IBM Support

IT27255: SERVER MAY HANG WHEN RUNNING EXPORT SERVER AND CLIENT SESSION SIMULTANEOUSLY, AND THE SAME NODE IS PROCESSED.

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • An EXPORT SERVER process may hang when it processes a node that
    also has an active client session running at the same time.
    
    Under these circumstances, other commands will also hang.
    For example, the "show session" will also hang.
    
    Customer/L2 Diagnostics :
    The "show lock" output will show that the thread for the
    EXPORT SERVER process will hold the 17001 lock while also
    waiting for the same lock.
    
    LockDesc: Type=17001(admin node name), NameSpace=0,
      SummMode=isLock, Key='NODEABC'
      Holder: (admutil.c:11088 Thread 111) Tsn=0:000000123,
    Mode=isLock
      Waiter: (admutil.c:11088 Thread 222) Tsn=0:000000456,
    Mode=xLock
      Waiter: (admutil.c:11088 Thread 111) Tsn=0:000000789,
    Mode=sLock
    
    In this case, thread 111 is the thread for the EXPORT SERVER
    command. For example :
    
    Thread 111, Parent 110: XiExport, Storage 1912127,
     AllocCnt 35582 HighWaterAmt 1927853
     tid=1faf2, ptid=13cef, det=1, zomb=0, join=0, result=0,
     sess=0, procToken=2732, sessToken=936451
      Stack trace:
        0x0900000000572200 _cond_wait_global
        0x0900000000572e88 _cond_wait
        0x090000000057382c pthread_cond_wait
        0x000000010000a834 pkWaitConditionTracked
        0x000000010006801c IPRA.$WaitForLock
        0x00000001000664f0 tmLockTracked
        0x00000001000836c0 admGetNodeConvState
        0x0000000100dcc318 IPRA.$DoProcessQueryNode
        0x0000000100dc11bc XiProcessDefinitions
        0x0000000100daff74 ExportData
        0x0000000100dae428 XiExport
        0x000000010000f600 StartThread
      Awaiting cond waitP->waiting (0x1bb067c70),
     using mutex TMV->mutex (0x111646758), at tmlock.c(2525)
     Thread context:
       COMMAND: EXPORT SERVER
       SCHEDULE_TYPE: ADMIN
    
    The show lock will also that the other waiter is a thread
    for a client session of the same node. For example, thread 222
    is this case shows :
    
      Thread 222, Parent 200: psSessionThread, Storage 626664,
      AllocCnt 161 HighWaterAmt 626664
       tid=22446, ptid=cb0d, det=1, zomb=0, join=0, result=0,
       sess=123456, procToken=0, sessToken=123456
        Stack trace:
          0x0900000000572200 _cond_wait_global
          0x0900000000572e88 _cond_wait
          0x090000000057382c pthread_cond_wait
          0x000000010000a834 pkWaitConditionTracked
          0x000000010006801c IPRA.$WaitForLock
          0x00000001000664f0 tmLockTracked
          0x0000000100083cc4 AdmLockNode
          0x000000010026bc10 admUpdateNodeSignOn
          0x00000001004a4a64 IPRA.$HandleNodeSession
          0x000000010048f30c smExecuteSession
          0x00000001004fcd48 psSessionThread
          0x000000010000f600 StartThread
        Awaiting cond waitP->waiting (0x11b55f6b0),
        using mutex TMV->mutex (0x111646758), at tmlock.c(2525)
       Thread context:
         COMMMETHOD: SSL
         THREAD_TYPE: SESSION
         SESSION_TYPE: NODE
         NODE_NAME: NODEABC
    
    The key indicator of the hang will show a call to AdmLockNode
    in several threads.
    
    The hang will resolve itself once the resource timeout period
    has elapsed. The default value is 60 minutes.
    
    Initial Impact: Medium
    
    Additional Keywords:
    hung wait deadlock 17001 TS001645100
    
    MDVPARTL - 7.1 - IC89590
    
    Versions Affected:
    IBM Spectrum Protect server V7.1 and 8.1 on all platforms.
    

Local fix

  • Avoid running a client session while the export server command
    is processing the same client nodename.
    Use the default value for the server resourcetimeout option to
    decrease the amount of time the server remains in a hang
    condition.
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED:                                              *
    * All IBM Spectrum Protect server users.                       *
    ****************************************************************
    * PROBLEM DESCRIPTION:                                         *
    * See error description.                                       *
    ****************************************************************
    * RECOMMENDATION:                                              *
    * Apply fixing level when available. This problem is currently *
    * projected to be fixed in level 8.1.8. Note that this is      *
    * subject to change at the discretion of IBM.                  *
    ****************************************************************
    

Problem conclusion

  • This problem was fixed.
    Affected platforms:  AIX, Linux, and Windows.
    

Temporary fix

Comments

APAR Information

  • APAR number

    IT27255

  • Reported component name

    TSM SERVER

  • Reported component ID

    5698ISMSV

  • Reported release

    81A

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    NoHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2018-12-10

  • Closed date

    2019-01-03

  • Last modified date

    2019-01-03

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

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

Fix information

  • Fixed component name

    TSM SERVER

  • Fixed component ID

    5698ISMSV

Applicable component levels

[{"Line of Business":{"code":"LOB26","label":"Storage"},"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SSGSG7","label":"Tivoli Storage Manager"},"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"81A"}]

Document Information

Modified date:
13 February 2021