The CICS transaction manager restart task could not complete because a necessary step failed. The task has done some essential recovery operations and abnormally terminated itself with code AKCB.
CICS writes a transaction dump for the transaction manager restart task.
CICS sends three messages to the console, one to identify the error detected by the transaction manager restart task, one to say that the task has failed, and one that gives you the option of cancelling CICS or letting it continue. Depending on the nature of the original error, you may see messages from some other system component (for example, an access method).
Use the messages and dumps to find out the cause of the failure.
DFHKCRP
The CICS transaction manager has abended the transaction because the purge threshold of its TRANCLASS has been reached. This is specified by the PURGETHRESH parameter on CEDA DEFINE TRANCLASS. See the CICS® Resource Definition Guide manual for more details of this parameter.
The transaction is abended and messages DFHAC2004 and DFHAC2236 are issued. The transaction dump is suppressed for this abend code.
Resubmit the transaction. The cause of the abend may be a temporary stress condition in the system.
If the problem persists, determine why the TRANCLASS purge threshold has been reached. Ensure that PURGETHRESH has been specified correctly. Also, ensure that the MAXACTIVE value of the TRANCLASS has not been set too low. Transactions attached after the MAXACTIVE limit has been reached are immediately queued subject to the PURGETHRESH limit.
If PURGETHRESH and MAXACTIVE are set correctly, look for a more general problem which has caused a decrease in the capacity of the system to execute transactions in the TRANCLASS. The decrease might, for example, be caused by a connected CICS region which processes requests for transactions in the TRANCLASS, if this connected region has slowed down.
Examine all resources (files, links, storage, and so on) used by the transactions in the TRANCLASS which is reaching the purge threshold and determine why the capacity of the system is reduced.
DFHXMAT, DFHXMCL
While CICS transaction manager was recording changes to a transaction or profile definition, a write to the system log failed.
CICS terminates the transaction with a transaction dump.
Use the dumps to find out why the write to the log failed.
DFHKCQ
While CICS transaction manager was recording changes to a profile definition, a write to the catalog failed.
CICS terminates the transaction with a transaction dump.
Use the dumps to find out why the write to the catalog failed.
DFHKCQ
Transaction manager has received an invalid request code. The last AP F000 trace entry before the program control program (PCP) ABEND TRACE entry (TRACE ID ‘F2’, request code X'6000') will contain the invalid transaction manager request code in the fifth byte of the first section of the trace.
The transaction is abnormally terminated with a CICS transaction dump.
Determine the cause of the invalid request code and correct the problem.
DFHXCP
A deadlock timeout condition has been detected. This condition may occur within a transaction that specifies DTIMOUT to be nonzero on its installed transaction definition. Deadlock timeout occurs when a transaction has been waiting or has been suspended for longer than the time specified in DTIMOUT.
The abend may be driven by a variety of internal CICS events, for example:
The abend can also occur if CICS stops running for a time, for example while an sdump is taken. This is because deadlock timeout is based on total elapsed time, and not just the time CICS is executing.
Analysis: The transaction receiving the AKCS abend must have been suspended or must be waiting for a reason such as a short on storage, enqueued on a lock, a short on temporary storage, a suspend after RETRIEVE WAIT, a suspend after ALLOCATE, or an implicit ALLOCATE within function shipping or terminal sharing support. If none of these apply, the trace might reveal some event that has caused CICS to stop running for a time.
The transaction is abnormally terminated. A dump
is not provided unless the dump table entry for transaction dump AKCS indicates
that one should be taken.
The transaction should be reexecuted, and the situation causing the SUSPEND to occur may clear itself.
The AKCS abend is to be expected occasionally, unless DTIMOUT is set to zero. No special action is necessary.
DFHXCP
A terminal read-time-out condition has been detected. The transaction has been waiting for a terminal input message for an interval longer than specified in the RTIMOUT value for that transaction.
If an EXEC CICS HANDLE ABEND has been issued for this task, the read that was timed-out is still outstanding. To cancel this read you should issue an EXEC CICS ABEND at the end of the user exit routine so that CICS can clean up the terminal’s TCTTE.
The transaction is abnormally terminated. A transaction dump is not provided.
This abend is a normal one. Coding RTIMOUT in the PROFILE entry asks for the task to be abnormally terminated if the terminal does not send input within the specified time.
DFHXCP
A bad return code was passed as a result of the resume of a task suspended by ICP.
