IBM Support

OA59569: IST1222I WRITE DEVICE XXXX IS INOPERATIVE ABEND0C4 IN IOSVSSCQ HIGH 4K ECSA CSM

A fix is available

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • First we see a WRITE DEVICE xxx IS INOPERATIVE
    
    Later we see an Abend0c4 in IOSVSSCQ
    The CSM 4K ECSA is filled up
    CSM 4K ECSA is full of IUTIL PARMLISTs containing
    MSG_UDATA_REQ for an OSA that had suffered an INOP on the
    TRLE's WRITE device 0A01 many days earlier.
    
    OSA firmware will "kill" the read to inop the group and that was
    recognized.  The unexpected abend on the write inoped it, but
    the group never progressed to come down and clean up.
    VTAM never notified the stack the interface was down, so these
    QueryARP IP assists built up on the MPNCB
    found 38K of them on the NCBWORKq (TI buffers)
    The TI00 buffer pool is really high too.
    
    ANALYSIS:
    PIDS/5752SC1C3 RIDS/NUCLEUS#L RIDS/IOSVSSCQ AB/S00C4
    PRCS/00000038
     RIDS/IOSVSSCQ#R
     Symptom       Description
     -------       -----------
     PIDS/5752SC1C3   Program id: 5752SC1C3
     RIDS/NUCLEUS#L   Load module name: NUCLEUS
     RIDS/IOSVSSCQ    Csect name: IOSVSSCQ
     AB/S00C4      System abend code: 00C4
     PRCS/00000038    Abend reason code: 00000038
     REGS/918C2     Register/PSW difference for R09:-18C2
     RIDS/IOSVSSCQ#R   Recovery routine csect name: IOSVSSCQ
    OTHER SERVICEABILITY INFORMATION
     Recovery Routine Label: SSCQFRR
     Date Assembled:     19302
     Module Level:      UJ00950
     Subfunction:       IOS SSCH QUEUE ROUTINE
    Time of Error Information
     PSW: 04041001 80000000 00000000 010E0ACE
     Instruction length: 04  Interrupt code: 0038
     Failing instruction text: E3202000 00045820 2040C22B
     Translation exception address: 807AD039_FFB00800
     Breaking event address: 00000000_010E0A62
     AR/GR 0-1  00000000/000001EF_0299CE20
    00000000/000001EF_2EB7C1D8
     AR/GR 2-3  00000000/807AD039_FFB00D10
    00000000/00000000_025E1E78
     AR/GR 4-5  00000000/00000000_2EB7C1D8
    00000000/00000000_2EB7C1D8
     AR/GR 6-7  00000000/00000000_062EF1F0
    00000000/00000000_025E1E78
     AR/GR 8-9  00000000/00000000_062ED4A0
    00000000/00000000_010E2390
     AR/GR 10-11 00000000/00000000_00001BB8
    00000000/00000000_2EB7C018
     AR/GR 12-13 00000000/000001EF_000019B8
    00000000/00000000_062EF288
     AR/GR 14-15 00000000/00000000_00000020
    00000000/000001EF_00001900
    

Local fix

  • From TCP/IP, issue a Vary STOP command against the QDIO
    interface to cease the TI00 buffer pool and 4K ECSA storage
    growth that followed the QDIO write control channel INOP.
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED:                                              *
    * All users of the IBM Communication Server                    *
    * for z/OS Version 2 Releases 2, 3 and 4 SNA:                  *
    * QDIO                                                         *
    ****************************************************************
    * PROBLEM DESCRIPTION:                                         *
    * TI00 Buffer pool and 4K ECSA CSM buffer pool storage growth  *
    * after QDIO write control channel INOP.                       *
    *                                                              *
    * IST1222I WRITE DEVICE wwww IS INOPERATIVE, NAME IS trlename  *
    *                                                              *
    ****************************************************************
    * RECOMMENDATION:                                              *
    * Apply PTF                                                    *
    ****************************************************************
    The problem is summarized as follows:
    1) A QDIO interface was activated from a TCP/IP stack.
    2) At some point, the UCB control block structure for the write
    control channel was overlaid.  The root cause of this overlay is
    unknown.
    3) The overlay results in various Abends in I/O Supervisor code
    including:
         ABEND0C4 IN IOSVSSCQ
    4) VTAM issues message IST1222I to document the write control
    channel inop:
    
