1. Attempt to delete JECL continuation statement in SLI-member (by using sequence number) results garbage on the screen (1R33D message), or 1Q2CI Segment translation exception and POWER ABEND followed by dump creation.
2. Specification of 'RBM=n' operand in the '* $$ PUN' statement can result 1QZ0I (RC=0045) error message which follows dump creation.

1. When displaying the CRE queue, duplicate entries may be shown.
2. PDISPLAY Q takes to much time to show extended queue information.
3. PDISPLAYqueue,TAPE=cuu,OUT=CON shows header line for each entry.

* DO NOT INSTALL (PE) *
Please install DY46552.

1. When VSE/POWER's own 'TI00' Timer Exit routine is disabled by a user exit coded 'SETIME r1,r2' with ECB, or by replacement with OEM own Timer Exit routine, then the Time Event scheduling functions will come to a halt.
2. VSE/POWER Networking entries that are transmitted via RSCS or AIX nodes will produce a "transmission acceptance" message even though the entries did not specify the NTFY= operand.
3. A compile error occurred for the module IPW$$MM due to an error in the message 1QE7I.

* DO NOT INSTALL (PE) *
Please install DY46552.

1. For PALTER LST,H,USER=Rnnn with more than 500 entries in disp. or non disp. chain, command processor loops on writing to queue file - while DMB is unlocked periodically (fortunately).
2. If data is included via * $$ SLI and * $$ DATA statements and a /*$SLI statement is used to end data to be included, results are unpredictable.
3. If in a * $$ LST/PUN statement the PRI= operand is specified without a value, the priority value for the created queue entries shows up as a blank or ','.
4. Trace message 1RT2I ends sometimes with (ACT)NT) wherey the 'NT)' are meaningless.

1. When during profile processing of dyn. part. 'Xn' the partition is cancelled by operator, then 1Q6FA RC=0002 is issued and class 'X' is disabled for job selection.
2. For CRE-queue browsing and Restart to Active the q-rec field QRC2 is used at length of 16 bytes, but cleared in implicite QRC2 length of 32 bytes.

1. When data is received from JES2 (or RSCS) via a SNA link, ending blanks may be lost.
2. When BTRNC=NO has been specified in the PSTART PNET,node command and the hardcopy file wraps around, no information is available whether it has been specified or not.
3. Operating systems other than z/VSE do not support multihoming (as done by z/VSE offering the IPEXTRAD parameter in the PNODE macro) for TCP links. Connections can not be established, message 1RT5I is issued.

1. Palter command has to commit the changes to q-file on disk. When bulk alteration is requested, this may take seconds while DMB is locked - disallowing tasks to gain control that need the DMB resource.
2. The Dsect MAPPOWJB for GETFLD FIELD=POWJOB provides Jobname and Jobnumber of the active VSE/POWER job, but the Jobclass is missing.
3. When m/d/y pattern has been changed during VSE/POWER warm start (e.g. during FSU), CRDATE= specifications are interpreted in the current e.g. d/m/y pattern . Hence entries created before the warm start can not be manipulated by CRDATE= to isolate them.
4. The SORT=NEW/OLD display presents display lines in standard format. The sort criteria 'creation date + time' are hidden in the 2nd/3rd line of a FULL=YES display (see D=... and T=....).

1. PNET usage:
   1. The JES copy count, for example COPIES=(164,(24,...,17)) of 164 in NDHGDSCT is destroyed when sending 3800 output through an intermediate VSE/POWER PNET node (replaced by number of copy groups).
   2. Job/output data originating from non VSE/POWER (e.g. JES, RSCS) nodes are transmitted with trailing blanks but these are removed forever when data is spooled by VSE/POWER as final or intermediate node.
   3. If job/output data originating from non VSE/POWER (e.g. JES, RSCS) nodes contain trailing blanks and are forwarded to another JES, RSCS, the record length is different from the original record length.
   4. If output data originating from non VSE/POWER (e.g. JES, RSCS) nodes consist of various parts with different record lengths and are forwarded to another JES, RSCS, the record lengths are different form the original record lengths.
2. The VM/CMS user of the VSECMD command does not receive a VSE/POWER response message for the PACCOUNT command.
3. PDISPLAY command:
   1. The PDISPLAY ..,LIMIT=NN value allows only a maximum of 99, especially for PDISPLAY BIGGEST. This is too little for big queues.
   2. The PDISPLAY RDR/LST/PUN/XMT cmd presents q-entries sorted by class but not by creation-date, i.e. not by 'age' of the entries.
4. VSE/POWER recovery startup may fail 1Q2CI CC=21, ILLEGAL SVC, PHASE=IPW$$NU, TASK='P PS', even more in VSE/POWER shared environment.

1. When PNET receiver task stalls due to 1Q38A NO DASD AVAILABLE, then the PDRAIN command is not effective, but receiver remains in wait-for-Data-File-DBLK.
2. When SAS-PUT-Job/Out task stalls due to 1Q38A NO DASD AVAILABLE, then the PSTOP SAS command is not effective, but SAS-PUT task remains in wait-for-Data-File-DBLK.

1. Unprintable characters in HDR1-part of message 1QB9A may abend VSE/POWER with message 1Q2CI with SEGM. TRANSL. EXCEPTION in phase IPW$$MX.
2. The error handling of PNET SSL for the GSKSSOCINIT call will ignore an I/O error (RC=-00000010 or the equate FRCN0010). Also an ERRNO=00001102 (EWOULDBLOCK) handling leads to an IDUMP in the cases of
   • GSKSSOCREAD
   • GSKSSOCWRITE
   • GSKSSOCCLOSE
   • GSKSSOCRESET
3. PACCOUNT PUN creates IDUMP indicated by msg 1Q2JI in IPW$$02 when data file is full and no DBLKGPs are available.

1. DUETIME operand does not support to calculate a runtime dependant on time when job entered system.
2. The record count of a list output received from a JES2/3 is different than the record count displayed by JES2/3.

