COPYRIGHT (C) 1984-2021 MERRILL CONSULTANTS DALLAS TEXAS USA
MXG NEWSLETTER FIFTY-THREE
***********************NEWSLETTER FIFTY-THREE***************************
MXG NEWSLETTER NUMBER FIFTY-THREE, FEB 3, 2009
Technical Newsletter for Users of MXG : Merrill's Expanded Guide to CPE
TABLE OF CONTENTS
I. MXG Software Version.
II. MXG Technical Notes
III. MVS, aka z/OS, Technical Notes
IV. DB2 Technical Notes.
V. IMS Technical Notes.
VI. SAS Technical Notes.
VI.A. WPS Technical Notes.
VII. CICS Technical Notes.
VIII. Windows NT Technical Notes.
IX. z/VM Technical Notes.
X. Email notes.
XI. Incompatibilities and Installation of MXG.
See member CHANGES and member INSTALL.
XII. Online Documentation of MXG Software.
See member DOCUMENT.
XIII. Changes Log
Alphabetical list of important changes
Highlights of Changes - See Member CHANGES.
COPYRIGHT (C) 1984,2009 MERRILL CONSULTANTS DALLAS TEXAS USA
I. The 2008 Annual Version MXG 25.25 was dated January 28, 2008.
All sites were mailed a letter with the ftp download instructions.
The availability announcement was posted to both MXG-L and ITSV-L.
You can always request the current version using the form at
http://www.mxg.com/ship_current_version.
1. The current version is MXG 26.26, dated Feb 3, 2009.
See CHANGES member of MXG Source, or http://www.mxg.com/changes.
II. MXG Technical Notes
1. Recent measurements of the revised MXG QA Stream (no PROC COPYs).
With SAS V9.1.3 and MXG 26.09 on two z/Series machines:
On z/OS:
Machine Elapsed CPU Virtual SAS
SU_SEC minutes min storage OpSys Version
21621 90.7 16.50 123558K z/OS 9.1.3
9708 104.8 45.19 124340K z/OS 9.1.3
3.2GHz 9.4 6.2 146173k WinXP 9.1.3
3.2GhZ 10.0 6.25 101840k WinXP 9.2
III. MVS, a/k/a z/OS, Technical Notes.
21. One site experienced several minutes of extremely bad performance
that was caused by an HDS hard disk failure: HDS says the impact of
future failure can be lessened by turning on SOM (System Option
Mode) 359. When the controller hardware detects a DASD sector
error, it reassigns that sector to a spare sector and rebuilds it
there. If SOM 359=OFF, the time out value to indicate sector
failure and initiate reassignment of the sector is 4.5 sec and
after eleven sector failures the hard drive is blocked and disk
sparing initiated. With SOM 359=ON the time out value for sector
failure and reassignment is 3 sec and after three such failures the
hard drive is blocked and disk sparing initiated. HDS says that
"given the proactive approach to maintenance" at the site, they
recommend setting SOM359 to ON for the USPV controllers. This
might cause more hard drive sparing to be triggered in the
environment but looks like it would lower the adversely impacted
time when sparing occurs from about one minute to about ten
seconds.
20. APAR PK78309 for WebSphere SMF 103 subtype 2 record documents four
previously internal-use fields, but they have been decoded by MXG
for years in variables THWANSSL THWASSL THWAASYN THWAMSGQ
19. SYNCSORT 1.3 - Records Dropped by Sort program.
The sort processes to good End of Job, but records are dropped.
This problem occurs when one or more empty datasets are in the
SORTIN concatenation, and PAV aliases are assigned to the SORTIN
I/O, and SyncSort screws up their EOV processing.
Fix is EW6629
18. APAR PK77184 reports QPACCLS7_ZIIP for all packages is higher than
the value of QWACTRTT_ZIIP when there are repeated calls to a
trigger without nesting results.
17. APAR OA26693 reports High CSM 32K ECSA buffer usage caused storage
shortage, when HiperSockets users were running with either a zIIP-
enabled application such as DB2 performing socket API while on a
zIIP, or, running with TCP/IP zIIP IPSec enabled.