The transaction is terminated with a dump.
Check the response from the resume in the trace to determine the cause of the error.
DFHALP
An attempt has been made to run the CICS internal task CSSY as a user transaction.
CICS terminates the task with a transaction dump.
Investigate why the attempt was made to run CSSY as a user transaction.
DFHAPATT
A DFHKC WAIT request was issued when the ECB was already marked as waiting.
There is a probable user error. The transaction is abnormally terminated.
Correct the program that issued the request.
DFHXCP
A bad response has been received from a dispatcher (DS) domain call.
The transaction is abnormally terminated with a transaction dump and a trace entry.
Examine the trace entry for further information.
DFHXCP
The task has been purged, probably due to operator action such as a CEMT TASK PURGE command. The task might also have been purged as a result of CICS issuing a purge request.
The transaction is abnormally terminated with a transaction dump.
Use the transaction dump to determine why the task was purged. In particular, if the purge was operator initiated, the dump should be useful in determining why this task needed to be explicitly purged.
DFHXCP, DFHXMAT, DFHXMCL, DFHXMIQ, DFHXMTA
DFHKC RESUME should always be preceded by DFHKC SUSPEND. If this protocol is violated then the transaction is abnormally terminated with abend code AKC6.
Transaction is abnormally terminated with abend code AKC6.
Examine the trace entry for further information.
DFHXCP
A bad response has been received from a call to the kernel (KE) domain during the processing of a task purge request.
The transaction is abended with a transaction dump.
Examine the dump and any exception trace entries for further information.
DFHXCP
An incorrect response has been received from a call to the enqueue (NQ) domain during the processing of a DFHKC TYPE=ENQ or a DFHKC TYPE=DEQ request.
The transaction is abnormally terminated with a transaction dump.
Examine the dump and any exception trace entries for further information. Since the DFHKC service is only used for internal enqueues, this abend indicates an error in CICS. You need further assistance from IBM to resolve this problem. See CICS Problem Determination Guide for guidance on how to proceed.
DFHXCP
A program check has been detected by the kernel (KE) domain.
If an application is in control, the ASRA abend is presented to the application. Otherwise, the functional recovery routine of the CICS module in control at the time is given control. This recovery routine produces suitable diagnostics and may terminate CICS.
Look at the kernel domain section of the system dump to determine where the program check has occurred.
DFHKESTX
An operating system abend has been detected by the kernel (KE) domain.
If an application is in control, the ASRB abend is presented to the application. Otherwise, the functional recovery routine of the CICS module in control at the time is given control. This recovery routine produces suitable diagnostics and may terminate CICS.
Check the console for any MVS messages that may have caused this abend.
Look at the kernel domain section of the system dump to determine where the abend has occurred.
DFHKESTX
The kernel (KE) domain has detected runaway.
If an application is in control, the AICA abend is presented to the application. Otherwise, the functional recovery routine of the CICS module in control at the time is given control. This recovery routine produces suitable diagnostics and may terminate CICS.
Look at the kernel domain section of the system dump to determine where the runaway has occurred.
DFHKESTX, DFHKERRU
The kernel (KE) domain has been requested to initiate abend processing as a result of a deferred abend request.
Abend processing starts for the task that is subject to the deferred abend request.
The task is not abended with AKED but by an abend code specified by the requestor of the deferred abend. See the description of this abend for further guidance.
DFHKEEDA
The kernel (KE) domain has detected an error while processing a domain call. The error may have been caused by a domain gate that was not yet active during initialization
If an application is in control, the transaction terminates with a system dump. Otherwise, the functional recovery routine of the CICS module in control at the time is given control. This recovery routine produces suitable diagnostics and may terminate CICS.
See any related messages from the kernel domain.
Look at the kernel domain section of the system dump to determine where the error has occurred. Check that a call has not been made to a domain gate that has not yet been made active. Check that the caller has NOT specified KERNERROR(YES).
If the abend occurs during CICS system initialization, ensure
that the utility (DFHCCUTL) used to initialize the local catalog (DFHLCD)
is at the correct level. A sample job is provided in SDFHINST(DFHDEFDS).
DFHKERKE
The kernel (KE) domain issued an SM GETMAIN for kernel stack
storage, but the GETMAIN request failed.
If an application is in control, the transaction terminates with a system dump. Otherwise, the functional recovery routine of the CICS module in control at the time is given control. This recovery routine produces suitable diagnostics and may terminate CICS.