1. 1. When all DBLKGP cushion elements are used and FREE DBLK GROUP count is 0, the next request for a queue display via Spool Access Support, e.g. by IUI, fails with 1QZ0I RC=0025 in module IPW$$DQ.
2. 2. FULL queue display does not show creation time.

1. POFFLOAD SELECT for labelled tape will bring msg 1Q77I RC=000c if it searches beyond the first tape tape volume and a continued queue entry at the end of a searched volume is not eligible for selection.
2. For dynamic partitions a default * $$ LST/PUN output class cannot be selected as with static partitions, namely (A-Z or 0-9). Instead the dynamic part. default is 'A' or the execution class itself, provided the SET DYNOUTCL=DYNCL stmt. is active.
3. POFFLOAD LOAD/SELECT transports RDR entries of other nodes back to the own RDR queue, but when executed, their output may end up in the XMT queue instead of LST queue.

1. Any other VSE/POWER subtask than Async. Service Subtask (-AS), namely DS/LS/SD/SN/S1/TD/TI should take IDUMP when being cancelled under all circumstances - Cancel Code '09' (LIOCS) should not make an exception.
2. When long running CICS/VTAM/TCP-IP jobs with DISP=D/K are replaced by new version, the old job in DISP=* cannot be PALTER'ed to DISP=D so that it vanishes at system shutdown and allows the new version to be released at system startup. Instead the old version is released once more then.

1. Destroyed VSE/POWER class anchors or chain pointers to the free-q-rec chain result in display of quasi empty 'free' queue records, which may be many.
2. When being used to 'number-of-spooled-pages' for LST queue entries and altering them to the XMT queue, then no more pages but LINES are presented, what may be confusing.
3. 1) When remote node sends FCS bytes not equal to X'8FCF', a loop of IDUMP macro-calls and message 1RTLI may occur.
   2) When remote node sends incorrect BSC frames, the connection hangs and message 1RC1I is issued.

1. Sending small job from RSCS to PNET may cause 'wait-a-bit' re- turned to RSCS resulting in 20 sec delay.
2. MSG 1Q07I INVALID LOGICAL UNIT names 'filename' but not logical unit number when Q/D/A-file have incorrect DLBL/EXTENT/ASSGN.

1. For OUTEXIT (enabled for e.g. LPR by SET OUTEXIT=SAS) which inserts records and where SAS-GET is terminated prematurely by QUIT/CLOSE - the processed q-entry may remain in '*' state without any processing task active.
2. After APAR DY46114 the 'B'rowse column of XMT queue display shows 'D' instead of ' ' when no browser is active.

1. PUT-OPEN-APPEND/RESTART position at OPEN time to the logical-end/target by stepping fwd from begin via the SER records. Hence an existing big appendable output may cost many I/O's for positioning to 'end'.
2. Reloading dispatchable entries via OFFLOAD LOAD/SELECT may cost effort to avoid immediate execution or processing of these entries.
3. For networking any transmitted entry is recorded on console by msg 1RA0I - but for local LST/PUN tasks (especially with 'VM'option) there exists no automatic message for a processed queue entry. Such msg would help in revision.

1. PRELEASE queue,ALL does not touch DISP=L entries but makes DISP=H entries dispatchable (D) without warning/confirmation, whether this mighty function is really wanted on all classes of the named queue.
2. Since DY42661 control of local UR-devices by PSTART/PSTOP/PGO is allowed by Spool Access Support (SAS) users. However 'PDISPLAY M' is not allowed to find out, which task is e.g. 1Q40A bound and needs a 'PGO'.
3. Attempt to change the local node name fails with 1RE8I - but without any ACTIVE SYSID (at switch from non-shr (abn. terminated before) system to a shared system). e.g. 1Q40A bound and needs a 'PGO'.
4. When VSE/POWER queue records have been overwritten in storage or on disk, then VSE/POWER warm start recovery may be mislead by seemingly correct indications in the queue records and may overlay lowcore.

1. For 'write once' tapes/cartridges as 9346 or 3592 - when EOV hit - VSE/ POWER cannot backout the (last) partially written Q-entry by POFFLOAD BACKUP/PICKUP/SAVE, and cannot terminate the last-but-one by 2 TM's.
2. Since DY46067 the 'D A' cmd presents display of SEP/NOSEP/... for LST/PUN tasks. For exec. processor tasks however an additional comma is placed before jobname or 'INACTIVE'.
3. When centr. op. PDISPLAY cmd creates many display lines, then console router buffers may be exhausted & WTO fails & cmd is terminated prematurely.
4. Acc. to A&O description for PDISPLAY STATUS, the 'USED' DBLKGPs comprise 'in-creation','queued','in-deletion' and 'lost-due-err'. This may be incomplete while a big Q-entry is finally deleted from DEL-queue while its DBLKGPs do not yet appear as 'FREE' ones.
5. D STATUS presents shared time intervals when Q-file runs shared, but does not inform about account file usage (also shared or non shared).

1. Extending the data file fails with message 1QD3A DATA FILE EXTENSION FAILED FOR EXTENT NO. nnn, RC=0005 when an existing extent of the data file does not start on track boundary for Data File on CKD/ECKD.
2. Extending the data file fails with message 4783I INVALID LOGICAL UNIT IJDFILE SYSxyz=cuu volid when the new data file extent is defined with a different device geometry but the same device type. For example a 3380 and a 3390 emulated on a 2105 have the same device type x'6E' (meaning ECKD) but a different track size.

