* copyright (C) 1984-2019 merrill consultants dallas texas usa
Yüklə 28,67 Mb.
|
|
Mar 16, 2013 IBM field name in an INPUT statement, but its end-comment was missing, causing MXG to be mis-aligned, so CICSMDSA was "trashed" - easy to see since DSANAME was blank - and printed WARNING CICS STID=0 STILEN=0 message on the log. Thanks to Jim Horne, Lowe's, USA.
IMAC6ESS ESSOPTNS variable in TYPE6 dataset if you edit IMAC6ESS VMAC6 to cause it to not be DROPed. See IMAC6ESS comments. Mar 15, 2013 Thanks to Jessie Gonzales, California State Controller's Office, USA.
FORMATS but SAS didn't detect the syntax error in member FORMATS, Mar 15, 2013 until FORMATS was accidentally run without a //LIBRARY, and only then did SAS flag the invalid syntax. Change 31.045 ID=90 records subtype 6 and 7 "EVENTIME" is not the time VMAC90 of the SWITCH/HALT but is the IPLTIME of this SYSTEM. So VMAC90A variable IPLTIME is now kept in the TYPE9006 dataset and Mar 14, 2013 both the archaic TYPE90 and recommended TYPE90A members store EVENTIME into IPLTIME, but I kept EVENTIME since it might already be in use in your reports. Thanks to Perry Lim, Union Bank, USA.
VMACNMON were wrong because the INDEX test was underspecified and Mar 13, 2013 satisfied multiple times, so the last match was used. Mar 14, 2013 -Mar 14: BBBP029 still wrong, new "POWER SAVINGS MODE" Mar 15, 2013 for MODE. Realized all groups of IF INDEX tests could be replaced with ELSE IF INDEX tests for ordered testing and to improve performance. Added BBBP147-BBBP149 variables. -Mar 15: BBBP029 still wrong, new "MEMORY MODE" used, so new BBBP150-BBBP158 are created for all new entries and the test for 'MODE ' is ELSE IF after the above tests. Thanks to Steve Dyck, CDS, CANADA. Thanks to G. Delvecchio, Canadian Depository for Securities, CANADA Change 31.043 -The example in comments in UTILEXCL that reads CICS SMF IMACEXCL with UTILEXCL to create your tailored IMACEXCL, and then UCICSCNT uses TYPE110 to re-read the CICS SMF data using that new UTILEXCL IMACEXCL code, works on z/OS because the example's JCL Mar 12, 2013 //IMACEXCL DD DSN=MXG.USERID.USERID(IMACEXCL),DISP=SHR writes the IMACEXCL code to your "USERID" tailoring PDS that is in your //SOURCLIB concatenation, so the %INCLUDE of IMACEXCL in TYPE110 finds and uses the new IMACEXCL. However, on ASCII, the FILENAME IMACEXCL must be written to a file named imacexcl.sas (suffixed with .sas) and on unix, in lower case, and, similarly, the file must be written to a "tailoring" directory that is in the FILENAME SOURCLIB concatenation on ASCII.
with no suffix, UTILEXCL wrote to that filename, but when TYPE110 was included, there were the same errors and no observations, because the new IMACEXCL was not %INCLUDEd, because it did not have the ASCII-required .sas suffix.
