Download presentation
Presentation is loading. Please wait.
1
IBM Tivoli Workload Scheduler for Host 8
IBM Tivoli Workload Scheduler for Host 8.2 PK Z/OS V1R7 COMPATIBILITY FOR TWS PK TO IMPROVE SERVICEABILITY AND ERROR HANDLING IN AN END-TO-END ENVIRONMENT Steve Viola – Level 2 support
2
PK06227 – Compatibility for z/OS 1.7
Modification for EXIT7 (OPCAXIT7) New EXIT51 (TWSXIT51) Minor enhancements Access Register Initialization Console Initialization Change for EQQINIT 9/22/2018
3
PK06227 – Compatibility for z/OS 1.7
PTFS for 8.1: UK05574 and UK05582 PTFS for 8.2: UK05575 and UK05583 9/22/2018
4
PK06227 – Compatibility for z/OS 1.7
For z/OS 1.7, JES2 EXIT7 for TWS will not assemble without PK06227 being applied For z/OS 1.7, JES2 EXIT7 is not called for input phase processing. EXIT51 is used instead EXIT51 is needed for tracking of STCs and handling of JES input errors (example: bad JECL) 9/22/2018
5
PK06227 – Compatibility for z/OS 1.7
Installation Steps: Apply PTFs Run EQQJOBS for new SEQQSAMP members Assemble and link EXIT7 (all z/OS) and EXIT51(z/OS 1.7 only) SMPE: EQQJES2U and EQQJES2V Non-SMPE: EQQJES2 and EQQJES21 Define EXIT51 to JES2 (z/OS 1.7 only) Ensure subsystem modules are loaded at IPL 9/22/2018
6
PK06227 – Compatibility for z/OS 1.7
Define EXIT51 to JES2 (load module TWSXIT51) LOAD(TWSXIT51) EXIT(51) ROUTINES=TWSENT51,STATUS=ENABLED Ensure that TWSXIT51 is in LNKLST or LPALIB so that JES2 can load it Subsystem modules modified by PK06227 must be loaded after IPL: 8.1 EQQSSCME and EQQINITE 8.2 EQQSSCMF and EQQINITF 9/22/2018
7
PK06227 – Compatibility for z/OS 1.7
WARNING: The PTFs for PK06227 may be applied at ANY level of z/OS, however EXIT51 can ONLY be defined at z/OS 1.7 level or higher If PK06227 is installed on z/OS 1.7 and EXIT7 is reassembled but EXIT51 is NOT defined, most TWS functions will work except for tracking of STCs and invalid JECL statements. Example: /*ROUTE PRINX TSO 9/22/2018
8
PK06227 – Compatibility for z/OS 1.7
Error messages if EXIT51 used on pre-z/OS 1.7: $HASP466 PARMLIB STMT LOAD(TWSXIT51) $HASP003 RC=(31),LOAD(TWSXIT51) - MODULE COULD NOT BE LOADED *01 $HASP469 REPLY PARAMETER STATEMENT, CANCEL, OR END REPLY: END 9/22/2018
9
PK06227 – Compatibility for z/OS 1.7
Error messages if EXIT51 used on pre-z/OS 1.7: *$HASP451 ERROR ON JES2 PARAMETER LIBRARY *02 $HASP441 REPLY 'Y' TO CONTINUE INITIALIZATION OR 'N' TO TERMINATE REPLY : Y 9/22/2018
10
PK06227 – Compatibility for z/OS 1.7
Error messages if EXIT51 used on pre-z/OS 1.7: $HASP857 WARNING - EXIT051 NOT DEFINED WITHIN CURRENTLY LOADED JES2 MODULES $HASP858 EXIT ROUTINE TWSENT51 (EXIT051) NOT FOUND *$HASP859 REQUESTED EXIT ROUTINE(S) NOT FOUND *03 $HASP441 REPLY 'Y' TO CONTINUE INITIALIZATION OR 'N' TO TERMINATE REPLY : Y 9/22/2018
11
z/OS 1.7 compatibility: subsystem name table
At z/OS 1.7 level, JES2 R4 mode is no longer supported (JES2 is always ACTIVATED in Z2 mode). This means any pre-TWS subsystem definitions for controllers or trackers will cause S0C1 abends in JES2 (OPC 2.3 or lower) Before IPLing on z/OS 1.7, make sure all TWS subsystems use EQQINITE or EQQINITF 9/22/2018
12
z/OS 1.7 compatibility: subsystem name table
You CANNOT use BUILDSSX and SSCMNAME parameters to change a pre-TWS subsystem definition to a TWS 8.1 or 8.2 subsystem definition if JES2 is in Z2 mode. Even if no controller or tracker is started for a pre-TWS subsystem, JES2 will still abend S0C1 if JES2 is in Z2 mode. 9/22/2018
13
QUESTIONS & ANSWERS 9/22/2018
14
PK01415 : Serviceability and Error Handling for E2E
The following slides are based on a presentation given by TWS level 3: Paolo Falsi Silvia Fama’ Annarita Carnevale PK01415 is for TWS 8.2 only 9/22/2018
15
PK01415 : Serviceability and Error Handling for E2E
Additional APARS: Some problems discovered after PK01415 have been corrected by these APARS: PK11095 PK11182 PK11351 9/22/2018
