* copyright (C) 1984-2019 merrill consultants dallas texas usa
Yüklə 28,67 Mb.
|
|
filters are effective only if the SPG table is selected. New Keyword Aliases ------------- ----------------------------------------- SPGSTORGROUP= SPGSTORGRP=, STORGRUP=, SPGGRP=, SGPSNM=, SPGSG=
SPGVOLSER= SPGVOLI=, SPGVOL=, SPGSER=, SPGV= SPGAND None SPGOR None
are present is: 1) SPGSTORGROUP= (or any alias for SPGSTORGROUP=) 2) SPGVOLSER= (or any alias for SPGVOLSER=) Selection results from repeats of the same SPG filter keyword (or any of its aliases) are always logically ORed. -TUTORIAL: Ranges of the form keyword=first:last may be used with any of the above keywords except SPGAND and SPGOR. The colon character ':' is required for a paired range specification. All entries GE the first value and LE the last value are selected for output to the RMFBSAM file.
collating sequence or an error is flagged in message RMFV056E.
simply as keyword=first and in this case the colon ':' is omitted.
except SPGAND and SPGOR and include one or more Wild Card characters to match the respective SPG data field.
characters as follows: Wild Card Matches ---- ------------------------------------------------- * 0 or more characters % 1 Non-blank character + 1 Numeric character (0-9) _ 1 Alphabetic character or _ (a-z, A-Z, _) . 1 National character (@, #, $) ! 1 Special character (not a-z, A-Z, 0-9, @, #, $) ? A blank string if used by itself ? 1 Blank character (X'40') if used with any other characters Ranges may not be wild carded. If wild carded the range value becomes a pattern instead and is processed as such. See Section 25 "Ranges and Patterns" in the ADOCRMFV member or ASMRMFV source code for more details on usage of ranges and patterns.
Group name entries from 1-8 characters. Storage Group names are defined for System Managed Storage (SMS) and are 1-8 alphabetic characters including A-Z, 0-9, $, @, #, *, or %. The first character must be alphabetic (A-Z). Both ranges and patterns with wild cards may be specified. Up to 64 ranges and 64 patterns are supported. The default is SPGSTORGROUP=ALL. -In all of the following examples assume the following Storage Group names and Volume Serials are defined to SMS in this highly simplified configuration and that RMF Monitor III is actively measuring them: Storage Volume Group Serials -------- --------------------------- PRODPOOL PRD001 PRD002 PRODAA PRODBB TESTPOOL TST001 TST002 TESTAA TESTBB WORKPOOL WRK001 WRK002 WORKAA WORKBB -Examples for SPGSTORGROUP= : SPGSTORGROUP=PRODPOOL:TESTPOOL is a range that selects only Storage Groups with a name GE 'PRODPOOL' and LE 'TESTPOOL'. All the volume serials in these pools will be selected. No volumes in WORKPOOL are selected. SPGSG=P* is a pattern that selects only Storage Group names that begin with 'P'. Only volumes from the PRODPOOL are selected. Note use of the keyword alias SPGSG= for coding convenience. STORGRUP=A* is a pattern that selects only Storage Groups with a name that begins with an 'A'. No Storage Groups will be selected in this example. Note use of an alias.
with a name that ends with 'L'. All Storage Groups will be selected in this example. This is the default if no filter keywords are coded. Note use of an alias. See Section 25 "Ranges and Patterns" in the ADOCRMFV member or ASMRMFV source code for more details on usage of ranges and patterns.
Volume Serial number. Both ranges and patterns with wild card characters may be specified. Up to 64 ranges and 64 patterns are supported. The default is SPGVOLSER=ALL. Any valid 1-6 character Volume Serial with or without pattern characters may be specified. Per JCL syntax a Volume Serial Number is 1 through 6 alphanumeric, national ($,#,@), or special characters. NOTE: Since just about any keyboard character is valid in a Volume Serial please take extra care when coding to avoid unintended results in the MXG PDB data set ZRBSPG.
efficient than filtering by Storage Group because every volume in every Storage Group must be matched against the ranges and patterns provided. There are usually many more volumes than Storage Groups. Use of SPGSTORGROUP= instead is recommended if feasible. NOTE: If one or more SPGVOLSER= filters is coded and a Storage Group has zero volume data entries the Storage Group is filtered as this is considered a mismatch.
However, ASMRMFV will search every Storage Group until it is found.
