Professional Documents
Culture Documents
Cics Level 2
Cics Level 2
Cics Level 2
• To take advantage of this new functionality, see APAR PM02187 for CICS TS V4.1 then
apply the PTFs provided with the APAR.
• The CICS TS V4.1 information center has been updated to add three new scenarios to
the Web 2.0 scenarios section. These scenarios are:
• Creating an Atom feed from unstructured data in a temporary storage queue
• Creating an Atom feed from a file
• Creating an Atom feed from structured data in a temporary storage queue
• The topic 'How the Atom sample works' has been updated to include these additional
artifacts.
CICS TCP/IP port hangs until CICS is recycled after applying PTFs
for OMVS APAR OA29566
• You are using CICS Transaction Server for z/OS (CICS TS) and TCP/IP. Your CICS
TCP/IP port hangs and is not responding. There are no errors but all work on that
PORT ceases. You have to bring CICS down and back up to resolve the problem.
This occurs after you apply OMVS PTF UA51856 (z/OS V1.9), UA51857 (z/OS
V1.10), or UA51830 (z/OS V1.11) for APAR OA29566.
• Severity 1 OMVS APAR OA32088 has been open to address this problem. If
possible, you should not apply the PTFs for APAR OA29566 until the PTFs for
OA32088 become available. If you did APPLY one of the PTFs, REJECT it to back it
off. You will also have to IPL to remove the PTF which is in error.
New CICS TS 4.1 Abend AFDK
• AFDK
• A file control request was made against a NSR file while transaction isolation was active for the
task. Using NSR files with transaction isolation active is not supported. The TRANISO system
initialization parameter is YES and the transaction definition has ISOLATE set to YES.
• CICS Manuals at all supported releases indicate:
• VSAM nonshared resources (NSR) are not supported for transactions that use transaction
isolation. You should specify ISOLATE(NO) when you define transactions that access VSAM
files using NSR.
• NOTE: CICS OPEN APAR PM07304 will allow READONLY Files to use NSR and
Transaction Isolation.
DFHSI1596 AFTER RUNNING DFH$UMOD
• Problem: During shutdown of your region, you do not receive message DFHRM0204
There are no indoubt, commit-failed or backout-failed UOWs. This is due to an immediate
shutdown occurs when you specify SDTRAN=CESD in your CICS® system initialization
table (SIT).
• Solution: Take the steps necessary to prevent the shutdown assist program from turning
your normal shutdown into an immediate shutdown. If you want to know why your normal
shutdown is not completing 'normally' then you can code SDTRAN=NO. This will cause
CICS shutdown to hang if it can not be completed normally. Afterwards, you will be able to
review what tasks are active that are keeping CICS from shutting down. See MustGather:
Documentation for a wait or loop in CICS for details on how to get the dump if needed.
• NOTE: You will receive CICS message DFHTM1703 if CESD issues an Immediate
Shutdown
8
CICS TS 3.2 DEADLOCK AFTER TWO USER TASKS
WANT SMLOCK AND XMLOCK
• There is a deadlock between a transaction that owns the SMLOCK
and is waiting for the XMLOCK to be freed, and another that owns
XMLOCK and is waiting for SMLOCK
– CICS has attached the CJGC transaction to perform Java garbage collection
work. The XMAT attach logic requires the XMLOCK while it does its work
– Another task is issuing an EXEC CICS INQUIRE STORAGE call. DFHSMMC2
is called to perform INQUIRE_TASK_STORAGE. This holds the SMLOCK
within DFHSMMC2
– The attach of CJGC requires storage, so it waits for SMLOCK.
– The code in DFHSMMC2 needs to call XM to return transactional information,
so requires the XMLOCK
• DFHSMMC2 has been changed to release and reacquire the
SMLOCK around the calls to transaction manager for this operation
• Apply APAR PK65070 / PTF UK38082
9
USADLOCK held by one task and no signons occur
• Application is issuing EXEC CICS RETURN CHANNEL. This is causing the task's
associated channel and containers to be detached but never freemained. Each time
this command is issued, storage becomes orphaned. This eventually leads to CICS
going short on storage and issuing the message DFHSM0133
• The root problem is the EXEC CICS RETURN CHANNEL. The CICS Translator
should have generated an error against this command as the manual states that
CHANNEL is a sub-parm under TRANSID. CHANNEL can only be specified if
TRANSID has been specified as well
• Problem: Your CICS Transaction Server for z/OS (CICS TS) region issues
message DFHSM0002 severe error code X'030D' in module DFHSMGF.
You are running CA Intertest.
• If the file is OPEN ENABLED for readonly then CICS SPI returns a
value of RECOVERYSTATUS based on the RDO definition instead
of what is defined on the VSAM ICF Catalog.
• CICS should use the VSAM ICF Catalog for all inquiries of a RLS
dataset.