16
Problems addressed by this PK01415
CEEDUMP and SYSMDUMP are collected. USS files corruptions and/or contentions when multiple AS are generated for the server. Policy for process restart when an abend occurs. Lack of Problem Determination information and messages during the daily planning phase. Lack of Problem Determination information when a file corruption occurs. Wrong definitions of server and daily planning batch job users and groups. 9/22/2018
17
CEEDUMP and SYSMDUMP are collected
9/22/2018
18
CEEDUMP and SYSMDUMP are collected
Prior to PK01415, when a server started task abends in the C/C++ code, a CEEDUMP (LE Dump) of the original abend and a SYSMDUMP with completion code U4039 are taken. The CEEDUMP contains just partial data related to the address space and it's not enough for a complete error analysis. with PK01415 The SYSMDUMP of the original abend is now collected containing also the LEDATA and CEEDUMP information. 9/22/2018
19
Documentation changes
Diagnosis Guide and Reference Applied the following changes: • In Chapter 2. Initial Problem Analysis, section “Problem-Type Keywords”, delete the following sentence from the description of the ABEND keyword: If you are using the end-to-end feature, you could find the CEEDUMP.* dump file in USS in the /tmp or /homedir directory of the user to which the server started-task is associated (using the STC option). • In Chapter 3. Problem Analysis Procedures, section “Information Needed for All Problems”, delete item 8c from the list: Collect the CEEDUMP file if it exists 9/22/2018
20
USS file corruption and/or contention when multiple AS (address space) are generated for the server.
9/22/2018
21
USS file corruption and/or contention when multiple address spaces are generated for the server
Customer experienced event files corruption and/or contention when the TWS server generated processes with parent process id (ppid) equal to 1 (for instance Batchman process). The reason for this problem was the cancel of the server started task when multiple address spaces have been generated (only one Address Space must be generated for all the server tasks/processes/threads). with PK01415 To avoid the generation of multiple address spaces we made the following changes: Rework of the environmental variables handling; in particular the _BPX_SHAREAS variable is now always set to YES for all the processes and threads. Batchman, mailman and writers processes have the same PGID of the netman process. 9/22/2018
22
Environment variable checks (1 of 3)
The putenv() function adds a new environment variable or changes the value of an existing one Before z/os 1.2, the system copied the string inserted in the putenv call into system allocated storage. Now each setting of envvar requires memory allocation to be executed by the caller program. 9/22/2018
23
Environment variable checks (2 of 3)
PK01415 Allocate storage for each environment variable before calling putenv() function A check on each putenv return code has been added in order to check if a putenv of an environment variable gets an error 9/22/2018
24
Environment variable checks (3 of 3)
The following error messages are printed in the server MLOG if the putenv() return code is not equal to zero: EQQ3129E module_name PUTENV environment_variable_string FAILED EQQPT68E PUTENV() environment_variable FAILED ERRNO=error_number: error_message, REASON=reason 9/22/2018
25
_BPX_SHAREAS environment variable
By setting _BPX_SHAREAS to YES, the z/OS will run foreground processes in the same address space that the parent process is running in. The environment variables that affect spawn processing are the ones that are passed into the spawn syscall. The putenv() call, with _BPX_SHAREAS equal to YES, is present more that one time in the code. This situation could produce errors in setting this environment variable with PK01415 This value is set in the USS environment variable only in the Starter process and should be inherited by the child processes because are provided in the spawn syscall 9/22/2018