1. When JOB segments LST/PUN output in creation using VSE/POWER macro IPWSEGM or SEGMENT, a SAS GET BROWSE request to the new increation queue entry fails with RC=0004, FDBK=0001 and FDB2=000C. IUI dialogue "Manage Batch Queues" 6 = In-Creation Queue fails to access such a segment and issues
   IESV0089I FOLLOWING MESSAGE(S) FROM PROGRAM 'IESBQP2 ', OFFSET X'6D0A'. IESB0012I UNABLE TO CONTINUE - DIAGNOSTIC INFORMATION SAVED.
2. When SET statement DLSEPSAS=FORCE is used IUI "Manage Batch Queues" dialogue fails for each queue type.
   IESV0089I FOLLOWING MESSAGE(S) FROM PROGRAM 'IESBQUP ', OFFSET X'2748'.
   IESB0144I $$SSSSSSSSSS PPPPPPPPPPP LL IESB0144I PREVIOUS MESSAGE WAS RETURNED BY A SUBSYSTEM.

When a non shared VSE/POWER startup phase addresses a shared VSE/POWER queue/data file for warm start commented by msg 1QB3I then msg 1QB3D asks, if another CPU still operates on the spool files. At incorrect reply the sharing and the non shared system operate acc. to different rules and destroy the free DBLKGP subchains. This leads to many drastic errors, e.g. 1QZ0I RC=14 or 1QFAA & 1QH3I.

1. Inadvertently entered 'L LST,ALL' may address and erase thousands of entries without requesting operator confirmation. This is too dangerous.
2. Failing XPCC application may loop on CONNECT'ing to SYSPWR hence consuming all VSE/POWER part. Getvis & SETPFIX storage by created SAS tasks.

1. Influencing Spool Access (SAS) GET tasks via SET DLSEP to support and customize single sheet printing does not become active for SAS tasks although stated in A&O on page 181.
2. PSTART TASKTR with FULL option verifies POWER's TCB and BCW chain at every POWER Dispatch cycle - but not for (hundreds) of other partition dispatch events, that could destroy these chains long before POWER can identify the destruction and catch the culprit.

1. After PDISPLAY CRE VSE/POWER may terminate due to SEGMENT Translation Exception as shown in message 1Q2CI
2. IDUMP may not contain SVA-31, although it was requested by TASKTR with option FULL
3. TCP/IP nodes may be disconnected and restarted (messages 1RTWI,1RB0I, 1R03I, 1RB8I), if no data is exchanged between the nodes.
4. Warm start recovery may issue 1QZ0I RC=0013, although all is correct
5. The PSTART LST/PUN options 'SEP/..' are not reflected in D A as the VM option

1. After VTAM abnormal termination and restart of VTAM, RJE,SNA can not be restarted but returns
   1V=2I RC=0004.
2. Attempt to stop VSE/POWER's internal DBLK Group tracing should be rejected by new message 1QAGI:
   1QAGI 'PSTOP DBLKTR' OBSOLETE - NO LONGER NEEDED

1. When a totally filled POWER Account file resides on CKD and is saved to a equally sized CKD SD file by PACCOUNT DISK,dlblname the saving task may fail with message
   0P22I P TRACK FMT SYSxxx=cuu
   The Account file can not be saved thus the system is locked until it is saved to tape or deleted.
2. When TASKTR been Pstarted with FULL to verify the TCB/BCW chains at each VSE/POWER dispatch cycle, POWER may idle for seconds, where other programs might have destroyed POWER's critical storage, what can no longer be re- constructed from the already overflown DEBUG area(s).

1. D LST,,3 command (actually invalid) accepted as D LST, resulting in bulk amount of (CICS) console messages. When console router buffers exhausted POWER loops on wait-1-sec and retry WTO request, while spooling resource DMB remains locked, because display cmd follows a class chain.
2. When SAS spooling output indicates SPLORCFM=ASA, but data records contain MCC op-codes, POWER does NO verification (as documented in Appl. Progr.). When later SAS GET retrieves this q-entry with option SPLGOPT2.SPLGO2AC (convert ASA-->MCC), default action (skip 1 and write immed) is taken for any MCC-code ( = unknown ASA).

1. Implicit TREADY COND=CANCEL for LIBR subtask with CC='FE' does no longer lead to correct DETACH of subtask. Hence POWER task waits eternally for DETACH completion before re-attaching a new LIBR subtask - while ASWE remains locked and all SLI of other partitions grinds to a halt.
2. If youuseIPWSEGMKEEP=YES,JECL= .. to overwriteselected attributes of 1st output segment, and if you want to reestablish the overwritten attributes inasubsequentIPWSEGMKEEP=YES,JECL=... request, then you can do it only for attributes re-turned by IPW$MXD at offset X'70'. However e.g. target node/user info is currently not reflected by the IPW$MXD parm. list.

1. When big q-entry or many normal sized are deleted by one command, 'free-writing' of returned DBLKGPs may be lengthy. For that period spooling requests for a new DBLKGP have to wait and command task will not return to requestor.
2. When system crashes while dynamic class executes (one) JOB-P and syst. restarts with SET NORUN=YES, then jobs been in exec. are set DISP=X (also JOB-P), but only static part. enter 1Q36I // PAUSE mode (for op. re-arrangement of exec. sequence), while dyn. class(es) selects the next dispatchable job for execution.

When the queue file is placed beyond track 32768, VSE/POWER issues message 1QE8A erroneously. Also queue file re-allocation is no longer possible, when the queue file is placed beyond track 32768. It will fail with message 1QE2A RC=0005

When the account file is full, a PACCOUNT DEL/DISK/TAPE must be started, to free the account file. When PACCOUNT task starts, the TCB is located in PFIXed storage cushion. If no PFIXed storage is available, any further request for PFIXed storage issued by the PACCOUNT task fails and the task waits forever, because other tasks can't release storage until they can write their accounting records.

1. Segmentation of 3800 spooled output at PDUMP request can currently be avoided be SET PDUMP=NOSEGMNT. However a program check or cancel dump still creates segmented job output in the list queue.
2. When POWER abn. terminates while in partition A job A1 runs with with 'NORUN=IGN' and another disp. job A2 exists in same class, and when POWER recovery creates a job B1 with DISP=X (due SET NORUN=YES), then job A1 re-starts execution as desired, but subsequent job A2 enters a loop on msg 1R30I.

