is used as the DDNAME with the DATASET to become WORK.ASUMCELP.
VMACRMFV (ASICNM ASISCWN AISCRN ASIRNM ASIWNM) in dataset ZRBASI Oct 26, 2018 could be misaligned. Introduced in Change 36.196. -Some WPS 4.0 iterations failed with ERROR: The format name '$ CPUPHYAD' invalid because WPS didn't like the blank in FMTNAME ' CPUPHYAD' Removing the blank didn't impact SAS but it healed WPS. Thanks to Rodger Foreman, Black Knight, USA Change 36.200 Logic rearranged to put error checking at the top, some VMXG2DTE additonal error checking added and abort changed to a Oct 24, 2018 soft landing with error messages. Thanks to Tom MacCabe, Dominion Energy, USA Change 36.199 -Beta 93 Subtype 51 variables now output in BETA51 dataset FORMATS that are also input in BE97 BETA9751D dataset: VMACBETA BETALCMD='LINE*COMMAND' Oct 30, 2018 BETARPOS='POINTER*TO*VALUE*AREA' -BETA 97 subtype 22 variable B97RELTY in dataset BETA9722 has new values decoded by updated FORMAT MGBETET. -New BETA 97 subtype 6 will create new datasets when test records are available. Thanks to Andreas Menne, Finanz Informatik, GERMANY.
VMACVMXA more in VXUSEACT and VXUSEINT datasets, which caused that Oct 22, 2018 merge to create multiple obs, which caused deaccumulated VMDTTIME/VMDVTIME to be very large for those intervals. Yüklə 28,67 Mb. Dostları ilə paylaş:
|