26
Netman and all the processes started by it now have the same PGID
Process PGID Mailman, batchman and writer processes have a PGID different from netman Netman and all the processes started by it now have the same PGID PK01415 9/22/2018
27
PGID (process group id)
Each process in a process group shares a process group ID (PGID), which is the same as the PID of the first process in the process group. This ID is used for signaling related processes, for example a KILL signal (SIGKILL). D OMVS,A=ALL output shows PID and PPID but not the PGID. To see PGID values, use TSO OMVS command: ps –ef –o pid,ppid,pgid,comm (see output on next slide) 9/22/2018
28
Ps –ef display 1 PID PPID PGID COMMAND 1 0 1 BPXPINPR 2 1 2 EZBTCPIP
/bin/ps EZBTTSSL EZBTMCTL EZACFALG EZASASUB EZBTTMST 9/22/2018
29
Ps –ef display 2 (grep for TWS)
Ps –ef –o pid,ppid,pgid,comm | grep TWS /u/tws82bin/netman /u/tws82bin/translator /u/tws82bin/mailman /u/tws82bin/writer /u/tws82bin/starter /u/tws82bin/batchman 9/22/2018
30
Process display showing relationships
PID PPID PGID CMD EQQPHTOP starter netman translator writer mailman batchman 9/22/2018
31
Policy for process restart when an abend occurs
9/22/2018
32
Policy for process restart when an abend occurs
Before PK01415: In case of fatal errors (abends), the Starter process restarts his children indefinitely. with PK01415 NEW translator and netman process policy NEW mailman and batchman process policy 9/22/2018
33
NEW translator and netman process policy
The restart process has been changed in the following way: => If translator goes down then starter tries to restart it after no more than 5 minutes. => If netman goes down, then mailman, batchman and writers go down. Because translator is strictly related to batchman and mailman, translator goes down also. Also in this case starter tries to restart netman and translator after no more than 5 minutes. Starter tries to restart translator and netman just once; anyway if an abend occurs after more than 2 hours since the last process restart, a new restart is attempted. If the problem persists then message EQQPT63E is logged and starter closes. 9/22/2018
34
PK01415 Message flow related to the process policy
START TIME the following msgs related to processes will be written in MLOG EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/starter" has been started, pid is EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/translator" has been started, pid EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/netman" has been started, pid is the following msg related to netman process will be written in xxx_NETMAN.log AWSEDW075I Netman (pid= pgid= ) was started by the starter process (pid= pgid= ) the following msgs related to mailman/batchman/writer processes will be written in xxx_TWSMERGE.log AWSBCV138I Mailman (pid= pgid= ) was started by netman (pid= pgid= ) AWSBCW056I Writer (pid= pgid= ) was started by netman (pid= pgid= ) AWSBCV108I Started Batchman, pin 763 (old message) 9/22/2018
35
PK01415 Translator policy example (1 of 2)
TRANSLATOR DOWN (kill command issued) the following msgs will be written in MLOG EQQPT11I The Translator process (pid=xxxxxxxx) has been killed by signal SIGKILL TRANSLATOR RESTART EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/translator" has been started, pid is EQQPT20I Input Translator waiting for Batchman and Mailman are started EQQPT21I Input Translator finished waiting for Batchman and Mailman the following msgs related to mailman/batchman/writer processes will be written in xxx_TWSMERGE.log AWSBCV138I Mailman (pid=xxxx pgid=xxxx) was started by netman (pid=xxxx pgid=xxxx) AWSBCW056I Writer (pid=xxxx pgid=xxxx) was started by netman (pid=xxxx pgid=xxxx) AWSBCV108I Started Batchman, pin xxxx (old message) 9/22/2018