'TST000' and LE 'TST999'. In this example volumes TST001 and TST002 will be selected. Note use of an alias.
'WORK' followed by up to 2 more characters. In this example volumes WORKAA and WORKBB will be selected. Note use of an alias. SPGVOL=*+++ selects all volume serials ending in 3 digits. In this example volumes PRD001, PRD002, TST001, TST002, WRK001, and WRK002 are all selected. Note use of an alias. SPGVOLI=P*A selects all volume serials starting with 'P' that have a final character 'A' with up to 4 intervening characters. In this example only volume PRODAA is selected. Note use of an alias. -SPGAND (default) indicates that selection results from the two different SPG filter keywords (and any of their respective aliases) are logically ANDed.
different SPG filter keywords (and their respective aliases) are logically ORed. SPGOR must be coded if desired.
-With SPGAND (default) in effect: SPGSTORGROUP=WORKPOOL SPGVOLSER=WRK* only selects volumes in the WORKPOOL Storage Group AND that have a volume serial number that starts with 'WRK'. SPGSTORGROUP= may appear redundant since only one pool has WRK* volumes, but it keeps ASMRMFV from matching many Volume Serials in other Storage Groups if only SPGVOLSER= were present. SPGSTORGROUP=WORKPOOL SPGVOLSER=WRK* SPGVOLSER=WORK* only selects volumes in the WORKPOOL Storage Group AND that have a volume serial number that starts with either 'WRK' or 'WORK'. In this example omitting SPGVOLSER= would have produced the same result and is redundant. SPGVOLSER= in this case adds unnecessary overhead.
or TESTPOOL Storage Groups. Note the use of aliases. SPGSTORGROUP=WORKPOOL SPGVOLSER=PRD* selects NO volumes in this example using SPGAND because the PRD* volumes are in the PRODPOOL AND not in the WORKPOOL. There will be zero observations in the ZRBSPG data set in the result MXG PDB. SPGAND (default) logical ANDing provides more restrictive SPG entry filtering than SPGOR. -With SPGOR in effect: SPGSTORGROUP=WORKPOOL SPGVOLSER=WRK* only selects volumes that are in the WORKPOOL Storage Group OR that have a volume serial number that starts with 'WRK'. In this example omitting SPGVOLSER= would produce the same result and is redundant. SPGVOLSER= adds unnecessary overhead in this case with SPGOR because ASMRMFV will search all other Storage Groups trying to match that volume serial.
Group OR that have a volume serial number that starts with either 'WRK' or 'WORK'.
SPGVOLSER= keywords are redundant in this example. SPGVOLSER= adds unnecessary overhead in this case with SPGOR because ASMRMFV will search all other Storage Groups trying to match those volume serials.
or TESTPOOL Storage Groups. Note the use of aliases. This example produces the same result as with SPGAND because two keywords (or any of their aliases) for the same selection are always logically ORed.
any volumes that start with 'PRD' in the PRODPOOL Storage Group.
is in effect. With SPGAND no volumes are selected. ASMRMFV will search all Storage Groups for a match with the SPGVOLSER= value. The logical OR with SPGOR results in less restrictive filtering because any of the 2 conditions in this example results in data selection of a SPG volume data entry. -Support for a new multi-table selection filter VOLSER= (aliases VOLUME=, VOLI=, SER=, VOL=) to allow selection by Volume Serial from both the RMF Monitor III DVT and SPG tables with one keyword.
