* copyright (C) 1984-2019 merrill consultants dallas texas usa
Yüklə 28,67 Mb.
|
|
the TYPE6156 datasets, from which you can find the time when the z/OS dataset was deleted, but those records will have only the job name of the DB2 DBM1 address space. By adding DB2ACCT you can narrow in on who did it to the DB2 table in that time period, with QXDRPDB or QXDRPTA or QXDRPIC GT 0. Thanks to Chuck Hopf, Independent Consultant, USA.
ASUMSMFI of PDB.SMFINTRV to create PDB.ASUMSMFI. Dec 10, 2010 Thanks to Chuck Hopf, Independent Consultant, USA. Change 28.293 -Support for OPENSSH SMF 119 subtypes 96-98 creates new VMAC119 dataset ddddd description VMACSMF TYP11996 T11996 OpenSSH Server Transfer Complete Dec 9, 2010 TYP11997 T11997 OpenSSH Client Transfer Complete TYP11998 T11998 OpenSSH Login Failure -Technically, these new subtypes are INVALID SMF records because BIT 1 in SMFxFLG, which is the IBM indicator that the record contains subtypes, is not ON, causing VMACSMF to see these as SUBTYPE=0. Now, VMACSMF forces the input of SUBTYPE for ALL SMF ID=119 records. Thanks to John McKown, Health Markets, USA.
VMXGSRCH of the Modify date, so the search results are printed in Dec 8, 2010 the same order they appear on the SAS log. The MODATE=NO default prints the datasets alphabetically, as before. MODATE=YES was used to debug the multi-step TYPETMS5 code by selecting a VOLSER to follow, especially since the variable name that contains "volser" is different in different temporary WORK datasets. Additional parameters were also added to allow you to limit which datasets and which variables to be searched and printed, if you don't want to see all of them. New parameters: MODATE=NO Change to YES to sort on modify datetime DATASET= A list of full or partial datasets to be searched for the string VARS= A list of full or partial variable names to be searched/printed
VMACSMF but previously they were set to missing value so UTILGETM Dec 8, 2010 did not report them. For actual ID=102 processing, MXG uses the IFCID value so this had no real impact. -Cosmetic. Back-to-back ID=2 did not print the LAST RECORD IN GROUP message. SMFHDRCN now keeps track. -Reminder for reading only part of an SMF file: While using OPTIONS FIRSTOBS=100 OBS=500; can sometimes used, to read only those input records, if there is any post-processing (deaccumulate, sort, etc.) it won't work! Instead, use %LET SMFEXIT= FIRSOBS=100 OBS=500; which will used those on the INFILE and thus only impact which records are read, not touching the global options.
ANALDBJS reading PDB.JOBS and PDB.JESNR and sequencing DB2ACCT Dec 8, 2010 by JOB and SMFTIME to propagate those variables. Thanks to Jane Stock, USPS, USA. Change 28.289 Variables GATLIMIT and GATCNT are now KEPT in TYPE6156 so VMAC6156 that changes in GDG limits can be observed. Dec 8, 2010 Thanks to Jorge Fong, NYC Information Technology, USA. Change 28.288 Cosmetic. If no observations are found with the searched VMXGSRCH values, now, a note that no observations were found is Dec 7, 2010 printed on the log.
ANALTMVT TMVTSX dataset with the "Session Extended Information". EXTMVTSX New member ANALTMVT replicates for ASG-TMON VTAM reports. IMACTMVT
VMACTMVT VMXGINIT
Dec 6, 2010 Change 28.286 Variables QAINTS and QAINTE, interval start/end datetimes VMACTMMQ should have been kept in TMMQQAA dataset, and now are. Dec 6, 2010 Thanks to Homayoun Riazi, United Health Group, USA. Change 28.285 CICS STID=143 sub-subtype printed message that six bytes VMAC110 of new data was skipped, but there was no new data; MXG Dec 3, 2010 incorrectly input ECCEVCAP as only 4 bytes, when it is 8, so ECCEVCAP/ECCCAPFA/ECCEMIFA in CICECC Statistics Data set were wrong, and the subtract of 146 is now 152 after that correction. Thanks to Tom Kelman, Commerce Bank of Kansas City, USA.