1. When a VSE/POWER command is entered with an incorrect queue class (PSTART, POFFLOAD or PALTER), then some key-board symbols could cause an incorrect message 1R54I contents, or a program check and an IDUMP.
2. When using the POFFLOAD LOAD command and the POFFLOAD input tape contained incorrect queue entries, then an IDUMP could occur for each entry, causing the IDUMP file to quickly be filled.

When a continued * $$ LST statement in an SLI member is updated, following error message appears:
1R33D CORRECT FULL STATEMENT
If for example the following SLI-member SLIMEMB.M01 is used with the sequence numbers in column 73-80:
* $$ LST DISP=H, ... M0100001
* $$ CLASS=A, ... M0100002
* $$ DEST=(,HUS), ... M0100003
* $$ FNO=HUHU, ... M0100004
* $$ PRI=4 ... M0100005
and the job stream contains following statement:
* $$ SLI=SLIMEMB.01
* $$ LST DISP=L, ... -$SLIR001
* $$ CLASS=U, ... -$SLIR002
* $$ PRI=8 ... $SLIR005
the following error messages are issued
* $H CLASS=A, ..... M0100002
1Q49I INVALID DELIMITER , COL=00072..
1R33D CORRECT FULL STATEMENT BG
If SLIMEMB.M01 contains a non-blank character in continuation column 72

• either the following messages are issued:
  BG-0000 (blank line)
  BG-0000 1S01D INVALID STATEMENT.
  and thereafter the statements from the jobstream are displayed:
  BG 0000 * $$ CLASS=U,
  BG 0000 * $$ PRI=8
  and the list queue entry has the characteristics of the first * $$ LST statement of the jobstream and the second and subsequent statements of the SLI member, whereas the information of the second and subsequent statements of the jobstream is ignored.
• or no error message is issued and the list queue entry has the characteristics of all * $$ LST statements of the jobstream and the information of all statements of the SLI member is ignored.

When PSTOP cuu,FORCE is issued, the termination of the printer task will overlay the previously used save area. This may cause VSE/POWER abnormal termination indicated by message 1Q2CI.

1. When the STOP command is issued and the task is waiting for spool I/O completion, VSE/POWER is terminated abnormally and message 1Q2C is issued.
2. When short queue entries are transmitted via PNET, the console may be overflowing with message 1RA0I and message should be suppressed.

1. Job output was being submitted via the SAS PUT interface, and when the internally provided job and data set control records were opened and initialized by the IPW$$PC function, corrupted data provided by the caller caused a loop.
2. If IPEXTRAD has been specified and PSTART has not yet been issued on the local system for the remote node, the NAK control record received at the remote node may contain sometimes an ip-address unknown to the remote node.

VSE/POWER uses as local ip-address always the one specified in the TCP/IP statement SET IPADDR and passes it via NJE control records to the remote node which issues message 1RT5I if the ip-address does not match its definition. So PNET/TCP is unable to work behind a firewall.

1. When a VSE/POWER subtask fails while it executes a $$B-transient in Logical Transient Area (LTA), the VSE/POWER message 1Q2CI does not point to the failing instruction but to the SVC 2 instead.
2. Msg 1RTUA/VA (by SVC 0) appear in green but should be in red for man. deletion

&PWRTNOD will appear as blank, although the executing node has a non blank node name. Using * $$ LST DEST=* however will create &PWRTNOD with node name of executing system.

1. When VSE/POWER's own 'TI00' Timer Exit routine is disabled by a user exit coded 'SETIME r1,r2' with ECB, or by replacement with OEM own Timer Exit routine, then the Time Event scheduling functions will come to a halt.
2. VSE/POWER Networking entries that are transmitted via RSCS or AIX nodes will produce a "transmission acceptance" message even though the entries did not specify the NTFY= operand.

1. For PALTER LST,H,USER=Rnnn with more than 500 entries in disp. or non disp. chain, command processor loops on writing to queue file - while DMB is unlocked periodically (fortunately).
2. If data is included via * $$ SLI and * $$ DATA statements and a /*$SLI statement is used to end data to be included, results are unpredictable.
3. If in a * $$ LST/PUN statement the PRI= operand is specified without a value, the priority value for the created queue entries shows up as a blank or ','.
4. Trace message 1RT2I ends sometimes with (ACT)NT) wherey the 'NT)' are meaningless.

1. When during profile processing of dyn. part. 'Xn' the partition is cancelled by operator, then 1Q6FA RC=0002 is issued and class 'X' is disabled for job selection.
2. For CRE-queue browsing and Restart to Active the q-rec field QRC2 is used at length of 16 bytes, but cleared in implicite QRC2 length of 32 bytes.

1. When data is received from JES2 (or RSCS) via a SNA link, ending blanks may be lost.
2. When BTRNC=NO has been specified in the PSTART PNET,node command and the hardcopy file wraps around, no information is available whether it has been specified or not.
3. Operating systems other than z/VSE do not support multihoming (as done by z/VSE offering the IPEXTRAD parameter in the PNODE macro) for TCP links. Connections can not be established, message 1RT5I is issued.

1. Palter command has to commit the changes to q-file on disk. When bulk alteration is requested, this may take seconds while DMB is locked - disallowing tasks to gain control that need the DMB resource.
2. The Dsect MAPPOWJB for GETFLD FIELD=POWJOB provides Jobname and Jobnumber of the active VSE/POWER job, but the Jobclass is missing.
3. When m/d/y pattern has been changed during VSE/POWER warm start (e.g. during FSU), CRDATE= specifications are interpreted in the current e.g. d/m/y pattern. Hence entries created before the warm start can not be manipulated by CRDATE= to isolate them.

