* copyright (C) 1984-2019 merrill consultants dallas texas usa
Yüklə 28,67 Mb.
|
|
that was corrected by HotFix E9BC81 SAS Note SN-035112, and this change text was revised. -However, it was also discovered that the dataset created by the PROC CONTENTS OUT= operand does NOT contain the TRANSCODE attribute; that OUT= dataset is a major input for MXG's QA testing of variables. While the printed output of PROC CONTENTS does show the TRANSCODE attribute, it is ONLY printed for datasets that have a variable with TRANSCODE=NO specified. Fortunately, the SASHELP.VCOLUMN dataset does contain the TRANSCODE attribute for every variable, so UTILXRF1 was revised to use that dataset to add TRANSCODE to the OUT= XREFDATA dataset used for QA analysis. SAS Development has indicated their intention to correct the omission in PROC CONTENTS OUT= in a future release. -UTILCVRT can be used if the dataset was already moved to the new platform, to "un-translate" the $HEX variables; it's comments were updated. -And, for WPS, the VMXGINIT default sets MXGVIEW=NO. -New $MGNOTRA is created in FORMATS by reading the FORMATS member to identify MXG-created formats that have HEX data values, and then it is used in the QA tests to ensure all of those variables do indeed have TRANSCODE=NO specified. -Because this change can be recognized by the existence of &MXGNOTRA in each updated member, I did NOT change the "LAST UPDATED:" text at the top of the 300+ members that created a total of 9831 $HEX-containing variables. I also did not add the TRANSCODE attribute to the DOCVER nor to the ADOCxxxx descriptions, as there's no room for a new attribute, and because a PROC CONTENTS will show it if you really need to confirm its value for a dataset.
VMACTMMQ subtype; MXG expected 18 reserved bytes that don't exist. Feb 18, 2009 Thanks to Paul Volpi, UHC, USA. Change 27.012 This enhancement to select DB2 reports by DATABASE set a ANALDB2R new MXG world record, with fifteen iterations needed to VFMT102 resolve all possible permutations discovered in testing! Feb 17, 2009 Selection by DATABASE uses the HEX values of OBID/DBID to Feb 21, 2009 select the desired observations from DB2 datasets, but if IFCID=105 and 107 were traced, they are used to map hex values to text names. If you did not trace 105/107s, or no 105/107 for the OBID/DBID being reported was found, the reports will print the HEX value instead of the name. New documentation in comments for the AUDIT= subparameter list the IFCIDs needed for full AUDIT report generation, as well as those that do contain the Database ID field. -If only PMACC02 was specified, and if both ASUMDB2A and DB2ACCCT datasets exist in the DDNAME/LIBREF pointed to by the PDB=xxx operand of %ANALDB2R, and if USEACCT=YES was specified, ASUMDB2A was always used when DB2ACCT should have been used. Thanks to Scott Swindling, Nordstrom, USA, who tested each iteration! Change 27.011 Support for CTG V7.2 (INCOMPATIBLE, new CTG_RECID=7 made VMAC111 INPUT STATEMENT EXCEEDED RECORD LENGTH error). That new Feb 19, 2009 subtype is