36
PK01415 Translator policy example (2 of 2)
TRANSLATOR DOWN (kill command issued after 2 mins) the following msgs will be written in MLOG EQQPT16E The Translator process ended abnormally for twice. Starter and his child processes beginning to shut down EQQPT11I The Translator process (pid= ) has been killed by signal SIGKILL EQQPT12I The Netman process ended successfully EQQPT10I All Starter's sons ended TRANSLATOR DOWN (kill command issued after 3 hours) EQQPT11I The Translator process (pid=xxxxxxxx) has been killed by signal SIGKILL TRANSLATOR RESTART EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/translator" has been started, pid is xxxxxxx EQQPT20I Input Translator waiting for Batchman and Mailman are started EQQPT21I Input Translator finished waiting for Batchman and Mailman 9/22/2018
37
PK01415 Netman policy example
NETMAN DOWN (kill command issued) the following msgs will be written in MLOG EQQPT11I The Netman process (pid=xxxxxx) has been killed by signal SIGKILL EQQPT09E The Mailman and/or Batchman process (pid=Unknown) ended abnormally EQQPT33E Mailman or Batchman ended abnormally. Translator beginning to shut down EQQPT40I Output Translator thread is shutting down EQQPT53I Output Translator thread has terminated EQQPT40I Input Translator thread is shutting down EQQPT53I Input Translator thread has terminated EQQPT40I Input Writer thread is shutting down EQQPT53I Input Writer thread has terminated EQQPT12I The Translator process ended successfully NETMAN RESTART EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/translator" has been started, pid is EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/netman" has been started, pid is EQQPT20I Input Translator waiting for Batchman and Mailman are started EQQPT21I Input Translator finished waiting for Batchman and Mailman 9/22/2018
38
PK01415 mailman and batchman process policy
In case of error (abend) in the mailman or batchman processes the following message will be printed in the MLOG EQQPT33E MAILMAN OR BATCHMAN ENDED ABNORMALLY. TRANSLATOR BEGINNING TO SHUT DOWN After that also translator goes down and then its restart policy applies. 9/22/2018
39
Batchman policy example
BATCHMAN DOWN (kill command issued after 2 mins) the following msgs will be written in MLOG EQQPT09E The Mailman and/or Batchman process (pid=Unknown) ended abnormally EQQPT33E Mailman or Batchman ended abnormally. Translator beginning to shut dow EQQPT40I Output Translator thread is shutting down EQQPT53I Output Translator thread has terminated EQQPT40I Input Translator thread is shutting down EQQPT53I Input Translator thread has terminated EQQPT40I Input Writer thread is shutting down EQQPT53I Input Writer thread has terminated EQQPT12I The Translator process ended successfully BATCHMAN RESTART EQQPT01I Program "/usr/lpp/TWS/TWS820anna/bin/translator" has been started, pid is EQQPT20I Input Translator waiting for Batchman and Mailman are started EQQPT21I Input Translator finished waiting for Batchman and Mailman 9/22/2018
40
Documentation changes
Messages and Codes The following msgs have been added: EQQPT16E THE PROC PROCESS ENDED ABNORMALLY TWICE. STARTER AND CHILD PROCESSES BEGINNING TO SHUT DOWN EQQPT33E MAILMAN OR BATCHMAN ENDED ABNORMALLY. TRANSLATOR BEGINNING TO SHUT DOWN Tivoli Workload Scheduler Administration and Troubleshooting AWSBCV138I Mailman (pid = xxx, pgid= xxx) was started by netman (pid= xxx, pgid = xxx). AWSBCW056I Writer (pid = xxx, pgid= xxx) was started by netman (pid= xxx, pgid= xxx). AWSEDW075I Netman (pid = xxx, pgid = xxx) was started by the starter process (pid = xxx, pgid = xxx). 9/22/2018
41
Lack of Problem Determination information and messages during the daily planning phase.
9/22/2018
42
Lack of Problem Determination information and messages during the daily planning phase.