In MXG 30.30 (Change 30.283) UTILEXCL was enhanced for diagnostics: when an IMACEXCL (created by new UTILEXCL) is %INCLUDEd, a new MXGNOTE is printed on the SAS log: 177043 %INCLUDE SOURCLIB(TYPE110); MXGNOTE: SITE IMACEXCL CREATED AT 11MAR2013:21:07:59.09 207431 RUN; It was the absence of that MXGNOTE when I ran the user's program to diagnose why no observations were created that reminded me of the ASCII syntax requirement. Adding .sas to the filename caused the new IMACEXCL to be error-free. Examples in UTILEXCL were updated with ASCII syntax. -The UCICSCNT program that reads SMF to count CICS record types (transaction, statistics, dictionary), from each APPLID and VERSION, is enhanced to count dictionary by "triplet" values MCTSSDCN and MCTSSDRL, which matches the existing transaction count details. This can be very useful when MXG doesn't find the records your CICS guru told you were enabled, by showing him/her what records are ACTUALLY being created. So useful, that I added a %INCLUDE SOURCLIB(UCICSCNT); statement to each example in UTILEXCL that reads SMF data, since it would be useful to support@mxg.com if there is a perceived UTILEXCL problem. Thanks to Giuseppe Giacomodonato, EPVTECH, ITALY. Change 31.042 Change 30.275 added support for a USER FORMAT library but MXGNAMES the last line added in MXGNAMES was typo'd; it should be VMXGCNFG LET MXGFORMU=YOUR.HLQ.MXG.USER.FORMATS; /*OPTIONAL*/ Jan 22, 2013 instead of MXGFORMT, and MXGNAMES wasn't listed. Change 31.041 -Support for SHADOW USER SMF subtype 21 record creates new EXSHDW21 DDDDDD DATASET DESCRIPTION IMACSHDW
VMACSHDW SHDW21 SHADOW21 SHADOW MAPS VMXGINIT -Mar 19: added support for subtype 4 record. Mar 11, 2013 -Mar 23: corrected _BSHDW04 and _BSHDW21. Mar 23, 2013 Thanks to Stuard Wildey, MorganStanley, ENGLAND.
READDB2 to minimize the //WORK library disk space requirement Mar 9, 2013 NOTE: THE FILE WORK.DB2STATB (MEMTYPE=DATA) WAS NOT FOUND BUT APPEARS ON A DELETE STATEMENT. WARNING: NO MATCHING MEMBERS IN DIRECTORY. are no longer printed. -Also cosmetic: Change 30.257 didn't note that the use of each _S102nnn sort macro creates 23 lines of messages on the SAS log, or 9,320 more SYSOUT lines for IFCIDS=ALL. Thanks to Jerry Massey, Compuware, USA. Change 31.039 -The IBM TS7700 BVIR data does not contain a SYSTEM name; VMACBVIR variable SYSTEM is now created and populated using Mar 8, 2013 // EXEC MXGSAS, // OPTIONS='SYSPARM="ASYS"' with one, then eight, then five single quotes on z/OS, or you can set SYSPARM in your SYSIN stream with: OPTIONS SYSPARM='ASYS'; -Discovered subtype 32x caused INPUT STATEMENT EXCEEDED with BVIRLEN=8256; tests added in Change 31.016 for the LENGTH 8256 should have tested BVIRLEN instead. Thanks to Scott Barry, SBBWorks, Inc., USA.
VMACRACF RACF0120 dataset. Mar 8, 2013 Thanks to Bill Arrowsmith, Euroclear, BELGIUM. Change 31.037 Documentation enhancements for installing MXG on unix, AUTOEXEU AIX, Linux, etc. INSTALL Mar 7, 2013 Thanks to Al Sherkow, I/S Management Strategies, Ltd. Change 31.036 BETA93 Version 4.2.0 and 4.3.0 had invalid data values in VMACBETA dataset BETA1 and 4.2.0 had invalid large NRACCTFL value Mar 8, 2012 that also caused "DATA ERROR. CODE IN IMACACCT ....". Variables INPUT after the accounting fields were invalid. Thanks to Rudolf Sauer, T-Systems, GERMANY.
VMAC74 input dataset rather than the correct _WTY74ID macro Mar 7, 2013 name so if you modified the destination of the dataset using the WTY74ID macro variable, the sort failed. Thanks to Mike Georte, Northern Territory Government, AUSTRALIA.
ANALDB2R to the PROC SORT NODUP removal algorithm, but ANALDB2R Mar 7, 2013 wasn't tested with PMAUD02, which was not reverted: ERROR: The keyword parameter SYSTEM DBNAME OBNAME was not defined with the macro.