• You should have the WLM class for ICSF as high or higher
as CICS
• Problem: You have upgraded from CICS Transaction Server for z/OS (CICS TS)
V2.2 to CICS TS V2.3, or later, and receive message:
• Cause: The way CICS TCBs were dubbed was changed in CICS TS V2.3 and
above. Any daughter TCBs of the QR are now dubbed as a process. This
increases the number of processes and might require an increase in the
MAXPROCUSER value. For instance, "TCPIP Sockets for CICS" can use up to
300-400 TCBs at any time. These TCBs are not OPEN TCBs but rather use Unix
System Services (USS). When running under CICS TS V2.2, these were not
counted against MAXPROCUSER, but after CICS TS V2.2 they are counted
• You enter CEDA DI LI(*) on your CICS Transaction Server for z/OS
(CICS TS) V3.2 region for the DISPLAY LIST command, and you
see error message: E 'LI' is ambiguous, LIST has been assumed.
• The message displayed on the CEDA panel is:
• The reason for the change was due to the addition of the new
CEDA LIBrary command that was added to CICS TS V3.2. When
more than one command with the same prefix is added, another
letter is added to the prefix of each command.
878 or 80A abend with reason code x'0C' or x'10' when
connected to MQ
• Problem: You are running WebSphere MQ for z/OS. After upgrading to CICS
Transaction Server for z/OS (CICS TS) V3.2, you receive an abend 878 or 80A
with reason code x'0C' or x'10'. Getmains fail due to lack of below the line LSQA
storage
• Cause: A large proportion of LSQA storage had been allocated for PRHWs,
program request handler control blocks. PRHWs are used by CICS tasks to
communicate with MQ through the MQ-CICS adapter.
• In CICS releases before CICS TS 3.2, MQ allocated one PRHW for each of the
eight subtask TCBs used for communication between CICS and MQ. In CICS TS
3.2 and above, where CICS has control of MQ-CICS adapter and bridge
components, CICS attaches to MQ under the L8 Open TCB. Because of this there
is now no limit to the number of PRHWs that can be allocated. This leads to the
possibility of filling LSQA below the line.
• Problem: You are unable to deallocate the SYSIN dataset while CICS® is active. You specify
FREE=CLOSE on each SYSIN DD statement and expect the dataset to be deallocated from CICS once
the file is closed. Normally the close occurs just after CICS has read the SYSIN parameters during
initialization. However, the SYSIN datasets remain allocated to CICS until it is shutdown. This stops you
from being able to delete and reallocate the dataset while CICS is running. You are able to successfully
deallocate other datasets that specify FREE=CLOSE in the CICS JCL
• Cause: Your CICS SYSIN input consists of several concatenated PDS datasets. z/OS is causing
FREE=CLOSE to be ignored due to the dataset being a member of a concatenated group
• Solution: Use the ADYN transaction to deallocate a file from CICS while CICS is up. Prior to using
ADYN you must install group DFH$UTIL in your CICS region. This group contains the transaction
definition for ADYN and its associated program, DFH99. Here is an example of the commands that can
be used to deallocate the file "MYFILE.CICSTS13.SYSIN" and reallocate it:
• ADYN is documented to in the CICS TS information center under the dynamic allocation sample
program
Detection of OS/VS COBOL modules running
within CICS Transaction Server
Problem How do I find OS/VS COBOL modules running in my CICS® Transaction Server region?
Cause CICS Transaction Server V3.1 will not execute OS/VS COBOL application modules.
Solution CICS Transaction Server V3.1 (and any subsequent releases of CICS) will not execute OS/VS
COBOL application modules. Therefore, in order to prevent application failures within CICS TS 3.1 region, all
OS/VS COBOL modules must be recoded into a supported language (such as Enterprise COBOL). The first
step is to find the modules executing within a CICS region (which will, therefore, be CICS TS 2.3 or lower).
There are two approaches. One involving batch processing, and another within the CICS region.
The best approach is to install SupportPac CH1A. This provides an update to the CICS Load Module Scanner.
Run this with the OSVSONLY option against all load libraries used by CICS to create a listing of the OS/VS
COBOL modules. Alternatively, install some APARs. The effect of which will be to create a new message
DFHAP1301 when an OS/VS COBOL module is first used within the CICS region. However, if an OS/VS
COBOL module is not executed, the message will not be created, so detection may be incomplete.
For CICS TS 1.3, use APAR PK16285. For CICS TS 2.2 and CICS TS 2.3, use APARs PK10480 and
PK15436.
• For CMF (CICS Monitoring Facility) Performance Class data clock fields, the base unit of time is 16
microseconds. For the CPU time clock fields (i.e. USRCPUT), the amount of CPU recorded for a single
dispatch of a task is truncated down to the nearest 16 microsecond unit. So if a task used 24 microseconds
during a dispatch, CICS will add 16 microseconds to the running USRCPUT total. If the task used 31
microseconds or 17 microseconds during a dispatch, 16 microseconds will be added to the running
USRCPUT total. If the total CPU time was less than 16 microseconds, then 0 will be added to the running
USRCPUT. So for each dispatch of a task, CICS will under report the CPU used during that dispatch by an
average of 8 microseconds.