Added new MLOG messages for DP Batch, Controller and Server during the daily planning phase. Added new TWSworkdir/stdlist/logs/xxxxxxxx_E2EMERGE.log Server messages during the CPUs stopping. during the translator checkpoint file change 9/22/2018
43
New messages for MLOG during a DP extend(underlined) (1 of4)
BATCH CONTROLLER SERVER EQQ3131I WAITING FOR A CP BACKUP EQQN121I START OF DAILY PLANNING ACTIVITY EQQN051I A CURRENT PLAN BACKUP PROCESS HAS STARTED EQQN012I OPC JOB TRACKING EVENTS ARE NOW BEING…………….. EQQN090I THE JOB TRACKING LOG DATA SET………………….. EQQN115I WAITING FOR NCP EQQ3132I CREATING A NEW NCP EQQ3105I A NEW NCP HAS BEEN CREATED EQQ3133I INITIALIZING OF NEW SYMPHONY FILE (RUN NUMBER = &RUNNUMB) EQQ3106I Waiting for SCP EQQN116I A NEW NCP HAS BEEN CREATED (EQQN122I START OF SYMPHONY RENEW ACTIVITY) EQQN117I SYNCRONIZATION BETWEEN CONTROLLER AND SERVER STARTED (send the SYNC S event to Server) 9/22/2018
44
New messages for MLOG during a DP extend(underlined) (2 of4)
EQQPT30I Starting switching Symphony EQQPT75I Syncronization between Server and Controller started EQQPT39I Stopping Mailman and Batchman processes EQQPT12I The Mailman process (pid=xxx) ended successfully EQQPT12I The Batchman process (pid=xxx) ended successfully EQQPT39I Stopping Input Translator Thread activities EQQPT24I Syncronization between Server and Controller ended EQQPT22I Input Translator thread stopped until new Symphony will be available (send the SYNC E to Controller) EQQZ195I SYNCRONIZATION BETWEEN CONTROLLER AND SERVER ENDED (EQQ3091E OPC FAILED TO SYNCHRONIZE WITH THE END-TO-END DISTRIBUTED ENVIRONMENT ) EQQPT39I Sstopping all FTWs EQQPT70I The stop command has been sent to all the reachable FTWs EQQPT71I Waiting for new SCP 9/22/2018
45
New messages for MLOG during a DP extend(underlined) (3 of4)
EQQN051I A CURRENT PLAN BACKUP PROCESS HAS STARTED EQQN012I OPC JOB TRACKING EVENTS ARE NOW BEING…………….. EQQN118I NEW SCP HAS BEEN CREATED (send SYNC Y event to Server) EQQ3107I SCP is ready: Start jobs addition to Symphony file EQQ3108I JOBS ADDITION TO SYMPHONY FILE COMPLETED EQQ3087I SYMNEW FILE HAS BEEN CREATED EQQN090I THE JOB TRACKING LOG DATA SET…………………. EQQPT72I Current plan is executing again EQQN111I A new Symphony file has been created (send SYNC R event to Server) EQQPT73I New Symphony file (run numbers=xxx) is ready EQQPT74I Starting Mailman and Batchman processes 9/22/2018
46
New messages for MLOG during a DP extend(underlined) (4 of4)
EQQPT74I Starting Input Translator activities EQQPT31I Symphony successfully switched (send SYNC X event to Controller) EQQW090I The new Symphony file has been successfully switched EQQPT20I Input Translator waiting for Batchman and Mailman are started EQQPT21I Input Translator finished waiting for Batchman and Mailman EQQPT23I Input Translator thread is running 9/22/2018
47
New messages for TWSworkdir/stdlist/logs/xxxxxxxx_E2EMERGE.log
EQQPT64I STOP COMMAND SENT TO FAULT TOLERANT WORKSTATION CPUNAME EQQPT65I STOP COMMAND SENT TO OPCMASTER EQQPT69I SENDING STOP COMMAND TO FAULT TOLERANT WORKSTATION CPUNAME EQQPT66I Value contained in the Server checkpoint file: key = value 9/22/2018
48
EQQPT66I message description
EQQPT66I = Value contained in the Server checkpoint file: key = value EQQPT66I message prints some Translator Checkpoint file useful variables values at server startup and when these variables are changed. The variables (key) values can be the following: FirstValidSymRun: The First Valid Symphony Run Number used by the server. CPAvailable: Is the SCP (copy of the active Current Plan) available or not SymAvailable: Is a new Symphony available to the server SymRunNumber: Current Symphony Run Number CPRunNumber: The Symphony Run Number in relation to the active current plan. SpecialSynchStart: Is a “special” synchronization in progress 9/22/2018