ANALCOMP name in titles. Mar 7, 2013
TRND71
Mar 7, 2013 Min Max Avg Frames in Pool SMF71L1M SMF71L1X SMF71L1A Frames Not Used SMF71L2M SMF71L2X SMF71L2A Frames In Use SMF71L3M SMF71L3X SMF71L3A Objects Alloc SMF71LOM SMF71LOX SMF71LOA Thanks to Wayne Bell, UniGroup, USA. Change 31.031 Cosmetic. Label for SMF64RLM is corrected from ESDS to VMAC64 SMF64RLM='CA-S*RECLAIMED*IN KSDS*SINCE CLOSE/EOV'. Mar 6, 2013 Thanks to Karl Lasecki, Chemical Abstracts Service, USA. Change 31.030 Support for 12 new NTSMF objects and MANY new variables EXNTD064 added for existing objects. Find "ADDED BY CHANGE 31.030" EXNTD065 in VMACNTSM to identify the new variables. There are now EXNTD066 513 MXG datasets created from Windows NTSMF data. EXNTD067 EXNTD068 -New Objects now supported: EXNTD069 EXNTD070 DDDDDD Dataset/Object EXNTD071 NTD064 HYPER-V_VIRTUAL_SWITCH_PROCESSOR EXNTD072 NTD065 MSRS_2011_WINDOWS_SERVICE EXNTD073 NTD066 MSSQL_BATCH_RESP_STATISTICS EXNTD074 NTD067 MSSQL_FILETABLE EXNTSQSV NTD068 MSSQL_MEMORY_BROKER_CLERKS IMACNTSM NTD069 MSSQL_MEMORY_NODE VMACNTSM NTD070 MSSQL_QUERY_EXECUTION VMXGINIT NTD071 NUMA_NODE_MEMORY Feb 26, 2013 NTD072 NETWORK_ADAPTER NTD073 PHYSICAL_NETWORK_INTERFACE_CARD NTD074 SMB_SERVER_SHARES
NTACSR ACTVSRVR NT ACTIVE SERVER PAGES NTASPN ASPNET ASP.NET NTNETI NETWINTR NT NETWORK INTERFACE NTPPAC PPNETACC PER PROCESSOR NETWORK ACTIVITY NTPRIN PROCINFO PROCESSOR INFORMATION NTQLAM MSQACCES NT MSSQL:ACCESS METHODS NTQLBM MSQBUFMG NT MSSQL:BUFFER MANAGER NTQLBN MSQBUFND MSSQL:BUFFER NODE NTQLDA MSQDATAB NT MSSQL:DATABASES NTQLMM MSQMEMMG NT MSSQL:MEMORY MANAGER NTSERV SERVER NT SERVER NTSQSV SQLSSISV SQLSERVER:SSIS SERVICE 11.0 NTD011 CLUSTER_RESOURCE_CONTROL_MANAGER NTD015 HYPER_V_DYNAMIC_MEMORY_BALANCER NTD016 HYPER_V_DYNAMIC_MEMORY_VM NTD018 HYPER_V_HYPERVISOR_LOGICAL_PROCE NTD019 HYPER_V_HYPERVISOR_PARTITION NTD020 HYPER_V_HYPERVISOR_ROOT_PARTITIO NTD021 HYPER_V_HYPERVISOR_ROOT_VIRTUAL NTD030 HYPER_V_VM_VID_PARTITION NTD035 HYPER_V_VIRTUAL_NETWORK_ADAPTER NTD037 HYPER_V_VIRTUAL_SWITCH NTD038 HYPER_V_VIRTUAL_SWITCH_PORT -Corrections diagnosed and corrected by Phil: QLGS - NRDATA=24 supported QLAM - MSEXCART removed from INPUT, not kept, not there ASPA - ASPARWTB removed from NRDATA=85 INPUT, not there -Additional records changed: QLDA - NRDATA=26 supported. Thanks to Phil Henninge, Demand Technology, USA.
VMAC6156 to protect for INVALID SECOND ARGUMENT when the SMF 66 Feb 26, 2013 record for a RENAME of DSN=X.Y to DSN=A.B.G0002V00 had an entry name too short to be tested for a GOOVO. Thanks to Rudolf Sauer, T-Systems, GERMANY.