1. PNET usage:
   1. The JES copy count, for example COPIES=(164,(24,...,17)) of 164 in NDHGDSCT is destroyed when sending 3800 output through an intermediate VSE/POWER PNET node (replaced by number of copy groups).
   2. Job/output data originating from non VSE/POWER (e.g. JES, RSCS) nodes are transmitted with trailing blanks but these are removed forever when data is spooled by VSE/POWER as final or intermediate node.
   3. If job/output data originating from non VSE/POWER (e.g. JES, RSCS) nodes contain trailing blanks and are forwarded to another JES, RSCS, the record length is different from the original record length.
   4. If output data originating from non VSE/POWER (e.g. JES, RSCS) nodes consist of various parts with different record lengths and are forwarded to another JES, RSCS, the record lengths are different form the original record lengths.
2. The VM/CMS user of the VSECMD command does not receive a VSE/POWER response message for the PACCOUNT command.
3. VSE/POWER recovery startup may fail 1Q2CI CC=21, ILLEGAL SVC, PHASE=IPW$$NU, TASK='P PS', even more in VSE/POWER shared environment.

1. When PNET receiver task stalls due to 1Q38A NO DASD AVAILABLE, then the PDRAIN command is not effective, but receiver remains in wait-for-Data-File-DBLK.
2. When SAS-PUT-Job/Out task stalls due to 1Q38A NO DASD AVAILABLE, then the PSTOP SAS command is not effective, but SAS-PUT task remains in wait-for-Data-File-DBLK.

1. Unprintable characters in HDR1-part of message 1QB9A may abend VSE/POWER with message 1Q2CI with SEGM. TRANSL. EXCEPTION in phase IPW$$MX
2. The error handling of PNET SSL for the GSKSSOCINIT call will ignore an I/O error (RC=-00000010 or the equate FRCN0010). Also an ERRNO=00001102 (EWOULDBLOCK) handling leads to an IDUMP in the cases of
   • GSKSSOCREAD
   • GSKSSOCWRITE
   • GSKSSOCCLOSE
   • GSKSSOCRESET
3. PACCOUNT PUN creates IDUMP indicated by msg 1Q2JI in IPW$$02 when data file is full and no DBLKGPs are available.

1. DUETIME operand does not support to calculate a runtime dependant on time when job entered system.
2. The record count of a list output received from a JES2/3 is different than the record count displayed by JES2/3

1. 1. When all DBLKGP cushion elements are used and FREE DBLK GROUP count is 0, the next request for a queue display via Spool Access Support, e.g. by IUI, fails with 1QZ0I RC=0025 in module IPW$$DQ.
2. 2. FULL queue display does not show creation time.

1. POFFLOAD SELECT for labelled tape will bring msg 1Q77I RC=000c if it searches beyond the first tape tape volume and a continued queue entry at the end of a searched volume is not eligible for selection.
2. For dynamic partitions a default * $$ LST/PUN output class cannot be selected as with static partitions, namely (A-Z or 0-9). Instead the dynamic part. default is 'A' or the execution class itself, provided the SET DYNOUTCL=DYNCL stmt. is active.
3. POFFLOAD LOAD/SELECT transports RDR entries of other nodes back to the own RDR queue, but when executed, their output may end up in the XMT queue instead of LST queue.

1. Any other VSE/POWER subtask than Async. Service Subtask (-AS), namely DS/LS/SD/SN/S1/TD/TI should take IDUMP when being cancelled under all circumstances - Cancel Code '09' (LIOCS) should not make an exception.
2. When long running CICS/VTAM/TCP-IP jobs with DISP=D/K are replaced by new version, the old job in DISP=* cannot be PALTER'ed to DISP=D so that it vanishes at system shutdown and allows the new version to be released at system startup. Instead the old version is released once more then.

1. Destroyed VSE/POWER class anchors or chain pointers to the free-q-rec chain result in display of quasi empty 'free' queue records, which may be many.
2. When being used to 'number-of-spooled-pages' for LST queue entries and altering them to the XMT queue, then no more pages but LINES are presented, what may be confusing.
3. 1) When remote node sends FCS bytes not equal to X'8FCF', a loop of IDUMP macro-calls and message 1RTLI may occur.
   2) When remote node sends incorrect BSC frames, the connection hangs and message 1RC1I is issued.

1. Sending small job from RSCS to PNET may cause 'wait-a-bit' re- turned to RSCS resulting in 20 sec delay.
2. MSG 1Q07I INVALID LOGICAL UNIT names 'filename' but not logical unit number when Q/D/A-file have incorrect DLBL/EXTENT/ASSGN.

1. For OUTEXIT (enabled for e.g. LPR by SET OUTEXIT=SAS) which inserts records and where SAS-GET is terminated prematurely by QUIT/CLOSE - the processed q-entry may remain in '*' state without any processing task active.
2. After APAR DY46114 the 'B'rowse column of XMT queue display shows 'D' instead of ' ' when no browser is active.

1. PUT-OPEN-APPEND/RESTART position at OPEN time to the logical-end/target by stepping fwd from begin via the SER records. Hence an existing big appendable output may cost many I/O's for positioning to 'end'.
2. Reloading dispatchable entries via OFFLOAD LOAD/SELECT may cost effort to avoid immediate execution or processing of these entries.
3. For networking any transmitted entry is recorded on console by msg 1RA0I - but for local LST/PUN tasks (especially with 'VM'option) there exists no automatic message for a processed queue entry. Such msg would help in revision.

1. PRELEASE queue,ALL does not touch DISP=L entries but makes DISP=H entries dispatchable (D) without warning/confirmation, whether this mighty function is really wanted on all classes of the named queue.
2. Since DY42661 control of local UR-devices by PSTART/PSTOP/PGO is allowed by Spool Access Support (SAS) users. However 'PDISPLAY M' is not allowed to find out, which task is e.g. 1Q40A bound and needs a 'PGO'.
3. Attempt to change the local node name fails with 1RE8I - but without any ACTIVE SYSID (at switch from non-shr (abn. terminated before) system to a shared system). e.g. 1Q40A bound and needs a 'PGO'.
4. When VSE/POWER queue records have been overwritten in storage or on disk, then VSE/POWER warm start recovery may be mislead by seemingly correct indications in the queue records and may overlay lowcore.