16. APAR OA26761 reports missing channel paths in RMF IOQ data when new
channel paths that were added to existing devices by a Dynamic
Activation of an I/O configuration.
15. APAR OA26488 reports performance enhancements for queued XES locking
requests and corrects the count of requests that are CHNGD from
synchronous to asynchronous in the RMF Coupling Facility Structure
Activity report.
14. APAR OA26753 reports no Link Statistics from 2107 with Microcode R3.
13. APAR OA26842 reports that READTIME was binary zeros in SMF 42
subtype 9 records.
12. FMIDs identify the base level of an IBM product; for z/OS they are:
HBB7709 z/OS 1.6
HBB7720 z/OS 1.7
HBB7730 z/OS 1.8
HBB7740 z/OS 1.9
HBB7750 z/OS 1.10 (a guess)
11. APAR OA26842 reports SMF 42 subtype 9 field S42RDST, the time part
of READTIME is not being populated.
10. APAR OA26555 reports HIGH CPU IN VLF address space, and possible
causes.
9. APAR II13495 is a long discussion of "HOW DFSORT TAKES ADVANTAGE OF
64-BIT REAL ARCHITECTURE".
8. APAR OA08719 reports FMID HBB7709, z/OS 1.6, SMF 30 records contain
zeros in the step termination (subtype 4) record for IFA CPU time
fields "SMF30_TIME_ON_IFA" and "SMF30_TIME_IFA_ON_CP" which are MXG
variables CPUIFATM and CPUIFETM in PDB.STEPS an TYPE30_4 datasets.
Those fields were filled in from fields that were erroneously
cleared before the record was written.
7. What products execute on zIIPs?
This note will be updated as new programs/products are reported:
- IBM DB2 and DB2 Utilities
- IBM DFSORT for additional zIIP offload by DB2 Utilities
- IBM Communications Server for IPSEC
- IBM Communications Server for Hipersocket Large Messages.
- IBM XML, some parts of the parser.
- IBM Extended Remote Copy (XRC) data mover address spaces (OA23174).
- SYNCSORT
- IBM Scalable Architecture for Financial Reporting (SAFR)
- System Data Mover (SDM) benefits (z/OS Global Mirror).
- Encryption and Compression:
- CA-Vtape (option to use software compression to compress the cache)
- CA Brightstore Tape Encryption (both compression and encryption).
- CA Datacom
- CA Netmaster for IP Packet Analysis (also uses zAAPs)
- CA Insight for DB2 (when the SQL is already running on a zIIP.
- CA IDMS R17 offloads some system mode time to zIIPs.
- PKZIP
- Progress Software - Data Direct
- Phoenix Software - Most products, including (E)JES
- Shadow 7 (from DataDirect, uses zIIP and zAAP)
- BMC, most data collectors, converted from TCB to SRB for zIIP
Effective August, 2009, the majority of CMF data collectors
(BMC's replacement for RMF) were converted from TCB to SRB and
became eligible for dispatching to zIIP;
- Neon Enterprise Software - IMS Utilities - Out of business.
An earlier version of this list showed DFSORT but that is not
correct. IBM's official position, Jan 21, 2009 is stated:
At this time, IBM has no plan for enabling DFSORT to exploit the
system z9 Integrated Information Processor (zIIP). IBM realizes
DFSORT remains a prominent component of our customers' batch
workloads. However, the added controls that would need to be
implemented in order to maintain our high standards for
performance, reliability and system integrity are not justified
in view of estimations that there is a low offload potential and
the value to clients may be marginal. IBM will continue to
focus its DFSORT development efforts on the enhanced function,
performance, reliability and service items that we believe
provide the most value to our clients. The foregoing represents
IBM's current intent and is subject to change.
But, now, August 1, 2009, APAR PK85856, provides this statement:
"DFSORT support for additional zIIP offload by DB2 Utilities. In
conjunction with DB2 APAR PK85889, this PTF enables DB2 Utilities
to offload portions of sort workload to zIIPs when they are
available.
6. APAR OA26611 reports HyperPAV can cause RMF DASD Device TYPE74 data
can report misleading data for CONNECT, DISCONNECT, PEND & RESPONSE
times.