ANALCOMP by WPS. Code was added to the macros to make the use of VGETLIBS views conditional depending on the value of MXGVIEW that VGETWKLD is set by VMXGINIT based on the version of SAS or WPS VMXGFIND being run. VMXGSRCH -In the case of VGETWKLD, it was not correctly gathering Feb 24, 2013 the number of periods for TSO workloads because there are differences in the names of the response time vars between the older TSO workload and any new workloads you may have created with RMFINTRV. For these newer workloads, at the time VMXGRMFI was written the limit on variable names was 8 characters so a name like DB2P1RESP would not have been possible so RSP was used rather than RESP to keep the variable names down to 8 bytes. That limit no longer applies but changing old variable names would be a bad thing to do to you so VGETWKLD now checks correctly for the old TSO variable names.
VMAC102 trace datasets. Feb 22, 2013 Thanks to Tony Anderson, Blue Cross Blue Shield of Alabama, USA. Change 31.026 SAS 9.1.4 SP4 ONLY: Note 33063 states that SASHELP.VEXTFL PREINIT dataset only has the FIRST external file reference, if an VMXGINIT AUTOCALL is issued, which happens by design in VMXGINIT, Feb 22, 2013 so any subsequent reference to SASHELP.VEXTFL has only that first DDNAME/LIBREF. There is NO FIX for 9.1.3: "The only known solution is to process the dictionary table SASHELP.VEXTFL prior to any system AUTOCALL." You can use DATA WORK.VEXTFL; SET SASHELP.VEXTFL; in PREINIT to save the VEXTFL table into //WORK.
Feb 20, 2013 Change 31.024 Format $MGSMFID had "DEFINE" rather than "DEVICE" in the FORMATS description of the ID=11 SMF record. Feb 20, 2013 Thanks to Mike Mayne, HHSYS, USA. Change 31.023 Support for Software Diversified Services VFTP product EXVFTP01 user SMF record creates these four new datasets: EXVFTP02 EXVFTP03 DDDDDD MXG MXG EXVFTP04 DATASET DATASET DATASET RECORD IMACVFTP SUFFIX NAME LABEL SUBTYPE TYPEVFTP TYPSVFTP VFTP01 VFTPST01 VFTP01:SESSION AND TRANSFER 01 VMACVFTP VFTP02 VFTPST02 VFTP02:LOGIN FAILURE 02 VMXGINIT VFTP03 VFTPST03 VFTP03:SERVER REJECTED 03 Feb 18, 2013 VFTP04 VFTPST04 VFTP04:SUMMARY ACTIVITY 04
but there is no GMT Offset in the record, and the delta between the End time and the SMF time is as much as 2100 seconds which I believe is in error, and that prevents me from using that delta to heuristically create the offset. This is being discussed with the vendor, but you can use %LET MXGGMTOFF=-6*3600; %INCLUDE SOURCLIB(....); to circumvent and set your GMT Offset (remembering to change each Spring Forward and Fall Backward). Change 31.022 -Support for Software Diversified Services VIP Product SMF EXVIPAPL USER record creates these datasets: EXVIPDLR EXVIPDLW SDS DDDDDD MXG MXG EXVIPEEC DATASET DATASET DATASET DATASET RECORD EXVIPEEX NAME SUFFIX NAME LABEL SUBTYPE EXVIPFRM EXVIPHPR DLC1R VIPDLR VIPDLR VIPDLR:DLC READ QUEUE 222 EXVIPHPX DLC1W VIPDLW VIPDLW VIPDLW:DLC WRITE QUEUE 222 EXVIPIFC HPR1 VIPHPR VIPHPR VIPHPR:HPR 228 EXVIPLUG HPR1 VIPHPX VIPHPX VIPHPR:HPR EXTENSION 228 EXVIPOSA RTM1 VIPRTM VIPRTM VIPRTM:TCP TAPM OR TN3270 229,231 EXVIPOSX FRM1 VIPFRM VIPFRM VIPFRM:FRAG MONITOR 230 EXVIPRTM LUG1 VIPLUG VIPLUG VIPLUG:LU GROUP 232 EXVIPSTK OSA1 VIPOSA VIPOSA VIPOSA:OSA 239 FORMATS OSA1 VIPOSX VIPOSX VIPOSA:OSA EXTENSION 239 IMACVIP APL1 VIPAPL VIPAPL VIPAPL:APPLICATION 242 TYPEVIP IFC1 VIPIFC VIPIFC VIPIFC:INTERFACE 243 TYPSVIP STK1 VIPSTK VIPSTK VIPSTK:STACK 244 VMACVIP EEC1 VIPEEC VIPEEC VIPEEC:ENTERPRISE EXTENDER 245 VMXGINIT EEC1 VIPEEX VIPEEX