49
Lack of Problem Determination information when a file corruption occurs
9/22/2018
50
New trace mechanism to get useful information to be used during problem determination
API are provided to the developers to instrument the code Implemented for every USS Server process It is a “wrapping trace” Two trace types are available for every process: Short trace with record length = 3*fullword = 48 byte with records number = header record Long trace with record length = 13*fullword = 208 byte with records number = header record 9/22/2018
51
Actual instrumentations using the trace in memory
Written a record for every write/open, in order to do checks to understand a file corruption reasons and the corruption author A dump will be taken when an error occur accessing an event file showing one of the following messages: EQQPT67E Dump was taken for Problem Determination purpose AWSDDW008E A memory dump was taken to assist in determining the problem 9/22/2018
52
Wrong definition of server and daily planning batch job users and groups
9/22/2018
53
Wrong definitions of server and Daily planning batch job users and groups (1 of 3)
Wrong definition of users or groups may introduce serious errors in an E2E environment. Multiple checks have been introduced to prevent these errors and to signal wrong definitions to user. Using E2E, user assigned to Server or DP batch must have a correct definition on RACF database: · user must have defined an OMVS segment (UID) · his default group must have defined an OMVS segment (GID) Every user defined on RACF database with the same UID of user assigned to Server or DP batch must belong to a group with a defined GID. 9/22/2018
54
Wrong definitions of server and Daily planning batch job users and groups (2 of 3)
Checks added: At start Server and DP batch checks if user assigned has a valid definition; then checks if users with the same UID, belong to a group with a GID defined. Server makes the same checks every five minutes. Every user or group checked, that doesn’t have an OMVS segment assigned, is reported in EQQMLOG with an error message.If there is a RACF access error, the problem is reported with a warning message in EQQMLOG. If DP batch finds an error it stops with return code 12, except for Symphony Renew that stops with return code 8. if Server finds an error it doesn't stop, but issues an error message. 9/22/2018
55
Wrong definitions of server and Daily planning batch job users and groups (3 of 3)
Note: besides the security items checked with PK01415, every userid that runs a DP batch job and the controller and E2E server userids must belong to group eqqGID (specified in job EQQPCS05). However, this is NOT being checked even with PK01415 applied. Error messages if userid does NOT belong to group eqqGID are shown in the next slide: 9/22/2018
56
Errors if userid does not belong to eqqGID
ICH408I USER(USER7 ) GROUP(OMVS ) NAME(TEST USER) /var/TWS820/inst/Symphony CL(DIRSRCH ) FID(01D9F0F1F9F1C5000F B0000) INSUFFICIENT AUTHORITY TO OPEN ACCESS INTENT(--X) ACCESS ALLOWED(OTHER ) EFFECTIVE UID( ) EFFECTIVE GID( EQQ3088E THE SYMPHONY FILE HAS NOT BEEN CREATED 9/22/2018
57
Enhancement request to add eqqGID checking:
MR 9/22/2018
58
Additional APARs which correct problems after PK01415
PK01195 : MESSAGE EQQZ404W TEXT CONTAINS THE LITERAL &UUID BUT IT SHOULD HAVE THE VALUE OF THE USERID INSTEAD PTF: UK06934 is available 9/22/2018
59
Additional APARs which correct problems after PK01415