not documented in SC34-6961-00, but protection for any new subtype was added, with a NOTE that new data exists in the future. -CTG_RECID=3 for GD (Gateway Daemon) added new variables: CTGDFSRV='DEFAULT*SERVER*FOR*GATEWAY' CTGHOSNM='HOSTNAME*OF GATEWAY*COMPUTER' CTGIHAEX='CICS*REQUEST*EXIT*CALLS' CTGISYNC='SYNCONRETURN*FAILS' CTGLHAEX='CICS*REQUEST*EXIT*CALLS' CTGLSYNC='SYNCONRETURN*FAILS' Thanks to Renato Guerra, SGS Banco Popolare di Verona, ITALY. Thanks to Davide Marone, SGS Banco Popolare di Verona, ITALY. Change 27.010 Variable SMF70CIX is now output in PDB.TYPE70PR dataset, VMAC7072 as it is the CPU Pool Number. While it has been INPUT Feb 16, 2009 for years, it was only used as an index to find out the engine type of each LCPUADDR, to store into SMF70CIN; it was Martin's SlideShow that educated me to the fact that it is also the CPU pool number, and hence worth keeping. http://www.slideshare.net/MartinPacker/much-ado-about-cpu Thanks to Martin Packer, IBM, EUROPE. Change 27.009 -SYSLOG MSGID IEF233D was captured by ASMTAPEE in ML-43, VMACTMNT last summer, but VMACTMNT was not updated to decode it, Feb 13, 2009 causing "UNEXPECTED" messages to be printed on the log. That MSGID is now output by VMACTMNT in TYPESYMT dataset. Support for IEF233D in ASUMTAPE was already in place. -SYSLOG messages that do NOT contain DSNAME are created if your site has NOT specified COM='MONITOR DSNAME' in your SYS1.PARMLIB(COMMNDxx). This causes either blank DSNAMEs in PDB.TYPESYMT and PDB.ASUMTAPE, or, it caused DSNAME to have "MEDIAn", because the SYSLOG MSGID IEC705I was incorrectly parsed when DSNAME was not populated. The location of the DSNAME in IEC705I is variable, so MXG's parsing was revised to only INPUT the SYSLDSN if the text ends in 'MEDIAx'. A new warning message is printed if the last field is not MEDIAx, and a warning if the sss field is blank, so I can validate my parsing of that instance. Thanks to Yves Cinq-Mars, IBM Global Services, CANADA.
VMACVMXA MXG unconditionally INPUT variable PFXCPUTY in 4 subtypes Feb 13, 2009 but the other 8 subtypes input that field conditionally. These four subtypes are now also conditionally INPUT (0.1 SYTSYP, 3.2 STORSP, 3.20 STOSXP, 5.3 PRCPRP) based on the length of the segment. Thanks to Tom Draeger, Aurora, USA.
FORMATS Change 26.323, but several fields were not documented by VMACENQM IBM in the DSECT; the ENQFLAG1/ENQFLAG2 are now decoded, Feb 11, 2009 and the description of ATTR1 bit settings were deduced. Change 27.006 Several logrec character variables were length $200 due VMACEREP to that pre-SAS-V8-length-limit, or were input $CHAR200 Feb 14, 2009 but then reduced because of shorter $HEX format length. These revisions were made: New SDWA $400. $HEX800. Removed: SDWA1, SDWA1A. New SDWARC1 $456. $HEX912. Removed: SDWARC1A,1B New SDWARC2 $16. $HEX32. New SDWARC3 $32. $HEX64. New SDWARC4 $360. $HEX720. New SDWAVRA $255 $HEX510. Removed: SDWAVRA1. New LXC2SNI $228. $HEX456. Removed: LXC2SNI1,2 Old SLHIRB increased to $HEX128. Thanks to Ken W. Kasten, Embarq, USA.