Volume Serial parameter twice when the same volume from both tables is of interest. Both the DVT and SPG tables must be selected for this multi-table selection keyword to function completely. Otherwise only entries from the one selected table are filtered. Note that most RMF III tables do not contain common character data fields, but in this case the DVT and SPG do. -Example of VOLSER= : VOLSER=ABC* is equivalent to coding DVTVOLSER=ABC* SPGVOLSER=ABC* -ASMRMFV now supports keywords up to 14 characters in length up from 12. -Support +13 / -13 hour offset or +780 / -780 minutes GMTOFFSET= value for locations near International Date Line using Daylight Saving Time such as New Zealand. -MXG00 record version is now X'05' from X'04' and includes new range and pattern table statistics for SPG filtering. New ASMxxxxx variables added to ASM00 dataset. -Add new items to Section 2 "Terminology" : Enclave, Report Class, Resource Group, Service Class, Storage Group, and Workload. -Former Section 27 "Summary" is now Section 29. -Update documentation for SPGSTORGROUP=, SPGVOLSER=, and GMTOFFSET= support: Section 5 "Input Data Selection Parameters" Section 8 "Error Handling Parameters" Section 9 "JCL and SYSIN Parameter Usage" Section 12 "Messages" Section 13 "Filtered Records" Section 15 "Program and IBM Limitations" Section 25 "Ranges and Patterns" Section 29 "Summary" -New documentation Section 27 "GMT Offset Support" better explains use of the GMTOFFSET= keyword added by MXG Change 34.133 . -New documentation Section 28 "Collection of DASD Usage with RMF Monitor III" details the requirements and setup of DASD usage measurement in the SPG table. -New documentation Section 30 "Bibliography" lists IBM manuals and numbers for the Resource Measurement Facility (RMF) for z/OS 2.2 back to z/OS 1.10. -Count information messages and include in final RMFV999I message with documentation update. Change 34.190 COSMETIC. INVALID ARGUMENT TO FUNCTION DATEJUL(1900000) VMACTPMX and a hex dump was printed if the DUE OUT time values are Aug 10, 2016 all zeros. Had no impact on the output TPMX datasets. Thanks to Scott Wiig, USBank, USA. Change 34.189 MXG 34.05 ONLY. INPUT STATEMENT EXCEEDED if more than 3 VMAC119 ADDL HOME ADDR fields exist; Change 34.168 update didn't Aug 10, 2016 protect more than 3. Since this site had 4, now there are 4 fields kept, and the more than 4 is protected. Thanks to Scott Wiig, USBank, USA.
VMAC7072 statement did not specify $1 to set the array to CHAR; Aug 10, 2016 if the user dropped variable CAI0, WPS failed because it couldn't identify the type. While this is an internal WPS issue that will be resolved, adding $1 to the ARRAY statement removes the exposure. Change 34.187 Support for SMF 124 I/O Supervisor Information (z/OS 2.2) FORMATS DDDDDD DATASET DESCRIPTION EXTY1241 TY1241 TYPE1241 I/O Supervisor Information IMAC124 The SM124RETTIME field is an invalid value, with the four TYPE124 byte packed decimal date value first and then the time, TYPS124 reversed from the standard SMFSTAMP format with time and VMAC124 then the date. MXG code detects/corrects either format. VMXGINIT
Aug 15, 2016 Thanks to Scott Barry, SBBWorks Inc., USA Change 34.186 Support for Mainview for IP RTIN='34'x, TAC9I220 dataset. FORMATS DDDDDD DATASET DESCRIPTION IMACMVIP VMIP34 TAC9I220 PIUTRACE VMACMVIP (Note that IMACMVIP controls which RTIN values are read.) Aug 15, 2016
VMACRMFV SPGFREEP='PERCENT*OF VOLUME*FREE' Aug 5, 2016 SPGUSEDP='PERCENT*OF VOLUME*USED'