1. For 'write once' tapes/cartridges as 9346 or 3592 - when EOV hit - VSE/ POWER cannot backout the (last) partially written Q-entry by POFFLOAD BACKUP/PICKUP/SAVE, and cannot terminate the last-but-one by 2 TM's.
2. Since DY46067 the 'D A' cmd presents display of SEP/NOSEP/... for LST/PUN tasks. For exec. processor tasks however an additional comma is placed before jobname or 'INACTIVE'.
3. When centr. op. PDISPLAY cmd creates many display lines, then console router buffers may be exhausted & WTO fails & cmd is terminated prematurely.
4. Acc. to A&O description for PDISPLAY STATUS, the 'USED' DBLKGPs comprise 'in-creation','queued','in-deletion' and 'lost-due-err'. This may be incomplete while a big Q-entry is finally deleted from DEL-queue while its DBLKGPs do not yet appear as 'FREE' ones.
5. D STATUS presents shared time intervals when Q-file runs shared, but does not inform about account file usage (also shared or non shared).

1. Extending the data file fails with message 1QD3A DATA FILE EXTENSION FAILED FOR EXTENT NO. nnn, RC=0005 when an existing extent of the data file does not start on track boundary for Data File on CKD/ECKD.
2. Extending the data file fails with message 4783I INVALID LOGICAL UNIT IJDFILE SYSxyz=cuu volid when the new data file extent is defined with a different device geometry but the same device type. For example a 3380 and a 3390 emulated on a 2105 have the same device type x'6E' (meaning ECKD) but a different track size.

1. When JOB segments LST/PUN output in creation using VSE/POWER macro IPWSEGM or SEGMENT, a SAS GET BROWSE request to the new increation queue entry fails with RC=0004, FDBK=0001 and FDB2=000C. IUI dialogue "Manage Batch Queues" 6 = In-Creation Queue fails to access such a segment and issues
   IESV0089I FOLLOWING MESSAGE(S) FROM PROGRAM 'IESBQP2 ', OFFSET X'6D0A'. IESB0012I UNABLE TO CONTINUE - DIAGNOSTIC INFORMATION SAVED.
2. When SET statement DLSEPSAS=FORCE is used IUI "Manage Batch Queues" dialogue fails for each queue type.
   IESV0089I FOLLOWING MESSAGE(S) FROM PROGRAM 'IESBQUP ', OFFSET X'2748'.
   IESB0144I $$SSSSSSSSSS PPPPPPPPPPP LL IESB0144I PREVIOUS MESSAGE WAS RETURNED BY A SUBSYSTEM.

1. Inadvertently entered 'L LST,ALL' may address and erase thousands of entries without requesting operator confirmation. This is too dangerous.
2. Failing XPCC application may loop on CONNECT'ing to SYSPWR hence consuming all VSE/POWER part. Getvis & SETPFIX storage by created SAS tasks.

1. Influencing Spool Access (SAS) GET tasks via SET DLSEP to support and customize single sheet printing does not become active for SAS tasks although stated in A&O on page 181.
2. PSTART TASKTR with FULL option verifies POWER's TCB and BCW chain at every POWER Dispatch cycle - but not for (hundreds) of other partition dispatch events, that could destroy these chains long before POWER can identify the destruction and catch the culprit.

1. After PDISPLAY CRE VSE/POWER may terminate due to SEGMENT Translation Exception as shown in message 1Q2CI
2. IDUMP may not contain SVA-31, although it was requested by TASKTR with option FULL
3. TCP/IP nodes may be disconnected and restarted (messages 1RTWI,1RB0I, 1R03I, 1RB8I), if no data is exchanged between the nodes.
4. Warm start recovery may issue 1QZ0I RC=0013, although all is correct
5. The PSTART LST/PUN options 'SEP/..' are not reflected in D A as the VM option

1. After VTAM abnormal termination and restart of VTAM, RJE,SNA can not be restarted but returns
   1V=2I RC=0004.
2. Attempt to stop VSE/POWER's internal DBLK Group tracing should be rejected by new message 1QAGI:
   1QAGI 'PSTOP DBLKTR' OBSOLETE - NO LONGER NEEDED

1. When a totally filled POWER Account file resides on CKD and is saved to a equally sized CKD SD file by PACCOUNT DISK,dlblname the saving task may fail with message
   0P22I P TRACK FMT SYSxxx=cuu
   The Account file can not be saved thus the system is locked until it is saved to tape or deleted.
2. When TASKTR been Pstarted with FULL to verify the TCB/BCW chains at each VSE/POWER dispatch cycle, POWER may idle for seconds, where other programs might have destroyed POWER's critical storage, what can no longer be re-constructed from the already overflown DEBUG area(s).

1. D LST,,3 command (actually invalid) accepted as D LST, resulting in bulk amount of (CICS) console messages. When console router buffers exhausted POWER loops on wait-1-sec and retry WTO request, while spooling resource DMB remains locked, because display cmd follows a class chain.
2. When SAS spooling output indicates SPLORCFM=ASA, but data records contain MCC op-codes, POWER does NO verification (as documented in Appl. Progr.). When later SAS GET retrieves this q-entry with option SPLGOPT2.SPLGO2AC (convert ASA-->MCC), default action (skip 1 and write immed) is taken for any MCC-code ( = unknown ASA).