BUIL3001 of SMF Record IDs that were read by BUILDPDB (by SYSTEM). BUILD001 Now, it prints ID+SUBTYPE with the new IDANDSUB variable, BUILDPD3 e.g., IDANDSUB=110.002 for the ID=110 SUBTYPE=2 records. BUILDPDB Additionally, the output of the PROC FREQ is now created MONTHASC in the new PDB.SMFRECNT dataset. MONTHBL3 -The WEEKly and MONTHly PDB build programs were updated to MONTHBLD create the SMFRECNT there as well, sorted BY ZDATE. MONTHBLS -All of the WEEK/MONTH programs are enhanced so that they MONTHDSK will NOT fail when an expected dataset does NOT exist. MONTHWEK This means you can implement a new BUILDPDB on any day of WEEKBL3 the week (previously, the only safe day was to implement WEEKBL3D on the first day of your week, so all seven day-of-week WEEKBL3T PDBs had the new dataset for the new WEEKly job.), and on WEEKBLD any day of the month for the MONTHly job. Now, by using VMACID SAS OPTIONS NODSNFERR NOVNFERR instead of just NODSNFERR, Feb 12, 2009 the DATASET-NOT-FOUND is protected by NODSNFERR, and the BY-VARIABLE-NOT-FOUND is protected by NOVNFERR. Well, almost: A WARNING: BY VARIABLE NOT FOUND will be printed on the SAS log, but that warning does not set _ERROR_, so there is NO return/condition code created, which is exactly what is desired for this condition. However, it is inconsistent for a WARNING message to not set a condition code, but since we do NOT want that here, SAS Technical Support has said the "WARNING" text will be changed to "NOTE" in the future. Thanks to Diane Farias, IBM Global Services, CANADA
EXTY113 record, documented in SA23-2260,SA23-2261 and added by FORMATS APAR OA25755 (and possibly other APARs), creates TYPE113 IMAC113 dataset with six Basic, six Program State, sixteen Crypto TYPE113 and twenty-four Extended counters of Level One, Two and TYPS113 Three Data and Instruction Cache activity counts for each VMAC113 CPU engine. The counts are accumulated, so the TYPS113 VMXGINIT member, which does the de-accumulation in the _S113 SORT Feb 8, 2009 macro, must be used, or with BUILDPDB, you must add the _S113 product sort macro in your EXPDBOUT member. Thanks to IBM z/OS Developer Support Program for providing SMF data!!
EXTY114 Creates new TYPE114 dataset from subtype 1 record. FORMATS This was originally completed on Jan 11, 2009, as Change IMAC114 26.298, but I accidentally re-used/overwrote that change TYPE114 text. The was no change to the code delivered in 26.26. TYPS114
VMAC114 VMXGINIT
Jan 11, 2009 Feb 9, 2009 Thanks to Siegfried Trantes, Gothaer Systems GmbH, GERMANY.
READDB2 -READDB2 in MXG 26.26: might not create all datasets that Feb 11, 2009 you requested, and could impact ANALDB2R(PDB=SMF,...) as Feb 16, 2009 the READDB2 member is invoked by ANALDB2R to read SMF. Feb 17, 2009 Last minute changes for IFCID=255 and DB2STAT4 were made Feb 20, 2009 but validation focused only on that correction. -If VMXGTIME was being used, when READDB2 tried to resolve a local macro variable it got a bad value placed in a local macro variable by VMXGTIME. This generated a WARNING: MACRO VARIABLE NOT RESOLVED. The macro variable name was changed in READDB2 to prevent the conflict. -The Second READDB2 in the re-dated MXG 26.26 could fail if the DB2= operand was used to select DB2 SubSystems. Fortunately (for me!), this caused an actual 180 ERROR condition before any data was read, so it was NOT like the first (insidious) error that you could overlook until you tried to read a non-created DB2 dataset, thus this error did not require another refresh! But the revision revealed an unrelated/unreported inconsistency with the documentation, if you had multiple selection criteria, like %READDB2(DB2=DB2X DB2Y, PLAN=PLANA PLANB). Those criteria were ORed, so all PLANs from DB2X or DB2Y were selected, plus all subsystems with PLANA or PLANB were also selected. The logic is now revised to AND all of the selection criteria, so now only PLANA or PLANB from only either DB2X or DB2Y would be selected. And, a cosmetic change was made so that the selection criteria are printed on the SAS log, so you can see what was requested. -A pair of additional errors were also corrected. Only DB2ACCTP was created from ACCOUNT if IFCIDS=ALL was used, and calling READDB2 from ANALDB2R could cause this error: DATA=ZZDB2PST; % ERROR: OLD-STYLE MACRO NAME % MUST CONTAIN ONLY LETTERS, DIGITS, AND UNDERSCORES. -May 20: Writing SMFOUT on PC's &RECFM corrected. -May 24: The T102S225, DB2STAT4, IFCID=225 building was significantly redesigned in Change 27.097. Thanks to Raff Rushton, Kaiser Foundation Hospitals, USA. Thanks to Scott Chapman, American Electric Power, USA. Thanks to Tony Curry, BMC, USA.