VMAC110 -New variables added to CICSTRAN DCN=416 DRL=3588. Aug 7, 2016 but no dictionary records were produced at startup so the new CICSTRAN fields are unknown, and it is also unknown if new fields were inserted or appended. Current MXG code falls thru and uses 5.3 INPUT code for 5.4 records but that is WRONG if there were INSERTS. -Type 110 Subtype 2 STID 108 statistics CICTCPIP new vars: SORTCPIPSMAXPERSIST='MAXIMUM*PERSISTENT*CONNECTIONS' SORTCPIPSNONPERSIST='NON*PERSISTENT*CONNECTIONS' SORTOTALCONNS ='TOTAL*CONNECTIONS' SORNONPATMAXPERSIST='MADE*NON-PERSISTENT*MAXPERSIST*REACHED' SORNONPATTASKLIMIT ='NEW NON*CONN MADE*TASK LIMIT*EXCEEDED' SORDISCATTASKLIMIT ='DISCONNECTS*TASK LIMIT*EXCEEDED' SORDISCATMAXUSES ='DISCONNECTS*MAX USERS*EXCEEDED' SORCURRBACKLOG ='CURRENT*BACKLOG*Q-DEPTH' SORCONNSDROPPED ='CONNECTIONS*DROPPED' SORCONNLASTDROPPED ='DATETIME*LAST*CONNECTION*DROPPED' SORCURRMAXBACKLOG ='BACKLOG*CURRENTLY*IN USE' SORREQUESTS ='REQUESTS*PROCESSED' Change 34.182 MIMPRID=2 Record sample count variable MIMCFNBR=0, but VMACMIM numerous fields are divided by Sample Count, so their Aug 3, 2016 divide caused DIVIDE BY ZERO messages and missing values in those variables. Temporarily, those divides are now inside a IF MIMCFNGR GT 0 DO group so the numbers are not set missing, while vendor contact is pursued. Change 34.181 -Defective BMC CMF type 74 subtype 4 records with SMF744ML VMAC74 zero (no SCM segments) but with non-zero R744RISC (SCM Aug 3, 2016 segment expected) caused R744Mxxx SCM variables to be Aug 11, 2016 INPUT when there was no data. Now, SCM segment is INPUT only when both R744RISC and SMF744MN are non-zero. This has been reported as a defect to the vendor; one record has the triplet populated with no segments and the next record has the data but the triplet count is zero. -SCM Variables R744MRBT and R744MWBT are now converted to bytes from KB and formatted MGBYTES. like other SCM byte containing variables. -Missing values notes for R749PCIxxx variables eliminated by wrapping conditional INPUTs with DO Group and moving the multiply inside the appropriate DO Group. Thanks to Giuseppe Giacomodonato, EPVTECH, ITALY. Thanks to Paul Volpi, UHC, USA. Change 34.180 -Omegamon XE ATF datetimestamps are now on the LOCAL zone VMACATF and variable ATFTU2L, the GMT Offset, is now kept in all Jul 29, 2016 ATF datasets. Sep 5, 2016 -To correlate the MXG variables with the CSV files from Oct 14, 2016 IMS Performance Analyzer, these variables are now kept in all ATF datasets ATFCORI/ATFCORP/ATFCORT to match the IBM AESCORID/AESCORPPST/AESCORTIME and variable ATFCREKEY is concat of AESCREID/AESREOASN/AESRECOMN/ATFCRE. -Summary variables ATFSUD2N/D2E/D2C from IMSATFD2 into the IMSATFA0 dataset incorrectly included ITEM CODE 9, the TOTALs, which were double accounted. -Oct 14. ATFSTART corrected to local time zone. Thanks to Robert Gilbert, BNPParibasFortis, FRANCE. Change 34.179 -Updated CHECKSTN, failed only in JCLTES92 in ASUM113 QA CHECKSTN test, because PDB.TYPE70PR was presumed to exist; now the VMXGINIT program verifies it exists, or prints note on the log if Jul 28, 2016 it can't be executed, and why not. -The CHECKSTN program was added to the ASUM113 program to report if duplicates exist in your TYPE70PR data, but you you can suppress its execution with in your //SYSIN with %LET MXGSTNCK=NO; Thanks to Anon, Anon, USA.
VMAC80A SISCCNO, SISDMPID, SISCOMPY, SISECVT, SISFIRST, SISLAST. Jul 27, 2016 creates new variables TOKMCSRGUID/TOKMSISMIDDL/TOKMSISCCNO/TOKMEMPID/ TOKMSISCOMPY/TOKMSISECVT/TOKMSISFIRST/TOKMSISLAST in dataset TYPE80TK. Thanks to Michael Oujesky, DTCC, USA.
VMACTMO2 LENMONI=0. causing "SHORT RECORD" warnings on the log, Jul 27, 2016 but these records with TMMDREC='DD' are not the records wanted, so they are now deleted prior to the length text. Thanks to Rodger Foreman, Transunion, USA.
VMAC80A caused invalid segment skipped message because MXG only Jul 27, 2016 expected 11. Thanks to Harald Seifert, HUK-COBURG, GERMANY. Change 34.175 zVM 6.3.16.1 added 4 bytes to PRCPUP segment but MXG did VMACVMXA not protect correctly, causing PROBABLY DATA LOSS ERROR Jul 26,2016 on the SAS log. SKIP logic is corrected and the dataset VXPRCPUP is now output for each segment; previously, only the last segment was output with the Park/Unpark metrics. Thanks to Joseph J. Faska, DTCC, USA Change 34.174 New NOTYPE= parameter lets you specify a list of SMF IDs VMXGGETM to not be copied. They will be counted in the input but Jul 26,2016 not in the output counts. Column percentages were added to the output report. VMXGGETM creates an SMF output file selecting N records of each SMF ID.