VIPEEC:EE ROUTES 245 Feb 22, 2013 -The variable names in these datasets are the same as the names in the SDS-provided example SAS programs and the MXG dataset names are similar so the SDS-provided SAS reports should be easily adapted to use the MXG datasets. The MXG code is structured so all records can be processed in one pass (the SDS example program processed one subtype per program) and the MXG "dddddd" tokens are created to permit standard MXG tailoring of datasets and contents. -The SDS-example variable names are long with underscores embedded; the MXG label replaced the underscore with the normal asterisk character so you will almost always want to use PROC PRINT SPLIT='*' with VIP datasets to get nice column headings and alignment. Change 31.021 -RMF III Enhancements and Notes ASMRMFV -NOTE: Since MXG V30.03 it is required that SYS1.MODGEN JCLASM3 also appear in the SYSLIB DD concatenation for the JCLRMFV ASMRMFV assemble step as follows (use JCLASM3 example): JCLCRMFV VMACRMFV //SYSLIB DD DSN=SYS1.MACLIB,DISP=SHR Feb 15, 2013 // DD DSN=SYS1.MODGEN,DISP=SHR Feb 26, 2013 Mar 10, 2013 ASMRMFV displays execution environment data that needs the macros in SYS1.MODGEN to map certain z/OS control blocks. Omission of SYS1.MODGEN in the assembly step will result in ASMA prefixed error messages and the assembly will fail. -ASMRMFV and VMACRMFV must be at the same level; do not expect the old VMACRMFV to read RMFBSAM files created by the new ASMRMFV, and vice versa. -When the ASI table is selected in ASMRMFV the entries are now blocked up to use as much as possible of a 32K logical record. The number of ASI records output was reduced by up to 96% during testing. -The example members JCLRMFV and JCLCRMFV for building an RMF III PDB have been improved by adding a new example of using the SIZE parameter to examine RMF III VSAM data set disk space and index usage. Additional information about use of ASMRMFV parameter aliases is also now included. -VMACRMFV is upgraded to support blocked ASI entry input. -VMACRMFV now displays a message about ASMRMFV level and execution environment that created the RMFBSAM file. -MISSING VALUES message on SASLOG for CPUDTIME variable is removed in VMACRMFV. Input logic for CPUDTIME is improved. -NOTE: The use of archival RMFBSAM files as input to TYPSRMFV or TYPERMFV is NOT recommended as many missing values and other undesirable results can occur in the PDB since VMACRMFV is expecting data in the current input format. The RMFBSAM file is intended only as temporary work file for the PDB build process and should not be retained for future use. A better alternative is to instead restore the archived RMF III VSAM file and read it with the current ASM/VMAC RMFV program pair. -NOTE: If intending to include the UWD table datasets in the RMF III PDB %INCLUDE SOURCLIB(TYPSRMFV) not TYPERMFV should be used in PDB build step. There are naturally occurring duplicates in the UWD table that need to be removed by the sorting process. -When the ENC table is selected in ASMRMFV the entries are now blocked up to use as much as possible of a 32K logical output record. -VMACRMFV is upgraded to support blocked ENC entry input. -ASMRMFV now inputs all 6 ENC table entry triplets instead of only the first one. Each offset/length/count triplet points to a separate group of ENC table enclave entries. Users including the ENC table in their RMF III PDB can expect 5 times as many observations in ZRBENC dataset; documentation on the number of ENC triplets was unclear until