VMXGRMFI But ONLY if HiperDispatch is active, plus negative values Feb 11, 2009 in variables PCTOVHTD CPUACTTM CPUOVHTM MSUINTRV MSUPERHR MSU4HRAV and PCTOFHDW. A recalculation of CPUACTTM was not tested with RMF data with HiperDispatch active (i.e., when SMF70PAT/CPUPATTM were GT zero). That recalculation was removed by this change. Thanks to Chuck Hopf, Bank of America, USA. LASTCHANGE: Version 27. =========================member=CHANGE26================================ /* COPYRIGHT (C) 1984-2009 MERRILL CONSULTANTS DALLAS TEXAS USA */
First MXG Version 26.26 was dated Feb 3, 2009, thru Change 26.325 MXG Version 26.26 is the 2009 "Annual Version". MXG Newsletter FIFTY-THREE is dated Feb 3, 2009 MXG Version 26.12 was dated Jan 20, 2009, thru Change 26.308 MXG Version 26.11 was dated Jan 5, 2009, thru Change 26.296 MXG Version 26.10 was dated Dec 1, 2008, thru Change 26.271 MXG Version 26.09 was dated Oct 20, 2008, thru Change 26.240 MXG Version 26.08 was dated Sep 12, 2008, thru Change 26.209 MXG Newsletter FIFTY-TWO was dated Aug 24, 2008 MXG Version 26.07 was dated Aug 24, 2008, thru Change 26.197 MXG Version 26.06 was dated Aug 6, 2008, thru Change 26.176 Third MXG Version 26.06 was dated Aug 5, 2008, thru Change 26.175 Second MXG Version 26.06 was dated Aug 4, 2008, thru Change 26.173 First MXG Version 26.06 was dated Aug 1, 2008, thru Change 26.172 MXG Version 26.05 was dated Jun 18, 2008, thru Change 26.140 MXG Version 26.04 was dated Jun 4, 2008, thru Change 26.120 MXG Version 26.03 was dated May 11, 2008, thru Change 26.095 First MXG Version 26.03 was dated May 8, 2008, thru Change 26.093 MXG Version 26.02 was dated Apr 22, 2008, thru Change 26.075 MXG Version 26.01 was dated Mar 11, 2008, thru Change 26.037 First MXG 26.01 was dated Mar 10, 2008, thru Change 26.036 MXG Version 25.25 was dated Jan 28, 2008, thru Change 25.309
http://www.mxg.com/ship_current_version Your download instructions will be sent via return email.
II. SAS Version requirement information. III. WPS Version requirement information. IV. MXG Version Required for Hardware, Operating System Release, etc. V. Incompatibilities and Installation of MXG 26.26. VI. Online Documentation of MXG Software. VII. Changes Log
and is updated with new notes frequently. All Newsletters are online at http://www.mxg.com in the "Newsletters" frame.
Member CHANGESS contains all changes that have ever been made to MXG. All MXG changes are also online at http://www.mxg.com, in "Changes".
I. MXG Version 26.26, dated Feb 3, 2009, the 2009 Annual Version. Major enhancements added in MXG 26.26, dated Feb 3, 2009 TYPEDB2 26.311 DB2STATS4-IFCID=225 in DB2 V9 correction/additions. TYPE23 26.312 Support for APAR OA27163, new interval variables. TYPEOMMQ 26.319 Support for Omegamon XE MQ Export File TYPEENQM 26.323 Support for IBM's ENQ/DEQ Monitor flat file. ASMTAPEE 26.317 Enhanced SYSLOG message capture, MSGIDs read from DD. TYPETMDB 26.313 New subtypes for TMON for DB2 V4 and V4.1. READDB2 26.311 TEXT EXPRESSION LENGTH (65545) EXCEEDS corrected. VMXGINIT 26.310 The MXGWORK= argument of VMXGINIT is removed. TYPE70 26.308 TYPE70 PARTNICF/IFA/IFL/ZIP variables added. TYPEACF2 26.324 Enhancement for ACF2 support to add IHDRACF2 exit. UTILCVRT 26.322 UTILCVRT utility for z/OS to ASCII conversion revised