5. APAR OA26077 for z/OS 1.10 reports an undetected loss of data with
a multi-volume tape dataset with QSAM and BUFNO GT 1 or BSAM with
NCP GT 1 specified, impacting applications such as HSM, IEBGENER,
IDCAMS and DB2.
4. APAR OA25243 for RMF III notes that GDACSAHWM and GDAECSAHWM report
the high water marks of CSA and ECSA allocations from the (E)CSA
areas, but fail to take into account the use of (E)CSA for (E)SQA
allocations, giving a misleading representation of the amount of the
(E)CSA areas available for additional (E)CSA allocations.
3. APAR OA26027 for the IFASMFDL program (the "SMF DUMP" if you send
your SMF data to the System Logger instead of to a VSAM file) fixes
an incorrect use the START parameter; without the PTF for the APAR,
records were copied from a START of zero rather than your desired
START time, so many extra records could have been selected.
2. Information APAR II10549 for TYPE70PR lists the diagnostics and
instructions to open a "hardware PMH" if LCPUEDTM GT LCPUPDTM,
i.e., if Effective Dispatch time greater than Total Dispatch time,
usually due to a non-responding coupling facility.
1. APAR PK67436 corrects the TCP Port statistics SMF record, Type 119,
subtype 7, which had an incorrect number of current sessions.
IV. DB2 Technical Notes.
2. APAR PK74210 reports INCORRECT VALUES FOR QWACCLS1_ZIIP (IIP CPU
TIME) for RRS threads that do TCB agent switching. That is MXG
variable QWACZIS1 in DB2ACCT dataset.
1. APAR II14438 lists known issues with OMEGAMON DB2 that can cause
high CPU utilization by OMEGAMON, and the Information APAR has
some performance tuning tips.
V. IMS Technical Notes.
VI. SAS Technical Notes.
5. Do NOT use BUFNO= on any DD statement for a SAS Data Library, or
you can anticipate a S878-10 ABEND.
The DCB is primed with the BUFNO from the JCL, but SAS is not
clearing the BUFNO from the DCB in the open exit. The system then
allocates the buffers in virtual, but SAS uses its own area of
storage above the line for the I/O buffer. Any change would require
a significant redesign in SAS to clear the DCB BUFNO field, which
is a future objective in a new SAS Note on this ABEND.
4. SAS Usage Note 34071 reports that the use of DSNTYPE=LARGE
on a DD in JCL for a SAS Data Library may generate SAS Error:
ERROR: OPEN TYPE=J failed to position library data set
libname.xxxxxxxx.DATA to volume number 1.
The error is corrected in hot fix E9BC06 that is associated
with SAS Note 17038. Jan 12, 2009.
3. SAS V9.2 Warnings when combining datasets with different lengths.
As previously documented, SAS V9.2 prints new WARNING messages:
WARNING: Multiple lengths were specified for the variable XXXXXXX
by input dataset(s). This may cause truncation of data.
when datasets with different length variables are combined, i.e.
with SET, MERGE, UPDATE, etc.
MXG History:
Apr, 2008. MXG 26.03. Internal code changes in MXG eliminated
the warnings, but, also
Added VARLENCHK=NOWARN to VMXGINIT
Aug, 2008. MXG 26.07. Removed VARLENCHK=NOWARN to VMXGINIT.
because Hot Fix F9BA07 restored behavior to 9.1.3.
BUT: That Hot Fix was ONLY for 9.2 TS1M0.
Apr, 2010 SAS 9.2 TS2M0 and TS2M2 reinstated the WARNINGs by
default, but documented VARLENCHK=NOWARN as the
way to eliminate the warning.
MXG 28.02 reinstated VARLENCHK=NOWARN in VMXGINIT,
so that the WARNING will NOT be printed nor will the
condition code be set. This protects future MXG
versions when MXG has to change a variable's length
but also protects user code from the WARNING.
But not all variables with different lengths are WARNED. The two
datasets OLD and NEW contain 2 character and 2 numeric variables
with different LENGTHs as shown. Combining OLD and NEW to create
DATA BOTHOLD; SET OLD NEW; with OLD listed first printed warnings
only for variables CH2 and NR1; changing the order with NEW first,
DATA BOTHNEW; SET NEW OLD; instead printed warnings for variables
CH1 and NR2.