1. Implicit TREADY COND=CANCEL for LIBR subtask with CC='FE' does no longer lead to correct DETACH of subtask. Hence POWER task waits eternally for DETACH completion before re-attaching a new LIBR subtask - while ASWE remains locked and all SLI of other partitions grinds to a halt.
2. IfyouuseIPWSEGMKEEP=YES,JECL= .. to overwrite selected attributes of 1st output segment, and if you want to reestablish the overwritten attributes in asubsequentIPWSEGMKEEP=YES,JECL=... request, then you can do it only for attributes re-turned by IPW$MXD at offset X'70'. However e.g. target node/user info is currently not reflected by the IPW$MXD parm. list.

1. When big q-entry or many normal sized are deleted by one command, 'free-writing' of returned DBLKGPs may be lengthy. For that period spooling requests for a new DBLKGP have to wait and command task will not return to requestor.
2. When system crashes while dynamic class executes (one) JOB-P and syst. restarts with SET NORUN=YES, then jobs been in exec. are set DISP=X (also JOB-P), but only static part. enter 1Q36I // PAUSE mode (for op. re-arrangement of exec. sequence), while dyn. class(es) selects the next dispatchable job for execution.

1. When the account file is full, a PACCOUNT DEL/DISK/TAPE must be started, to free the account file. When PACCOUNT task starts, the TCB is located in PFIXed storage cushion. If no PFIXed storage is available, any further request for PFIXed storage issued by the PACCOUNT task fails and the task waits forever, because other tasks can't release storage until they can write their accounting records.

1. Segmentation of 3800 spooled output at PDUMP request can currently be avoided be SET PDUMP=NOSEGMNT. However a program check or cancel dump still creates segmented job output in the list queue.
2. When POWER abn. terminates while in partition A job A1 runs with with 'NORUN=IGN' and another disp. job A2 exists in same class, and when POWER recovery creates a job B1 with DISP=X (due SET NORUN=YES), then job A1 re-starts execution as desired, but subsequent job A2 enters a loop on msg 1R30I.

1. When a VSE/POWER command is entered with an incorrect queue class (PSTART, POFFLOAD or PALTER), then some key-board symbols could cause an incorrect message 1R54I contents, or a program check and an IDUMP.
2. When using the POFFLOAD LOAD command and the POFFLOAD input tape contained incorrect queue entries, then an IDUMP could occur for each entry, causing the IDUMP file to quickly be filled.

• either the following messages are issued:
  BG-0000 (blank line)
  BG-0000 1S01D INVALID STATEMENT.
  and thereafter the statements from the jobstream are displayed:
  BG 0000 * $$ CLASS=U,
  BG 0000 * $$ PRI=8
  and the list queue entry has the characteristics of the first * $$ LST statement of the jobstream and the second and subsequent statements of the SLI member, whereas the information of the second and subsequent statements of the jobstream is ignored.
• or no error message is issued and the list queue entry has the characteristics of all * $$ LST statements of the jobstream and the information of all statements of the SLI member is ignored.

1. When the STOP command is issued and the task is waiting for spool I/O completion, VSE/POWER is terminated abnormally and message 1Q2C is issued.
2. When short queue entries are transmitted via PNET, the console may be overflowing with message 1RA0I and message should be suppressed.

1. Job output was being submitted via the SAS PUT interface, and when the internally provided job and data set control records were opened and initialized by the IPW$$PC function, corrupted data provided by the caller caused a loop.
2. If IPEXTRAD has been specified and PSTART has not yet been issued on the local system for the remote node, the NAK control record received at the remote node may contain sometimes an ip-address unknown to the remote node.

1. It was not possible to specify BLDG,DEPT,ROOM, and PROGR on a * $$ LST/PUN statement
2. A * $$ LST/PUN statement cataloged in a VSE/AF library could not be continued in the next line.
3. For CFNO= a valid FNO had to be specified. There was no way to specify 4 blanks.

1. When RDR type queue entry is restarted during SAS-GET or -BROWSE, and entry consumes more than one DBLK Group, and 'position on lines' (PXRSOPOL) option is set, then the restart position is incorrect.
2. When POFFLOAD SELECT is entered specifying ALL,CDISP= this is rejected with 1R52I O OPERAND 02 MISSING OR INVALID.
3. The date of the last level set VSE/POWER APAR displayed by the 'SIR' command is always in mm/dd format - whatever DATE format option is active.

1. When starting the PNET TCP function, and the hex value of the Host ID is negative, then the startup fails with the error MSG1QZ0I RC=0052. This is due to the EZASMI GETHOSTID interface returning a Host ID whose hex value is negative, which was assumed to be an error condition.
2. When PSTART PARTITION command in Autostart Procedure does not respect the sequence of READER/PRINTERS/PUN.., but mentions e.g. PRINTERs first, then the misleading message 1R91I PSTART TOO MANY OPERANDS ... does not describe the real cause. .

1. When a LST task loads an FCB into VSE/POWER partition GETVIS, a VSE subtask is attached an uses module IPW$$AS to execute the LOAD macro. This macro modifies IPW$$AS to save pointers. When 2nd LST tasks needs to LOAD a FCB phase to another storage area, the 2nd subtask modifies the same code, so that the pointers for the 1st subtask are destroyed. This causes various problems, when the 1st subtask continues to LOAD with the maimed pointers. VSE/POWER part. GETVIS may be destroyed by FCBs loaded to wrong locations.
2. $LSTnnnn entries (PDISPLAY ALL,LST) or $TAPnnnn entries (Tape DISPLAY) are created with DISP=D & CLASS=A. Often they are processed by printers of class A, before they can be viewed by the operator. Entries should be created with at least DISP=H.

