DFHFC0152 date time applid termid tranid An attempt to retain locks for data set within unit of work X'uowid' failed. VSAM return code X'rrrr' reason code X'cccc'.
Explanation
Unit of work uowid for transaction tranid has failed backout for one of its data sets. CICS has attempted to convert all the RLS locks owned by this unit of work that are associated with the failing data set into retained locks. This attempt has failed because VSAM has detected an error.
The IDARETLK response is rrrr and the reason code is cccc.
termid identifies the terminal running this transaction.
This message is followed by message DFHFC0312 which identifies the failing data set.
System action
CICS continues shunting this unit of work. Some locks may remain as active locks (which cause other transactions to wait until their timeout value is reached) rather than as retained locks (which cause other transactions to encounter LOCKED responses).
Message DFHME0116, which contains the symptom string for this problem, is produced.
User response
Use the VSAM codes to determine the cause of the problem. For the meaning of the VSAM codes, see z/OS DFSMS Macro Instructions for Data Sets.
The most likely reason for the failure to convert locks into retained locks is that the SMSVSAM server was not available. It is also possible to get a failure because the specified logical unit of work ID does not exist for the subsystem (that is, the unit of work does not hold any locks) during lost locks recovery, or after a CICS restart which specified OFFSITE=YES as a system initialization override. If you are performing RLS lost locks recovery, message DFHFC0555 will have been issued when lost locks recovery started; if you are performing RLS offsite recovery, message DFHFC0574 will have been issued during file control initialization. Other VSAM codes may indicate a more severe error.
Normally no other action is necessary. When the condition that caused the backout failure has been resolved, the backout of this unit of work is retried. If the attempt to retry the backout succeeds, all locks are released.
Message DFHFC4701 specifies the cause of the backout failure. The most common cause of backout failures is a hardware problem causing I/O errors. In this case the data set needs to be restored and forward recovered. If CICSVR (or a functionally equivalent product) is used to perform forward recovery, and the data set was being accessed in RLS mode, units of work that have failed backout for this data set are retried automatically. If the data set was quiesced, you need to unquiesce it to allow the backout to succeed. When the data set is unquiesced, CICS automatically retries the backout.
Backouts may also be retried using
CEMT SET DSNAME RETRY
or
EXEC CICS SET DSNAME(dsname) ACTION(RETRY)
.
The only problem that you may encounter is that some locks may have been left as active locks. This can cause a slow response (and eventual failures) from transactions that wait for these locks and have to wait for their full timeout interval.
In this case, consider releasing all locks held against this data set using the CEMT SET DSNAME RESETLOCKS command. This command should only be considered in extreme cases because it discards both the retained locks held by this CICS system against the named data set and all associated log records. The consequence is that the corresponding backout operations are never performed and data integrity is lost.
If you cannot resolve the problem or the problem recurs, there may be a more severe error. In this case, you will need assistance from IBM. See IBM problem support for guidance on how to proceed.
Module
XMEOUT parameters/Message inserts
- date
- time
- applid
- termid
- tranid
- X'uowid'
- X'rrrr'
- X'cccc'
Destination
CSFL and Console