ANALZPCR 26.297 Execution errors in mismatched Tags corrected. TYPE72GO 26.299 MXG 26.10-26.11. PERFINDX missing for R723TYPE=2. RMFINTRV 26.303 Capture Ratios for zAAPs,zIIPs added to PDB.RMFINTRV TYPE102 26.298 DB2 IFCID=22 INPUT STATEMENT EXCEEDED error fixed. FORMATS 26.304 Internal format $MGUTILD for DOCVER created. Major enhancements added in MXG 26.11, dated Jan 5, 2009 Many 26.289 QA Cleanup, variable LENGTHs changed, zdate added. VMXGPRAL 26.293 Print all datasets with variable name + label heading TYPERMFV 26.287 Support for RMF III CPUG3 z/OS 1.9 (INCOMPAT). TYPE92 26.277 Support for APAR OA24208 new subtype 15 for ID=92. TYPEOPCN 26.280 Support for Open Connect user SMF record. ANALZPCR 26.283 zPCR failed with z/OS V1R9. SELECT=CECTIME supported. IMACICMD 26.284 BMC Optional CMRDB2 segment increased to 256 INCOMPAT TYPENMON 26.279 DISKSERV,DISKWAIT,MEMPAGESxxx supported, some fixes. TYPE78 26.288 TYPE78CU variables for Aliases could be wrong. ASUMUOW 26.282 APPLID could be blank in PDB.ASUMUOW. BUILDPDB 26.281 IFAUNITS,IFEUNITS added to PDB.STEPS and PDB.JOBS. ANALDBJO 26.278 Example analysis JOins DB2ACCT + DB2ACCTP, expensive. TYPE72GO 26.276 Negative one value for CPUUNITS due to resolution. BLDSMPDB 26.275 New WEK2KEEP=,MTH2KEEP= controls for keeping PDBs. TYPEDB2 26.274 Some DB2 V9-only QWACxxxx vars were INPUT with V8. TYPEXAM 26.272 Variable SIZE in HSTMEM incorrectly INPUT, is RB4. Major enhancements added in MXG 26.10, dated Dec 1, 2008 ANALZPCR 26.264 Support for IBM zPCR model input from MXG PDB data. PDB TYPE70,TYPE70PR,TYPE74 are read to create "External Study" files for input to IBM's capacity modeling tool; by default, MXG selects the RMF interval from each system with peak CPU usage to be modeled. This should make your use of IBM's excellent zPCR tool even easier!
TYPE70PR 26.243 Support for OA21140 RMF HiperDispatch enhancements. Because of these changes MXG 26.10 is now required for HiperDispatch
TYPE120 26.262 Support for WebSphere Version 7, new subtype 9 data. TYPEMVCI 26.254 Support for MAINVIEW for CICS 6.1 CMRDETL (INCOMPAT). TYPEBVIR 26.250 Support for eight clusters in BVIR33 dataset. TYPEHSM 26.249 Sorting HSM ABARS datasets caused NOT SORTED errors. TYPE113 26.247 SMF 113 data records needed to finish support. TYPE70 26.269 CPUWAIxx/MVSWAIxx for CP Engines 33-63 were missing. TYEPRMFV 26.246 RMF III ASIPHTxx SRB CPU times wrong by x1000. TYPETPMX 26.245 ERROR VARNAME=$JXSLMJ_ in Thruput Mgr SMF corrected. GRAFWRKX 26.244 MIPS was not calculated for WORKLOAD=0, uncaptured. ANALDB2R 26.256 %ANALDB2R(PDB=SMF); failed with error. VMXGOPTR 26.242 Internal utility enhanced for multiple options. TYPEVMXA 26.241 LINUXKRNL '02'x caused BROKEN CONTROL RECORD error. TYPE112 26.257 TYPE112 now reads both V550 & V560 subtype 203 data. TYPEOMCI 26.257 TYPEOMCI supports subtype 200,201,203, but only V550. VMXGINIT 26.252 Forward Slash in a unix