VMACIAM Change 34.008A in MXG 34.01 added support for V9.2 with Jul 25, 2016 segment lengths of 292 and 264 bytes for IAMIAINL and IAMIASTL, but V9.0 has shorter 204/148 segment lengths that are now detected and protected. Thanks to Paul Naddeo, Fiserv, USA. Thanks to Bernie Ethridge, Fiserv, USA.
IHDRMVIP the MACMVIH macro variable, which also needed to be VMXGINIT defined in the %GLOBAL statement in VMXGINIT. Jul 24, 2016 Change 34.170 Support for WebSphere Liberty Batch SMF 120 subtype 12. EXT12012 Creates new dataset: VMAC120 dddddd dataset description VMXGINIT T12012 TYP12012 WAS LIBERTY BATCH Jul 23, 2016
VMACVMXA must also create VXMTRPRP to populate the PFXCPT array. Jul 19, 2016 The _Tdddrrr macros read VMINPUT and the _Xdddrrr macros read the MWINPUT file to create each VXdddrrr dataset. Thanks to Scott Barry, SBBWorks Inc., USA
VMAC119 adds these first three instances to dataset TYP11906: Jul 16, 2016 IFADDLIN1='IFADDINTFNAME*1' IFADDLIH1='IFADDINTFHOME*1' IFADDLIN2='IFADDINTFNAME*2' IFADDLIH2='IFADDINTFHOME*2' IFADDLIN3='IFADDINTFNAME*3' IFADDLIH3='IFADDINTFHOME*3' Thanks to Wolfgang Kueller, S-Itsolutions, AUSTRIA
ASUM113 ONLY NEEDED IF YOU HAVE LPARs WITH THE SAME SYSTEM NAME. CHECKSTN VMAC7072 -New SOLUTION FOR THE ASUM113 PROBLEM: VMXGINIT The text below shows SMF70STN can not be used to match Jul 16, 2016 TYPE1131 data with TYPE70PR data, but an alternative to Jul 25, 2016 identify which TYPE70PR obs belong to this SMF record: Jul 28, 2016 IF PARTISHN(SMF70PTN)=LPARNUM(SMF70LPN) in TYPE70PR, that LPAR is the LPAR of this SMF 70 record, which is then selected to be merged with TYPE1131s. BUT YOU MUST have BOTH 70s and 113s, and ONLY from one LPAR, for ASUM113 to create valid PDB.ASUM1131 dataset. -NO SOLUTION FOR THE ASUM70PR PROBLEM: If you have duplicate SMF70STN values with CHECKSTN, there is NO SOLUTION to use ASUM70PR to combine the multiple LPAR's TYPE70PR data; those duplicates cause PCTCPUBY to be incorrect (over 100%) with other metrics also wrong. YOU MUST PROCESS EACH DUPLICATED LPAR's SMF/RMF data into a SEPARATE PDB FOR EACH OF THE LPARs. -The new CHECKSTN program can be run to read PDB.TYPE70PR to produce a report ONLY if duplicate SMF70STN values for different LPARNAMEs are found in your data: // EXEC MXGSAS94 //PDB DD DSN=YOUR.TYPE70PR.PDB,DISP=SHR //SYSIN DD * %INCLUDE SOURCLIB(CHECKSTN); -Original Change text, prior to Jul 25: IN GENERAL, MXG CAN NOT HANDLE MULTIPLE SYSTEM NAMES THAT ARE FOR UNIQUE SYSTEMS TO BE COMBINED. YOU MUST CREATE SEPARATE PDB DATA LIBRARIES FOR EACH SYSTEM AND THEY CAN NOT BE MERGED/COMBINED. The SMF70STN (LPAR's SYSTEM name) is needed in ASUM113 as it is the only mapping from the z/OS SMF SYSTEM name to that systems LPARNAME, and must be used there so the TYPE70PR LPAR utilization variables can be added to the PDB.ASUM1131 dataset. But RMF data with the same SMF70STN for different LPARNAMEs has occurred and that corrupts the PDB.ASUM1131 dataset with incorrect values and creating multiple LPARNAMEs when there was only one system's SMF 113 records. The original solution required you to tell MXG the real SYSTEM name of those LPARNAMEs that are duplicated, using Yüklə 28,67 Mb. Dostları ilə paylaş:
|