Look at the kernel domain section of the system dump to determine why sufficient storage was not available.
If the short-on-storage condition persists, consider increasing
the size limit of the CICS DSA. You can vary the DSA dynamically using the
DSALIM parameter on the CEMT master terminal command.
DFHKESGM
The transaction was purged while running outside the control of CICS.
CICS terminates the transaction abnormally.
The EXEC CICS HANDLE ABEND command can not handle this abend.
Investigate the reason why the transaction was purged.
Possible reasons are
DFHKESTX
The kernel (KE) domain has detected runaway while the transaction is running outside the control of CICS.
If an application is in control, the AICA abend is presented to the application. Otherwise, the functional recovery routine of the CICS module which was last in control at the time of runaway detection is given control. This recovery routine produces suitable diagnostics and may terminate CICS.
The EXEC CICS HANDLE ABEND command can not handle this abend.
See the CICS Problem Determination Guide for guidance on dealing with loops.
DFHKESTX
A user attach has failed because there are insufficient kernel tasks available. This indicates an internal logic error.
Message DFHKE0001 is issued and a system dump is taken. The attach of the user transaction fails.
You need further assistance from IBM to resolve this problem. See CICS Problem Determination Guide for guidance on how to proceed.
DFHKETA
A kill request has been actioned when a transaction was not protected from purge or force purge. The transaction is either in a dispatcher suspend, the deferred abend processor is in control, the application is calling CICS, CICS is returning to the application, or the application is outside of the control of CICS.
If an application is in control, the abend is presented to the application. Otherwise, the functional recovery module of the CICS module in control at the time is given control. This recovery routine produces diagnostics and might terminate CICS. This abend code cannot be handled by the application.
Notify the system programmer to determine why the task has been killed.
DFHDSSR, DFHKEEDA, DFHEIP, DFHKEDS
A kill request has been actioned when a transaction was not protected from forcepurge but was protected from purge. The transaction is either in a dispatcher suspend, the deferred abend processor is in control, the application is calling CICS, CICS is returning to the application, or the application is outside of the control of CICS.
If an application is in control, the abend is presented to the application. Otherwise, the functional recovery module of the CICS module in control at the time is given control. This recovery routine produces diagnostics and might terminate CICS. This abend code cannot be handled by the application.
Notify the system programmer to determine why the task has been killed.
DFHDSSR, DFHKEEDA, DFHEIP, DFHKEDS
A kill request has been actioned when a transaction was protected from force purge. The transaction is either in a dispatcher suspend, the deferred abend processor is in control, the application is calling CICS, CICS is returning to the application, or the application is outside of the control of CICS.
If an application is in control, the abend is presented to the application. Otherwise, the functional recovery module of the CICS module in control at the time is given control. This recovery routine produces diagnostics and might terminate CICS. This abend code cannot be handled by the application.
Notify the system programmer to determine why the task has been killed.
DFHDSSR, DFHKEEDA, DFHEIP, DFHKEDS
A CEKL purge has been requested. Abend processing has started for the task that is subject to the deferred abend request.
The task is abended with abend code AKKD.
Notify the system programmer to determine why the task has been purged.
DFHKEEDA
A CEKL force purge has been requested. Abend processing has started for the task that is subject to the deferred abend request.
The task is abended with abend code AKKE.
Notify the system programmer to determine why the task has been purged.
DFHKEEDA
The kernel (KE) domain has detected a kill request from the runaway exit program. The task was not protected from runaway when the kill request was actioned.
If an application is in control, the abend is presented to the application. Otherwise, the functional recovery module of the CICS module in control at the time is given control. This recovery routine produces diagnostics and might terminate CICS. This abend code cannot be handled by the application.
Notify the system programmer to determine why the task has been killed.
DFHKESTX, DFHKERRU, DFHKEKIL
The kernel (KE) domain has detected a kill request from the runaway exit program. The task was protected from runaway when the request was actioned.
If an application is in control, the abend is presented to the application. Otherwise, the functional recovery module of the CICS module in control at the time is given control. This recovery routine produces diagnostics and might terminate CICS. This abend code cannot be handled by the application.
Notify the system programmer to determine why the task has been killed.
DFHKESTX, DFHKERRU, DFHKEKIL
A user has generated an addition to the keyword table, but code has not been added to process this keyword.
The transaction is abnormally terminated and a dump is taken.
Add code to process the keyword.
DFH99KC