So the WARNING is only printed for the variable that is truncated
(i.e. when the shorter-length is in the first input dataset, as
that shorter-length sets the variable's length in the output data.
So BOTH:OLD-NEW's variables CH2 and NR1 are truncated and WARNED,
and BOTH:NEW OLD's variables CH1 and NR2 are truncated and WARNED.
Dataset: OLD NEW BOTHOLD:OLD-NEW BOTHNEW:NEW-OLD
CH1 $8 $4 $4 $8 W
CH2 $4 $8 $8 W $4
NR1 5 8 8 W 5
NR2 8 5 5 8 W
Because lengths are taken from the 1st dataset in the SET statement
it is STRONGLY recommended that you install a new MXG version into
production on the FIRST day of a new week (e.g., if your PDBs run
from MON to SAT, install on Tuesday morning so the MONDAY PDB will
have any new variable lengths, so that that first-day-of-week PDB
will set the length of the next WEEK's PDB library).
2. What SAS Procedures are included in SAS Base Product with SAS V9.2?
This list of SAS Procedures in Base SAS V 9.2 is taken from that
version's SAS Procedures Manual:
APPEND BMDP CALENDAR CATALOG CHART CIMPORT
COMPARE CONTENTS CONVERT COPY CORR CPORT
DATASETS DBCSTAB DISPLAY DOCUMENT EXPLODE EXPORT
FCMP FONTREG FORMAT FORMS FREQ IMPORT
ITEMS JAVAINFO MEANS MIGRATE OPTIONS OPTLOAD
OPTSAVE PDS PDSCOPY PLOT PMENU PRINT
PRINTTO PROTO PRTDEF PRTEXP PWENCODE RANK
REGISTRY RELEASE REPORT SCAPROC SOAP SORT
SOURCE SQL STANDARD SUMMARY TABULATE TAPECOPY
TAPELABEL TEMPLATE TIMEPLOT TRANSPOSE TRANTAB UNIVARIATE
VAXTOINTEG WEBMDDB
1. For z/OS, in a LIBNAME statement: If you specify a GDG as +0 (which
is valid syntax) SAS refuses to allocate the LIBNAME saying it does
not support adding members to a GDG, but +0 is not adding one.
So,
LIBNAME PDB 'MXG.PDB(+0)' DISP=SHR; fails
LIBNAME PDB 'MXG.PDB(0)' DISP=SHR; works
VI.A. WPS Technical Notes.
1. WPS cannot read a z/OS RECFM=VB file when RECFM=VBS is specified.
ERROR: Failed to open SMF : EDC5129I No such file or directory.
The RECFM=VBS parameter is INTENTIONALLY used in the VMACSMF code,
because VBS is a superset of both V and VB files on z/OS, and so
the VBS specification should always process any of the three file
types that might be presented. This is a WPS design error that
has been reported to WPS technical support and this note will be
revised when the error is corrected.
VII. CICS Technical Notes.
1. MQ TCB CPU time included in CICSTRAN and TYPE30 and TYPE72.
With CICS/TS 3.2 and Websphere MQ 5.3.1 or later, using the CICS
Attachment Facility, the MQ TCB CPU time consumed on behalf of a
CICS transactions (which will be under the L8 and/or KY8 TCBs) is
included in IBM SMF 110 subtype 1 fields USRCPUT, L8CPUT, and
KY8CPUT, which are MXG variables TASCPUTM, L8CPUTTM and KY8CPUTM in
MXG dataset CICSTRAN. That MQ CPU time is also recorded in the
address space of the CICS region, so it is also in the CPUTCBTM in
the type 30 records for that "job" and in the TYPE72GO for that
address space's service class.
VIII. Windows NT Technical Notes.
IX. z/VM Technical Notes.
X. Email notes.
1. Outlook Restore Line Breaks corrupts received ftp instructions.
Our ftp instructions for MXG download are sent by email from
an Outlook client. One MXG user noticed that his Outlook
client was combining two lines of text into a single line.
The two LOCSITE ftp control statements in Appendix Ds example
were combined, while the identical pair of LOCSITE statements
in Appendix A were correctly seen as two lines.