libref for WORK supported. TYPECTLG 26.255 Enhancements to processing Catalog records. Many 26.252 %QUPCASE(xxx) vs %UPCASE for forward slash protect. Many 26.259 QA Stream revised to eliminate return code & warnings WPS 26.258 WPS 2.3.4 now required for MXG, ARRAY(256,512) error. Major enhancements added in MXG 26.09, dated Oct 20, 2008 TYPE70 26.236 HiperDispatch CPUPATTM, PCTMVSBY can be wrong TYPE70. TYPE7072 26.222 Large CPUIFATM IFAUNITS when op varied CP on/offline. ASUMMIPS 26.216 ZIPUSED MSU was incorrect, ZIP/ZAP metrics fixed. TYPENMON 26.224 NMON variables without decimal point may be wrong. TYPESVC 26.221 Support for IBM DS8000 2107 SAN Disk SVCPerfStats. TYPENTSM 26.213 Support for new data in NTDS and ASP.NET App objects. TYPETMDB 26.210 Support for ASG/Landmark DB2 Monitor V4.1 raw data. TYPETNG 26.223 Support for NSM VMWARE ESX 2.5.5 new objects. FORMATS 26.231 MEMLIMIT '00000FFFFFFFF000'x value is NOLIMIT. READDB2 26.233 Dataset DB2STAT4 and T102S225 created for IFCID=225. ASUMSTGP 26.228 Example to report DASD storage by Storage Group. TYPERMFV 26.218 RMF III ASIRNM,ASIRDE (reporting class) names blank. TYPENDM 26.215 NDM-CDI subtype 'UC' is now output in NDMAE. TYPE1415 26.214 Invalid extended segment protection enhanced. UPRINDOC 26.238 Utility to PROC PRINT the LABEL and VARIABLE NAME.
TYPEDB2 26.201 Support for DB2 V9.1 SMF 100,101 (COMPAT MXG 25.25+) TYPE1415 26.199 INVALID SMF1415 RECORD, even with Change 25.228, fix. TYPEBVIR 26.198 All BVIR32 Pool 00-31 are now Pool 01-32 variables. TYPETPMX 26.207 Support for Thruput Manager Subtype 7, new fields. IMACICMR 26.206 Optional BMC CMRDATA increased in CICS/TS 3.2. WEEKBLDT 26.205 SYSNAME incorrectly added to BY List for TYPE892. TYPESHDW 26.204 Support for new subtypes, fields Shadow USER SMF. BUILDPDB 26.208 Variables SMF30MLS, MEMLIMIT now kept in PDB.STEPS.
MXG code that protected an earlier IBM error in the number of triplets caused z/OS 1.10 TYPE72GO to have zero observations, so MXG 26.07 is REQUIRED to support z/OS 1.10 records. Sorry! MXGSAS92 26.191 New JCL proc for SAS V9.2 with new z/OS DSNAMES. VMXGINIT 26.189 SAS V9.2 Hot Fix F9BA07 eliminates new WARNINGs MXG Version 26.03+ provided circumvention for new WARNING messages that set condition Code 4 with SAS V9.2, but SAS Hot Fix F9BA07 now eliminates the need for that MXG circumvention. TYPE42 26.187 Support for APAR OA25205 adds SMF 42 subtypes, data. TYPEINSY 26.182 Support for MACRO4 INSYNC SMF user record. ASMIMSL6 26.190 Support for IMS Log record 0A (CPI-CI Drive PGM). TYPEIMS7 26.190 Support for IMS Log record 0A (CPI-CI Drive PGM). ASUMCEC 26.188 HiperDispatch subtracts SMF70PAT from SMF70ONT ASUMDB2P 26.183 Revised summary/trending of DB2ACCTP example. TYPERMFV 26.178 RMF III z/OS 1.9 changed length of ASI segment.
UNDUPSMF 26.152 Utility removes duplicate SMF records, output is VBS. RMFINTRV 26.165 New RMFWKLRV: RMFINTRV Workload-only dataset created. TYPEQACS 26.166 Support for AS/400 Version 6.1.0 (COMPATIBLE). TYPETPF 26.163 Support for TPF PUT22 changes, and corrections. Yüklə 28,67 Mb. Dostları ilə paylaş:
|