PH21495: EYUTS0004I CMAS1B TOPOLOGY FOUND WUI1F ALREADY ACTIVE WHEN WUI RESTARTS ON A DIFFERENT LPAR 20/03/27 PTF PECHANGE

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• A WUI is restarted on a different LPAR but
  fails to connect to the CMAS on that LPAR.
  Error message is issued in the CMAS log  :
  EYUTS0004I CMAS Topology Found WUI already Active.
  Problem scenario :
  WUI WUI1F and CMAS CMAS1D are connected on LPAR D.
  CMAS1D is shutdown. WUIIF remains active on LPAR D
  CMAS1B is started on LPAR B
  WUI1F terminates on LPAR D
  When WUI1F restarts on LPAR B, the topology connect with
  CMAS1B fails with message :
  EYUTS0004I CMAS1B Topology Found WUI1F already Active.
  Time line and messages as follows :
  
  LPAR D
  12/22/2019 22:32:47 EYUTS0003I CMAS1D Topology Connect for WUI1F
  Complete - APPLID(WUI1F) CICSplex(PROD).
  ........
  12/29/2019 22:00:43 EYUXL0011I CMAS1D CMAS shutdown in progress.
  12/29/2019 22:00:45 EYUXL0015I CMAS1D Message Services
   termination requested.
  LPAR B
  12/29/2019 22.01.50 STC26876 IEF695I START CMAS
  12/29/2019 22.01.50 STC26876 DFHPA1101 CMAS1B DFHSIT6$ IS BEING
  LOADED.
  LPAR D
  12/29/2019 22:03:16 EYUVS0004I WUI1F CICSPLEX SM WEB USER
  INTERFACE TERMINATION COMPLETE.)
  LPAR B
  12/29/2019 22.03.27 STC26948 IEF695I START WUI   WITH
  12/29/2019 22.03.27 STC26948 DFHPA1101 WUI1F DFHSIT6$ IS BEING
   LOADED
  12/29/2019 22:03:32 EYUTS0001I CMAS1B Topology Connect for
   WUI1F Initiated - APPLID(WUI1F) CICSplex(PROD).
  12/29/2019 22:03:32 EYUTS0004I CMAS1B Topology Found
   WUI1F already Active.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CPSM Users with version V5R3M0,V5R4M0    *
  *                 and V5R5M0.                                  *
  ****************************************************************
  * PROBLEM DESCRIPTION: This APAR solves two problems for CPSM  *
  *                      asynchronous topology processing for    *
  *                      JOIN and TERMINATE.                     *
  *                      Problem 1. A MAS shuts down and         *
  *                      restarts on another LPAR, connecting to *
  *                      another CMAS, CMAS issues message       *
  *                      EYUTS0004 indicates the MAS is already  *
  *                      active.                                 *
  *                      Problem 2. A MAS restarts when the      *
  *                      system is quite busy, the CMAS issues   *
  *                      message EYUTS0025 indicates a           *
  *                      CONNECT/JOIN is terminated due to       *
  *                      disconnect request.                     *
  ****************************************************************
  * RECOMMENDATION: After applying the PTF resolving PH21495,    *
  *                 all CMAS regions need to restart to pick up  *
  *                 the change. Note that the restarts do not    *
  *                 need to occur at the same time.              *
  ****************************************************************
  This APAR addresses two problems of CPSM asynchronous processing
  MAS JOIN and TERMINATE.
  Problem 1. A CMAS region shuts down, but the connected MAS
  regions do not also terminate. These MAS regions issue message
  EYUXL0018I:
  EYUXL0018I applid LMAS restart in progress
  The CMAS is then restarted and performs a CPSM WARM restart.
  This initiates CPSM topology warm start processing for the MAS
  regions which remained active. Message EYUTI0009I is issued to
  the EYULOG for each of these MAS regions:
  date time EYUTI0009I applid Topology warm start for sysname
            initiated - APPLID(masapplid) CICSplex(cicsplex)
  These MAS regions are considered to be in lost contact status
  until a full connection can be re-established. An EENDCONN event
  is generated and passed to the topology long running task (TIST)
  to initiate the new connection. Other CMASes in the network will
  generate an EENDJOIN event, which will be passed to the TIST
  task. In these CMAS regions the MAS JOIN timestamp
  (CSDB_STRT_TSSCJ) is stored into the CICS System Descriptor
  Block (CSBD), even though the join method (TSSJ) will not be
  invoked until later. Before the connect process can complete,
  one of these MAS regions terminates. The other CMASs in the
  network will generate an EENDEXIT event (and set CSDB_STAT_TERM
  in the CSDB), but this will not be processed correctly due to
  the CSDB_STRT_TSSCJ state.
  The MAS is restarted on another LPAR, connecting to one of these
  other CMAS regions. The connect is unable to complete as the
  CSDB in that CMAS is now in an inconsistent state.
  Message EYUTS0004I will be issued to the CMAS EYULOG, and
  EYUTS0002E will be issued to the EYULOG and JESMSGLG.
  date time EYUTS0004I applid Topology Found sysname already
            Active.
  date time EYUTS0002E applid Topology Connect for sysname
            Failed - APPLID(masapplid) CICSplex(cicsplex).
  .
  Problem 2. This problem occurs when MAS terminates and restarts
  quickly.
  When a MAS region terminates, a termination event EENDDISC is
  generated and passed to the topology long running task TIST to
  initiate termination. The CSDB for the MAS is updated with a
  termination timestamp (CSDB_STRT_TSST). When TIST task gets to
  process the MAS termination event, the CSDB_STRT_TSST is
  updated. If the TIST task is not able to respond quickly, then
  the MAS may restart and attempt to reconnect before the
  termination event can be processed. At this time, CMAS considers
  the termination is requested after the connect request, which is
  not the actual request order.
  