These four lines of text in the original ftpxxxx.txt file:
BINARY
LOCSITE LRECL=1024 RECFM=FB BLKSIZE=6144 UNIT=SYSDA
LOCSITE PRIMARY=5000 SECONDARY=300 GET ... (replace
CLOSE
QUIT
Were received as
BINARY
LOCSITE LRECL=1024 ... UNIT=SYSDA LOCSITE PRI... SECONDARY=
300 GET ter2605.ter 'MXG.MXG2605.TERSED' (replace CLOSE QUIT
The received message also contained the "Restore Line Breaks"
message, and Microsoft KB articles did document that that was
a flag that this issue was unique to "Plain Text" messages, and
suggested that use of html would circumvent the problem.
But html has these exposures:
-It destroys collimation, unless I change to a fixed font,
-HTML documents can contain executable stuff, and ours has
many instances of the same url, which can trigger corporate
spam filters to false-positive on our ftp instructions, i.e.
you don't get our instructions, and
-the url you see and can finger in an html document can be
quite different than the actual link - you have to display
the html document in plain text to see if the actual url/ip
address is what was displayed, i.e., html documents can
contain false-url-pointers.
MicroSoft also suggested you could eliminate this problem
by disabling this Outlook option
"Remove Extra Line Breaks in Plain Text Messages"
by unchecking that option under
TOOLS ==> EMAIL OPTIONS == Uncheck Remove Extra Line Breaks...
But: if you didn't know to remove that option, then your ftp
download example's text was corrupted.
The solution was to alter the email creation:
There were TWO sets of those five lines in the JCL examples in
our ftp download instructions, but only the second set of lines
were corrupted. I finally noticed that the first pair had
blanks before the text, while the failing pair started in
column 1. By experimentation, I discovered that insertion of
two blanks (but not just one!), BEFORE (and not after, as some
have claimed) eliminated the corruption in the received email,
no matter what you had done with that Outlook default option.
Now chuffed that I had found a circumvention, if not a real
solution and now Googling the EXACT text of the Outlook option.
I found this circumvention reported at least as early as 2005
December 17, 2005. Robin Good's Email Newsletter.
BUT NOT BY MICROSOFT, EVEN NOW, AT THE END OF 2008.
Now, I add two blanks before each line in the MXG ftp download
instructions, and, for those of you using the MicroSoft Outlook
(or I suspect Exchange Server) as your email client, the
control statements in the MXG JCL examples will not be
merged, even using the (ill-advised) default Outlook option.
XI. Incompatibilities and Installation of MXG vv.yy.
1. Incompatibilities introduced in MXG 26.yy (since MXG 25.25):
See CHANGES.
2. Installation and re-installation procedures are described in detail
in member INSTALL (which also lists common Error/Warning messages a
new user might encounter), and sample JCL is in member JCLINST9 for
SAS V9.1.3 or JCLINST8 for SAS V8.2.
XII. Online Documentation of MXG Software.
MXG Documentation is now described in member DOCUMENT.
XIIV. Changes Log
--------------------------Changes Log---------------------------------
You MUST read each Change description to determine if a Change will
impact your site. All changes have been made in this MXG Library.
Member CHANGES always identifies the actual version and release of
MXG Software that is contained in that library.
The CHANGES selection on our homepage at http://www.MXG.com
is always the most current information on MXG Software status,
and is frequently updated.
Important changes are also posted to the MXG-L ListServer, which is
also described by a selection on the homepage. Please subscribe.
The actual code implementation of some changes in MXG SOURCLIB may be
different than described in the change text (which might have printed
only the critical part of the correction that need be made by users).
Scan each source member named in any impacting change for any comments
at the beginning of the member for additional documentation, since the
documentation of new datasets, variables, validation status, and notes,
are often found in comments in the source members.
Alphabetical list of important changes after MXG 24.24 now in MXG 25.01:
Dataset/
Member Change Description
See Member CHANGES or CHANGESS in your MXG Source Library, or
on the homepage www.mxg.com.
Inverse chronological list of all Changes:
Changes 25.yyy thru 25.001 are contained in member CHANGES.