1. No encryption is performed, although ENCRYPT=STRONG has been used in PNODE macro. The VSE/POWER command PINQUIRE=remote-nodeid displays a used cipher of X'0001' which means no encryption is performed. In some situations a used cipher of X'0009' is displayed(normal encryption)

   1R56I LOCAL NODE IS ACTING A SERVER,
    .. USED CIPHER=X'0001' 208090A62 02
   

2. No information is available which key and which ciphers are used when the SSL feature is initialized.
3. If 'TLS31' has been used for operand SECTYPE of the PNODE macro, the connection fails with following message

   1RT2I  TCP SSL: EZASMI MACRO-REQUEST
     SSL-SOK-INIT FAILED, RC=-0001335
     ERRNO=+000000, SS=08 FOR
     NODE POWER262 (CLIENT)
   

1. DDS device name or additional log. destinations R000-R250 (restricted in PSTART dev,... for RJE-id's) can be specified but will not post DDS in waiting-for-work state) when LST output added to queue.
2. SAS Direct GET OPEN request to XMT-Q reflects SPLGQI='X' in Verification SPL but does not identify the entry as 'job' or 'output' type.

1. Message 1RV4I and 1RV6I contain header 'TCP SSL' instead of 'TCP/IP'
2. Message 1Q2CI identifying PROTECTION EXCEPTION for module IPW$$TD and SUBTASK=TD after socket call BIND failed (for example with ERRNO=1115)
3. Message 1RT7I and 1RTJA are issued every 20 seconds until TCP/IP interface is established.
4. Trace message 1RTTI not very helpful to debug special error situations.
5. Message 1RT2I missing in special error situations.
6. Loading an NDT with SSL nodes and starting TCP/IP for VSE after 30 minutes produced message: 1QH7A REAL/PFIXED STORAGE CORRUPTED - SHUTDOWN SYSTEM AND RE-IPL

1. at begin of the next VSE/POWER queue entry for NETEXIT, OUTEXIT, and XMTEXIT
2. after an active reader task has been stopped and been re-started for JOBEXIT.

HYPER

1. After PDISPLAY CRE VSE/POWER may terminate due to SEGMENT Translation Exception as shown in message 1Q2CI
2. IDUMP may not contain SVA-31, although it was requested by TASKTR with option FULL
3. TCP/IP nodes may be disconnected and restarted (messages 1RTWI,1RB0I, 1R03I, 1RB8I), if no data is exchanged between the nodes.

When a totally filled POWER Account file resides on CKD and is saved to a equally sized CKD SD file by PACCOUNT DISK,dlblname the saving task may fail with message
0P22I P TRACK FMT SYSxxx=cuu
The Account file can not be saved thus the system is locked until it is saved to tape or deleted.

1. D LST,,3 command (actually invalid) accepted as D LST, resulting in bulk amount of (CICS) console messages. When console router buffers exhausted POWER loops on wait-1-sec and retry WTO request, while spooling resource DMB remains locked, because display cmd follows a class chain.
2. When SAS spooling output indicates SPLORCFM=ASA, but data records contain MCC op-codes, POWER does NO verification (as documented in Appl. Progr.). When later SAS GET retrieves this q-entry with option SPLGOPT2.SPLGO2AC (convert ASA-->MCC), default action (skip 1 and write immed) is taken for any MCC-code ( = unknown ASA).

LEVELSET

1. Implicit TREADY COND=CANCEL for LIBR subtask with CC='FE' does no longer lead to correct DETACH of subtask. Hence POWER task waits eternally for DETACH completion before re-attaching a new LIBR subtask - while ASWE remains locked and all SLI of other partitions grinds to a halt.
2. IfyouuseIPWSEGMKEEP=YES,JECL= .. to overwrite selected attributes of 1st output segment, and if you want to reestablish the overwritten attributes inasubsequentIPWSEGMKEEP=YES,JECL=... request, then you can do it only for attributes re-turned by IPW$MXD at offset X'70'. However e.g. target node/user info is currently not reflected by the IPW$MXD parm. list.

When the account file is full, a PACCOUNT DEL/DISK/TAPE must be started, to free the account file. When PACCOUNT task starts, the TCB is located in PFIXed storage cushion. If no PFIXed storage is available, any further request for PFIXed storage issued by the PACCOUNT task fails and the task waits forever, because other tasks can't release storage until they can write their accounting records.

1. Segmentation of 3800 spooled output at PDUMP request can currently be avoided be SET PDUMP=NOSEGMNT. However a program check or cancel dump still creates segmented job output in the list queue.
2. When POWER abn. terminates while in partition A job A1 runs with with 'NORUN=IGN' and another disp. job A2 exists in same class, and when POWER recovery creates a job B1 with DISP=X (due SET NORUN=YES), then job A1 re-starts execution as desired, but subsequent job A2 enters a loop on msg 1R30I.

1. When a VSE/POWER command is entered with an incorrect queue class (PSTART, POFFLOAD or PALTER), then some key-board symbols could cause an incorrect message 1R54I contents, or a program check and an IDUMP.
2. When using the POFFLOAD LOAD command and the POFFLOAD input tape contained incorrect queue entries, then an IDUMP could occur for each entry, causing the IDUMP file to quickly be filled.
3. When updating a non-continued JECL statement within an SLI member by a continued statement in the jobstream using $SLIxnnn sequence numbers, following messages may be issued:
   * $H DEST=(,HUS), ...
   1Q49I INVALID DELIMITER , COL=00018 ..
   1R33D CORRECT FULL STATEMENT F4
   and if the statement has been corrected by a reply to message 1R33D following messages may be issued with the replid of the user partition:
   0004 // JOB DY4594A2 +
   0004 1S03D INVALID STATEMENT.

HYPER

When PSTOP cuu,FORCE is issued, the termination of the printer task will overlay the previously used save area. This may cause VSE/POWER abnormal termination indicated by message 1Q2CI.

When a continued * $$ LST statement in an SLI member is updated, following error message appears:
1R33D CORRECT FULL STATEMENT
If for example the following SLI-member SLIMEMB.M01 is used with the sequence numbers in column 73-80:
* $$ LST DISP=H,                          M0100001
* $$ CLASS=A,                              M0100002
* $$ DEST=(,HUS),                        M0100003
* $$ FNO=HUHU,                          M0100004
* $$ PRI=4