        IST1222I WRITE DEVICE wwww IS INOPERATIVE, NAME IS trlename
    
    
    5) VTAM should INOP the MPC group from this failure, but failed
    to do so.
    6) A display of the QDIO TRLE whose name is listed in messages
    IST1222I shows the write subchannel INOP while the read
    subchannel is ACTIVE:
    
    D NET,ID=trlename
    
    IST097I DISPLAY ACCEPTED
    
    IST075I NAME = trlename, TYPE = TRLE
    
    IST486I STATUS= ACTIV, DESIRED STATE= ACTIV
    
    IST087I TYPE = LEASED             , CONTROL = MPC , HPDT = YES
    
    IST1954I TRL MAJOR NODE = trlmajnd
    
    IST1715I MPCLEVEL = QDIO       MPCUSAGE = SHARE
    
    IST2263I PORTNAME = portname   PORTNUM =   0   OSA CODE LEVEL =
    ABCD
    IST2337I CHPID TYPE = OSD      CHPID = C4  PNETID = **NA**
    
    IST1577I HEADER SIZE = 4096 DATA SIZE = 0 STORAGE = ***NA***
    
    IST1221I WRITE DEV = wwww STATUS = INOP       STATE = ONLINE
    
    IST1577I HEADER SIZE = 4092 DATA SIZE = 0 STORAGE = ***NA***
    
    IST1221I READ  DEV = rrrr STATUS = ACTIVE     STATE = ONLINE
    
    
    7) Because the MPC group was not INOPed, VTAM failed to notify
    the stack of the outage.
    8) At thirty second intervals, the TCP/IP stack sends two IP
    Assist commands to query ARP information from the OSA-Express
    interface.  These requests requires two TI00 buffers and two 4K
    ECSA CSM pages.
    9) These are packaged as IUTIL requests and each request
    occupies 4K of ECSA CSM along with a VTAM TI00 buffer.
    10) The IUTIL tipac buffer is queued to the MPNCB NCBWORKQ for
    processing.  Because the write device is INOPed, the data is
    never processed.  This process expects the MPC Group is being
    brought down because of this condition.
    11) Because VTAM failed to INOP the MPC group for this
    condition, the Write device stays in INOP state and the READ
    device stays in ACTIVE state from the VTAM perspective.  The
    TCP/IP stack still shows the OSA interfaces as READY.
    12) Over time ECSA growth occurs because of this exposure.
    13) CSM tracking (CSMUSE) shows the 4K ECSA CSM storage growth
    is accummulating for component ID CMPID_IUT_Req  ('20'X).  Also,
    the TI00 buffer pool is growing too.
    14) A TCP/IP stop command must be issued for the OSA express
    interface in each of the owning TCP/IP stacks to cease the  ECSA
    storage growth.
    

Problem conclusion

  • ISTLLCM8 - Has been updated to correctly INOP the MPC group for
    a QDIO TRLE when an INOP occurs for the write control channel
    and the read control channel is still active.
    ISTLLCIE - Has ben included for maintenance purposes.
    

Temporary fix

Comments

APAR Information

  • APAR number

    OA59569

  • Reported component name

    VTAM MVS/ESA

  • Reported component ID

    569511701

  • Reported release

    230

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    NoHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2020-05-12

  • Closed date

    2020-05-21

  • Last modified date

    2020-08-03

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

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

    UJ02978 UJ02979 UJ02983

Modules/Macros

  • ISTLLCM8 ISTLLCIE
    

Fix information

  • Fixed component name

    VTAM MVS/ESA

  • Fixed component ID

    569511701

Applicable component levels

  • R220 PSY UJ02978

       UP20/07/30 P F007

  • R230 PSY UJ02979

       UP20/07/30 P F007

  • R240 PSY UJ02983

       UP20/07/30 P F007

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.

[{"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Product":{"code":"SSSN3L","label":"z\/OS Communications Server"},"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"230","Line of Business":{"code":"LOB35","label":"Mainframe SW"}}]

Document Information

Modified date:
05 August 2020