ASUMUOWT (_LCICTRN _LMONTSK _LDB2ACC) do not exist, or it will VMXGUOW construct the macros based on the presence or absence VGETENG of the corresponding &P****** macro. If for example Dec 3, 2010 &PCICTRN is empty or the DDNAME is not found, the _LCICTRN macro is set to CICSTRAN. If it is found, the macro is set to &PCICTRN..CICSTRAN. If neither CICSTRAN nor MONITASK exist, a warning is printed, but the output ASUMUOW dataset is created with OBS=0. -ASUMUOWT (for TMON instead of CICSTRAN input) will now call VMXGUOW, so there is only one macro to maintain; this obsoletes VMXGUOWT as no longer required. -VGETENG NOEXIMSG=YES added as a default. NOEXIMSG=NO suppresses the MXGNOTE on the log as it does in the other %VGET macros. Thanks to Ken Goodis, Emblem Health, USA. Change 28.283 Many RACF segments have a variable containing CLASS*NAME VMAC80A for that specific RACFTYPE/SMF80DTP, but some had only Dec 2, 2010 "CLASS*NAME" for their label value. Now, the RACFTYPE value is included to make the LABEL value unique. Thanks to John Matson, EPSON, USA.
EXTY89I to dataset TYPE89, and which adds new Intersect Data that IMAC89 creates new TYPE89I dataset for Measured Usage reporting. VMAC89 -Variable SMF89SYN is added to each of the BY lists as the VMXGINIT last variable; if you have duplicate SYSTEM names in your Dec 2, 2010 SYSPLEX, then SMF89SYN will be different than SYSTEM. Change 28.281 PDB.ASUMCICR dataset contains the count/average response ASUMCICR time by DATE for each REGION/APPLID, and can be created Dec 1, 2010 from transaction detail MONITASK or CICSTRAN or ASUMUOW datasets, or the summary PDB.CICS dataset (built by ASUMCICS/ASUMCICX), or, if the input is WEEK.ASUMCICR or if INDATA= MON.ASUMCICR ... SUN.ASUMCICR, the prior sums will be re-summed to include partial days for each DATE. And, if NODATE=YES, is specified, whatever input is in the INDATA= argument will be summarized only by APPLID, (in case your manager thinks a weekly average of all of the week's transactions in a region is a useful metric!). The PDB.ASUMCICR dataset also summarizes TASCPUTM. Note: These values may be of little use, if your site uses Multi-Region-Option MRO and you read transaction detail datasets (MONITASK/CICSTRAN), where the counts will be inflated and false, since each one of the multiple observations of an MRO transaction (one TOR, one-to-many AOR, one-to-many DOR/FOR obs) will each be counted as a separate "transaction", which they aren't! On the other hand, if there are very few MRO trans, and your APPLIDs are stable, these counts/averages might be useful for tracking quantity and response. Thanks to Ken Goodis, Emblem Health, USA.
ANALRMFR some BY lists with repeats of SYSNAME were corrected. Nov 30, 2010 Thanks to Bruce Hewson, Citibank N.A., SINGAPORE. Change 28.279 ASUM113 used SMFTIME to define each interval, but SMFTIME VMAC113 can have multiple values in the records for an interval; Nov 29, 2010 it can take a second to write all of the records for one interval. If SMFTIME had different .01 second values, ASUM113 incorrectly created "near duplicate" observations with wrong values. As no "start of interval" flag exists in SMF 113 records, this revision uses the time value in the SM113CPU=0 and SM113CST=1 and SM113CPT=0 observation to populate SM113STM, the Interval Start Time, for each interval. An additional error when the GMT OFFSET was was positive that could cause a one-second error in the converted timestamps was corrected. Thanks to Adnan Can, Garanti Teknoloji, TURKEY.