• This can become a significant amount of underreported CPU when a single task has hundreds of
thousands of dispatches. CICS has always done the CPU reporting this way.
• This effect is sometimes noticed when using CMF data to compare the CPU used by a task when its
program is defined THREADSAFE verses when its program is defined QUASIRENT.
– If the program makes hundreds of thousands of SQL calls with no interleaving non-threadsafe CICS
commands, and is defined QUASIRENT, the task will constantly switch from QR to L8 and back. Each
switch is a new dispatch and so the task will have a very high dispatch count and so the USRCPUT
time recorded will be significantly less than was actually used.
– But with the same program defined THREADSAFE, the task will not constantly switch, it will have a
low dispatch count and so the USRCPUT reported will be very close to the amount used.
• If the CPU used in DB2 and in the application is very small, it is possible for the USRCPUT reported for the
QUASIRENT case to be less than the USRCPUT reported for the THREADSAFE case.
• How do you find CICS product documentation for CICS Transaction Server for z/OS
(CICS TS) and the CICS tools?
• The best place to find CICS product documentation is to use the CICS TS
information centers. Documentation for most of the CICS Tools is also included in
the CICS TS information centers. Each release of CICS has its own information
center that you can view on the www.ibm.com Web site or install on a workstation or
server.
• Here is a link to an item that has all the Information Centers online, and directions to
download to a workstation:
– http://www-
01.ibm.com/support/docview.wss?rs=1083&context=SSGMGV&uid=swg21200934
The CICS catalog manager example application
• The base application has a 3270 user interface, but the application comes
with Web service support, which is designed to illustrate how you can extend
an existing application into the Web services environment. Find out more at:
– http://publib.boulder.ibm.com/infocenter/cicsts/v3r2/index.jsp?topic=/com.ibm.cics.ts.webservic
es.doc/sampleapp/topics/dfhxa_t100.html
Fix lists for CICS TS
• Fix list documents summarize all of the APARs and PTFs for a
particular product Version
• Fixes by version document 7008833 contains a link to the
available Fix list documents
• http://www.ibm.com/software/htp/cics/tserver/support/
• http://www.ibm.com/software/htp/cics/tserver/support/
• Flashes for all CICS New Function (NF) APARs are posted to
CICS Support page and included in MySupport email
• CICS item 1238275 has a listing of links to all New Function
APARs for CICS Transaction Server and all CICS Tools
• http://www.ibm.com/software/htp/cics/tserver/support/
– http://www.ibm.com/software/htp/cics/tserver/support/
You want to know when the end of service is for any IBM Software release.
You can find the End of Service dates by navigating to the following URL:
http://www.ibm.com/software/info/supportlifecycle/list/c.html
CICS Transaction Server for z/OS V2.2 5697-E93 25 Jan 2002 30 Apr 2008
CICS Transaction Server for z/OS V2.3 5697-E93 19 Dec 2003 30 Sep 2009
CICS Transaction Server for z/OS V3.1 5655-M15 25 Mar 2005
CICS Transaction Server for z/OS V3.2 5655-M15 29 Jun 2007
Order processing for CICS products and
maintenance
You would like to order a CICS® product or maintenance. You want to know
what options are available for ordering the product, individual PTFs, and
cumulative maintenance for CICS or any of the CICS tools
See the following document for all product or maintenance ordering
concerns
http://www-1.ibm.com/support/docview.wss?fdoc=aimcics&rs=1083&context=SSGMGV&uid=swg21049360
Must Gather Documentation
MustGather documents aid in problem determination and save time resolving problem management
records (PMRs). These documents are located on the CICS® Web site and contain instructions
about what documentation to gather for specific problems.
Collecting MustGather data early, even before opening a PMR, helps IBM® Support quickly determine if:
You can find the MustGather – Read first document on the CICS Home Page or go directly to:
http://www-1.ibm.com/support/docview.wss?rs=1083&uid=swg21208053
PE and HIPER APARs
• Problem: You receive a program check that has an offset of X'FFFF'. For example,
DFHAP0001 An abend (code 0C4/AKEA) has occurred at offset X'FFFF' in module
DFHAPDM. Since the X'FFFF' indicates that CICS® was unable establish the location of
the program check, you would like to locate the PSW and registers in your CICS system
dump.
• Solution: Review the following CICS Document for debugging information:
http://www-1.ibm.com/support/docview.wss?rs=1083&context=SSGMGV&uid=swg21176381
CICS support news is now available on Twitter
– http://www-
01.ibm.com/support/docview.wss?rs=1083&context=SSGMGV&uid=swg21384915
Additional Product Resources
• Tracking specific CICS APARs or CICS APARs by component id:
http://www.ibm.com/support/docview.wss?uid=swg21422149