PK11182 : MESSAGE EQQZ404W (ADDED BY APAR PK01415 ) MAY BE ISSUED INCORRECTLY. BPX_DEFAULT_USER IS NOT CHECKED APAR is currently open, but an APARFIX is available from level 2 support More information on the security considerations corrected by PK11182 later in this presentation 9/22/2018
60
Additional APARs which correct problems after PK01415
PK11351: AFTER PK01415 IS APPLIED, A SYMPHONY RENEW JOB ENDS WITH RC=12 (should be RC=08) PTF: UK07035 is still open, but an APARFIX is available from level 2 support 9/22/2018
61
Security issues resolved by PK11182
After applying PK01415, security problems resulted for the following environments: If BPX_DEFAULT_USER is set up so that ANY user without an explicit OMVS segment inherits the OMVS segment from the default user If IRRIRA00 has been executed so that STAGE 3 is in effect (see z/OS Security Server RACF System Programmer's Guide) 9/22/2018
62
BPX_DEFAULT_USER If the default user is set up with a valid shell (for example: PROGRAM('/bin/sh') ) then any userid is allowed OMVS access. However, the checking done by PK01415 expects an EXPLICIT OMVS segment to be defined. If a user that picks up the default segment attempts to run a CP batch job, these messages are issued: EQQZ401E USER BPXDEF HAS NO VALID UID EQQZ400E A USER ID DEFINITION ON RACF CLASS UNIXMAP IS WRONG EQQZ400I CORRECT THE PROBLEM AND RESTART 9/22/2018
63
BPX_DEFAULT_USER If the default BPX user is set up with an invalid shell program, like PROGRAM= /bin/echo , then any user without an explicit OMVS segment is FAILED if they attempt OMVS access. In this case, PK01415 does not present any new problem. 9/22/2018
64
BPX_DEFAULT_USER To display the BPX default user information, use command: rlist facility bpx.default.user This will include this information: APPLICATION DATA OEDFLTU/OEDFLTG (default user/default group) 9/22/2018
65
BPX_DEFAULT_USER Next, do an LU (list user) on the default user id,for example: LU OEDFLTU OMVS NORACF Resulting display: OMVS INFORMATION UID= HOME= / PROGRAM= /bin/echo . 9/22/2018
66
IRRIRA00 (Stage 3) Any RACF database created at OS/ or later will be at Stage 3. However, if a RACF database was migrated, it could still be a stage 0, 1, or 2 Only stage 3 creates problems if PK01415 is applied, since at stage 3 RACF does not use mapping profiles for UID, GID, SNAME, and UNAME associations. Commands such as ADDUSER no longer maintain the old mapping profiles. 9/22/2018
67
IRRIRA00 continued The following JCL may be executed to determine what stage the RACF database is in: //TEST EXEC PGM=IRRIRA00 //SYSPRINT DD SYSOUT=* Sample output: IRR66017I The system is currently operating in stage 3. 9/22/2018
68
IRRIRA00 continued The problem caused is an incorrect message EQQZ404W at E2E server startup and when CP batch jobs are run: EQQZ404W RACF ACCESS ERROR WHILE CHECKING USERS WITH UID U0 EQQZ404I SAF RC: 0004; RACF RC: 0008; RACF REASON CODE: 0000 However, the E2E server and CP batch jobs continue to run correctly 9/22/2018
69
PK01415 and related APARS – documentation
After the PTFs for PK01415 is applied, the documentation changes are in SEQQMISC member EQQPDFST Download the SEQQMISC member to a PC with extension .pdf in binary mode, and use Adobe Acrobat to read the documentation 9/22/2018
70
PK01415 and related APARS – PTF availability
PK PTFS UK04908, UK04925, UK04927 USS Fix Pack7 is prereq for PK USS Fix Pack8 supercedes PK01415 PK PTF UK07035 currrently OPEN PK11095 – PTF UK06934 PK11182 – APAR is OPEN 9/22/2018
71
TWS 8.2 – recent USS fix packs
APAR PQ PTFs UQ96309 and UQ96295 fix pack7 : APAR PK04260 – PTFs UK02459 and UK02460 fix pack8 : APAR PK10713 – PTFs UK06627 and UK06629 9/22/2018
72
QUESTIONS & ANSWERS 9/22/2018
Similar presentations
© 2025 SlidePlayer.com Inc.
All rights reserved.