VMACRMFV
Nov 28, 2010 Thanks to Matthew Chappell, Dept. of Transport Main Roads, AUSTRALIA Change 28.277 Variables NETSNAME and UOWTIME are created in DB2ACCT so VMACDB2H that DB2 observations with QWHCATYP=4, i.e., CICS, can be Nov 25, 2010 merged with CICSTRAN to create PDB.ASUMUOW. Those vars Nov 29, 2010 are now populated ONLY for DB2ACCT observations from CICS (i.e. QWHCATYP=4). Changes to NETSNAME creation in DB2 in MXG 28.05 caused non-CICS DB2ACCT obs to have changed values in last 4 characters that caused no harm except to show up as differences in PROC COMPARE, but as there is no value in creating NETSNAME for these observations, to avoid confusion, they are no longer populated. -However, some values of NETSNAME were not correctly created, if the last four characters happened to contain a period in those hex values. The logic was revised to only scan the first 16 bytes of QWHCTOKN for the period. -Also, if there WAS a period in the first 16 bytes, then the resultant NETSNAME value was non-blank in the last four bytes; now it is populated with only the first 16 bytes of QWHCTOKN in that instance. -These MXG errors could cause PDB.ASUMUOW to have fewer observations than it should. Thanks to Paul Volpi, UHC, USA. Thanks to Matthew Chappell, Queensland Dept of Transport, AUSTRALIA. Change 28.276 ASUMHSM enhanced with optional macro to select HSM data ANALHSM to be summarized, and ANALHSM also enhanced to support ASUMHSM selection with BEGTIME/ENDTIME. Nov 18, 2010 Dec 11, 2010 Thanks to Doug Medland, IBM Global Services, USA. Change 28.275 -Support for NMON FCREAD/FCWRITE/XFERIN/XFEROUT records EXNMONFC creates new dataset PDB.NMONFC for the Fibre Channel data VMACNMON for AIX and Linux. VMXGINIT -Support for NMON DISKXRFER (disk transfer reads per sec) Nov 17, 2010 creates new variable DISKXRFR in PDB.NMONDISK. Dec 2, 2010 -Support for DISKBUSY/READ/WRITE/XFER/BSIZE/SERV/WAIT with Dec 9, 2010 suffixes thru 21. -Invalid UARG record with only four fields and the fourth containing text of PPID COMMAND THCNT USER GROUP COMMAND is detected and printed on the log and deleted. Thanks to Xiaobo Zhang, FISERV, USA.
VMACNMON 512 bytes to capture the full text of commands, and the Nov 17, 2010 elements of the WORDS array are also increased to $512. Thanks to Xiaobo Zhang, FISERV, USA. Change 28.273 MXG support for CA NSM/TNG only output 4-digit values in VMACTNG the generated %LET statements with number of Instances, Nov 13, 2010 creating %LET xxxxxx=12E3; which is not valid syntax for the macro language. The %LET counters now put 6 digits. Thanks to Xiaobo Zhang, FISERV, USA.
VMAC7072 documented that the last 3 nybbles of those 8-byte TOD VMAC89 Clock fields can be non-zero but are NOT used by SCRT. Nov 13, 2010 MXG input those fields, which caused fractional seconds that did not exactly match SCRT reports. Now knowing the idiosyncrasy of these fields, MXG now zeros those last three nybbles prior to their input to match SCRT. Thanks to Al Sherkow, I/S Management Strategies, Ltd., USA. Thanks to Charles E. Hackett, SCRT-IBM, USA.
VMAC113 in CICS exit years ago, but when they added SMF 113 to Nov 11, 2010 their daily BUILDPDB, it died because that name was used as an ARRAY name in VMAC113, an unacceptable use. While the site easily renamed their temporary variable to avoid the conflict, I changed COUNTER to CNTR113. Change 28.270 Documentation. The successful JCLTEST9 execution prints JCLTEST9 UNINITIALIZED variable messages in two places. There are Nov 11, 2010 335, mostly with variable names AD0nnxxx immediately prior to NOTE: Dataset WORK.SV01EV01 has 0 observations. There are 120 more with various names before the NOTE: Dataset WORK.AIXPTXIN has 0 observations. Thanks to Andrew Woods, Interactive Data, ENGLAND.
WEEKBL3 for DST, combining daily PDBs all created by the same MXG WEEKBL3D Version! Discovered GMTOFF72 was in BY list in VMAC7072 WEEKBLDD for TYPE72DL/TYPE72GO/TYPE72MN/TYPE72SC datasets, but not WEEKBLDT in the WEEKBLDs. Have now added GMTOFF72 after STARTIME WEEKBLD in all WEEKBLDs. Nov 8, 2010 Jan 16,2011: Now, see Change 28.324.
UTILNPRT in the formatted value of all character variables in all Nov 8, 2010 SAS datasets in a SAS data library. SAS Enterprise Guide before 4.22 failed on non-printable DB2 data, so this was written to examine what variables could cause problems. Most variables that contain hex data in characters are formatted with $HEX precisely to eliminate the non-printables, or $EBCDIC field's '00'x are changed by MXG to blanks. But some variables have mixtures of EBCDIC and HEX; while these could be formatted $HEX, sometimes that EBCDIC text is useful in PROC PRINTs, and it would be lost in hex characters with $HEX, so I leave the variable unformatted, leading to this kind of exposure, hence the utility. If you have a problem, just add a FORMAT vvvvvvvv $HEXnn. statement in the EXdddddd exit for the dataset, where nn is twice the length of the character variable. Thanks to Stephen Hughes, Excellus, USA. Change 28.267 Optional argument NOEXIMSG=YES added so that messages VGETENG could be suppressed when not wanted, used internally by Nov 5, 2010 other MXG programs that use VGETENG.