actual data provided the actual documentation that that there were 6 triplets and not just 1. -Note: The total number of ENC table records output will now be at least 6 times the number of tables input since each of the 6 triplets causes a new record to be output. -REQUIREMENT: In order to receive the performance benefits from ASI and ENC entry blocking, the current ASMRMFV utility program with this MXG change must be installed. See MXG SOURCLIB member JCLASM3 for sample JCL for the assembly and link-edit install steps. -Tutorial: This item has been mentioned before, but with the ASI and ENC blocking and increase in observations, you may need to use the EXdddddd dataset exit to filter which observations are created, and your logic in those exits should always use this code structure: IF something THEN DO; OUTPUT _Wdddddd; END;
because when "something" is true, they stop the read of this current record, skipping any un-read segments from being tested for "something". This consideration applies in particular to the following RMF III tables that have blocked input data to a PDB build and to their respective output exits:
------------- ----------- ASI EXZRBASI CPD EXZRBCPD CSR EXZRBCSR DVT EXZRBDVT ENC EXZRBENC ENT EXZRBENT OPD EXZRBOPD RED EXZRBRED SHD EXZRBSHD SPG EXZRBSPG UWD EXZRBUWD Thanks to Perry Lim, Union Bank, USA Thanks to Betty Wong, Bank of America, USA
VMACTMMQ new variables to the TMMQQA dataset. Feb 15, 2013 Thanks to Paul Volpi, UHC, USA. Thanks to Homayoun Riazi, UHC, USA.
ANALCAPD to your PDB library. By default they were temporary and Feb 13, 2013 are deleted by ANALCAPD. This change deletes the delete but this text will be updated when the PDBOUT= option is. Thanks to Ralph Belamy, ???, ???
ASMIDMPA DDDDDD DATASET DESCRIPTION EXIDMPAC IDMPAC PACEIDMS PACE IDMS USER SMF RECORD IMACIDMP The SMF record is created in an IDMS Exit by ASM code TYPEIDMP contributed by Blake Leggett, PACE Applied Technology, TYPSIDMP their product KOMAND IDMS Charging System, KOMAND/IDMS. VMACIDMP See the ASMIDMPx members documented in Change 31.068. VMXGINIT However, neither PACE nor MERRILL will guarantee ASMIDMP Feb 20, 2013 exit code will always work with future IDMS changes; this Mar 9, 2013 now works fine, but your IDMS staff must be prepared to ASMIDMPJ investigate and repair any exit problems in the future. ASMIDMPS If your guru can fix the exit, I can update the SAS code! ASMIDMPU Thanks to Blake Leggett, PACE Applied Technology, Inc., USA. Thanks to Trevor Rowe, Bell Alliant, CANADA.
ANALCOMP member used when TRNDRMFI is invoked; VGETWKLD extracts VGETLIBS your workload names by reading DICTIONARY.TABLES to get VGETWKLD your variable names from PDB.RMFINTRV by using VMXGFIND PROC SQL; VMXGSRCH CREATE VIEW VGETWKLD AS SELECT NAME,LABEL Feb 12, 2013 FROM DICTIONARY.COLUMNS. . . . and then reading dataset VGETWKLD with this logic POINT=LENGTH(NAME)-2; IF SUBSTR(NAME,POINT,3)='ZIP'; but that logic failed with WPS, perhaps due to the VIEW, as views are not yet supported in WPS, or perhaps because WPS created a different LENGTH value, but in either case POINT1=INDEX(NAME,'FRTM'); IF POINT1; now circumvents the error, even with "VIEW" in the code. The FRTM suffix test was used in place of ZIP as not all sites create a xxxxZIP workload variable. And just to be safe, the "VIEW" text was removed from VGETWKLD and the other four members. Change 31.016 -Support for BVIR Version 2.0, 2.0a, and 2.1. VMACBVIR -BVIR30 adds two new variables: Yüklə 28,67 Mb. Dostları ilə paylaş:
|