  Messages EYUTS0001 and EYUTS0003 are issued to the
  CMAS EYULOG and JESMSGLG showing the MAS has been terminated
  successfully. Then, messages EYUTS0001 and EYUTS0025 are issued
  to CMAS EYULOG and JESMSGLG indicating that a topology Connect
  for a MAS is terminated due to Disconnect request.
  date time EYUTS0001I applid Topology Disconnect for sysname
            Initiated - APPLID(masapplid) CICSplex(cicsplex).
  date time EYUTS0003I applid Topology Disconnect for sysname
            Complete - APPLID(masapplid) CICSplex(cicsplex).
  date time EYUTS0001I applid Topology Connect for sysname
            Initiated - APPLID(masapplid) CICSplex(cicsplex).
  date time EYUTS0025I applid Topology Connect for sysname
            terminated due to Disconnect request
            - APPLID(masapplid) CICSplex(cicsplex).
  

Problem conclusion

• UI63618 UI63619 UI63620
  
  EYU0TIST has been updated to enhance CPSM synchronous processing
  for MAS JOIN and TERMINATE.
  1. Update MAS JOIN timestamp (CSDB_STRT_TSSCJ) settings to make
  sure it is set only when TSSJ is actually invoked;
  2. Remove redundant terminate timestamp (CSDB_STRT_TSST) setting
  in TIST, so that the request termination timestamp is always
  consistent with the time termination event generated.
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

  UI68664 UI68665 UI68666

Modules/Macros

• EYU0CPAM EYU0CPEJ EYU0CPLT EYU0TIST EYU0TSSA EYU0TSSC EYU0TSSJ
  EYU9TSR3 EYU9TSR4 EYU9TSRU EYUT2459 EYUT2461 EYUY2459 EYUY2461
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R00M PSY UI68664

     UP20/04/03 P F004 {

• R10M PSY UI68665

     UP20/04/03 P F004 {

• R20M PSY UI68666

     UP20/04/03 P F004 {

Fix is available

• Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.