IEBUPDTE directory of files from an IEBUPDTE-format input file now Nov 4, 2010 looks for either './ ' or '.XY' in columns 1-3 to flag a new "member", so the PRODTEST member can be read directly without EDITing those '.XY' into the './ ' that is needed by the z/OS PGM=IEBUPDTE. The MXG source library has to have '.XY' inside these members that contain a PDS in IEBUPDTE-format, because there might still be someone actually using PGM=IEBUPDTE on z/OS to create their MXG.SOURCLIB PDS, if they chose to download the (ARCHAIC) ebcvvnn.ebc EBCDIC MXG install file to z/OS, as those './ ' would create unwanted new PDS members on z/OS. On z/OS the tervvnn.ter tersed-PDS MXG Install File should be downloaded instead.
it's for performance when writing in-place, so it was removed. -The INFILE option TERMSTR=CRLF is now in comments, as it doesn't exist in SAS V9.1.3 nor WPS, and it is only needed if this IEBUPDTE program is executed on unix to read a windows-created CRLF-terminated text file. On unix, TERMSTR=LF is the default text line terminator. Thanks to MP Welch, Aprize, Inc, USA.
ASUMCACH DEVMODEL='3380K' (i.e., with alpha character) when the Nov 4, 2010 new statement MODEL=INPUT(DEVMODEL,HEX8.) was executed. Now, that statement is protected if DEVMODEL is not a hex value (e.g., MODEL=3380x for DEVMODEL='3380K'. Thanks to Tom Heller, CINCOM, USA. Change 28.264 Support for DB2 Version 10. COMPLETELY INCOMPATIBLE: EXDB2ACR MXG 28.06 was required to process the V10 data, but now, EXDB2ACW MXG 28.07 has full support plus the below documentation. FORMATS
IMACDB2 -DB2 V10 Records can be compressed. Member EXITCICS is VMACDB2 updated to decompress SMF 110-1 and SMF 100,101,102s. VMACDB2H VMACSMF -INVALID DATA FOR QWHSRELN is proof you have DB2 V10 SMF; VMXGINIT QWHSRELN is not a valid "PK" value; it now has 'A1'x, so VMXGWORL VMACDB2H was revised. The value is 10.1, not 10.0. Jun 16, 2010 -And INPUT STATEMENT EXCEEDS RECORD error ABENDs may occur Jun 19, 2010 because fields were inserted rather than added at the end Jun 21, 2010 where MXG would have automatically skipped them. Nov 4, 2010 -Subtype in SMF Header INCOMPATIBLY increased from one to two bytes (because that's what it should have been all along. However, only SMF 100 and 101 records populate the subtype in the SMF Header. VMACSMF was updated to get the DB2 version and then input the subtype correctly. (MXG has always stored the IFCID value in SUBTYPE for the DB2 102 trace records, since they don't have a subtype.) -New DB2ACCGW dataset is created for each QWAR segment, which can be used to correlate rollup records. -Multiple SMF 100 Subtype 1 (DB2STAT1) IFCID=2 records are now supported. These records contain only QBST or QBGL segments and are written when more than 25 buffer pools are used in an interval. -Macro _SDB2STS was redesigned to correct DUPLICATE MERGE VALUES errors that occurred only if DB2 was restarted, by removing QWHSACE QWHSMTN from the _BDB2STS "BY list", by interleaving the four input datasets to create STATSGROUP to use as merge variable (QWHSSTCK cannot be used as it not exact in all four records for each interval), and by conditionally merging T102S225 (DB2 V8) if it exists. The _SDB2STY macro is now a null macro and no longer used. The new sort order for the PDB.DB2STATS dataset is now SYSTEM QWHSSSID QWHSSTCK. A cosmetic enhancement to VMXGWORL, NOWARN=YES, is used to suppress the MXGNOTEs when a tested dataset is known to not always exist (used for T102S225 in this change). -Many new variables added to DB2ACCTx/DB2STATx by V10: DB2ACCT:
QLACRLNU QXSTCWLP QXSTCWLR QXSTCWLM QXSTCWLD DB2ACCTP: QPACAWLH QPACANLH QPACRLNU QWACPCTT QWACPQRY QWACAWLH QWACARLH DB2ACCTB: QWACPCTT QWACPQRY QWACAWLH QWACARLH DB2ACCTP: QPACAWLH QPACANLH QPACRLNU QWACPCTT QWACPQRY QWACAWLH QWACARLH DB2ACCTG: QWACPCTT QWACPQRY QWACAWLH QWACARLH DB2STAT0: Q9STCDMD QDSTNQWC QDSTNARD QDSTMARD D64POST A64POST A64WAIT M64DISNU M64DISPG SGETR64 SGETEXT6 SGETDEXT SFREER64 SFREEDEX DISCARDM QWS1MCPU QWS2MCPU QWS3MCPU QWS4MCPU QXSTCWLP QXSTCWLR QXSTCWLM QXSTCWLD DB2STAT1: QISEKSPG QISEKSPA QISEKLRU QISEDLRU QISESQCB QISESQKB QISESQCA QISESQKA QISTRCCI QISTRCCD QISTRCCU QISTWMXA QISTWMXU QISTWCTO QISTW4K QISTW32K QXSTCWLP QXSTCWLR QXSTCWLM QXSTCWLD DB2STATB: QBSTFLG
DB2GBPST: QBSTFLG
-SMF 102 IFCIDs 172 and 196 were compatibly updated. -SMF 102 IFCIDs 267, 268, 317, and 337 are now decoded. New formats are created by the updated FORMATS member. -These other IFCIDs in user-sent SMF files are presumably the trace records that are normally written or needed. They have been examined and none were change in V10: 4,5,55,87,105,140,141,173,196,199,250,254,258,261,262 -See the text of Change 28.146, which made changes to DB2 processing that were independent of the Beta. Thanks to IBM DB2 V10 Beta for both EARLY DATA AND DOCUMENTATION!! Change 28.263 Support for IBM/OMEGAMON XW MQ flat file (INCOMPAT) adds VMACOMMQ new formats for UTF-8 data, and MXG only tested for up to Nov 2, 2010 20 datasets, but there can be 26 in total. Thanks to Michael Reffler, Credit-Suisse, USA Change 28.262 Support for CA MIM RESOURCE SHARING R11.7 (COMPAT) adds VMACMIM new variables to several datasets, and many variables Nov 2, 2010 are now correctly input and formatted, especially times in MIMCF dataset. Subtypes 0/1/2/4/5/7/8/9/256 and 256 have been tested. Thanks to Tony Curry, BMC, USA. Change 28.261 SAS/GRAPH example that uses PDB.ASUMCELP (LPAR-CEC level) GRAFCEC (built by ASUM70PR) to create charts of CEC Utilization Nov 1, 2010 for each of the engine types (CP,IFA,ZIP,IFL).
VMACTPX TPX Version 4.0 but were overlooked; now they are input Oct 29, 2010 in 05/06/07/08 subtypes, when possible: -Invalid subtype 7 records, LENGTH=101, LRECL=104, with LEN07=93 in bytes 58-59 of the data portion indicating the record should contain IP Address and Port, but the record has only 44 bytes remaining in the record, i.e. the IP Address/Port field are not present. -Subtype 8 record with a IP Port field that is not a valid EBCDIC numeric (>?01) with an IP Address with manually typed all text characters (ABCDEFGHI...) caused hex dump and error message, now both suppressed with ?? modifier. -These records were supposedly created by TPX 5.3, but the Version value in the records is 4.0. Thanks to Dennis Longnecker, State of Washington Courts, USA.
UTILEXCL messages were printed by _BLDEXCL because only the first Oct 29, 2010 to the three repeated (for emphasis) PUT statements was conditional; the 2nd-3rd PUTs always printed a pair of this warning message. Only if there are THREE adjacent warning messages (then and now) does the warning apply. The three PUTs are now inside a conditional DO-Group. Thanks to Mrs. Brigitte Wallbaum, FINANZ INFORMATIK GMBH, GERMANY. Thanks to Mr. Dieter Haak of FINANZ INFORMATIK GMBH, GERMANY.
VMAC120 been validated (and rewritten) with actual data records. VMACSMF -ID=120 SUBTYPE=20 records are INVALID because they do NOT Nov 1, 2010 set the "RECORD CONTAINS SUBTYPES" bit in Byte One of the SMF header (all other 120 subtypes DO set that bit), so VMACSMF had a missing value for SUBTYPE for ID=120-20s. Now, VMACSMF always reads the 2-byte subtype for 120s, whether or not the bit is enabled. -Dataset TY12020 is now populated, and the Start and Last Datetimes are converted from GMT to Local Time Zone. -There is an undocumented duration field SM120XZ at the Yüklə 28,67 Mb. Dostları ilə paylaş:
|