PQ63439: ECSA AND CSM DATASPACE STORAGE GROWTH AND TI BUFFER STORAGE GROWTH 02/07/29 PTF PECHANGE

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• Using EE ( Enterprise Extender ) over non QDIO devices ECSA,
  CSM dataspace and TI buffer growth can occur.
  

Local fix

• KEYWORDS:  IUDR
  ADDITIONAL SYMPTOMS:
  4K ECSA CSM buffers will be owned by VTAM
  32k CSM dataspace growth may also occur
  569511701
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of the Communications Server for   *
  *                 OS/390 Release 10 IP and z/OS                *
  *                 Version 1 Release 2:  Enterprise Extender    *
  ****************************************************************
  * PROBLEM DESCRIPTION: VTAM uses excessive CSM storage when    *
  *                      running Enterprise Extender.  TI buffer *
  *                      and CSM storage allocation failures.    *
  *                                                              *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  CSM storage growth in IO and TI buffers occurs when running EE
  (Enterprise Extender).
  
  EE traffic outbound over a Packing Claw device fails to release
  the TI buffers.  The number of xbufls in the dlru is not
  incremented in the packing code when sending outbound over a
  claw device.  The release logic in EZBIFIF1 does not release
  the storage because the number of free requests never matches
  the number of xbufls.
  
  EE segments that are larger than the outbound device MTU are
  fragmented.  The dlrumsg@ is updated for each xbufl and should
  only be updated one time unless the segment is packed.  The
  incorrect updating of the dlrumsg@ for non-packed segments
  sets dlrumsg@ to zero or causes an abend0C4.  If the dlrumsg@
  is zero, EZBIFIF1 will not release the TI buffer used to pass
  the original unfragmented segment.
  
  Requests for new TI buffers and/or CSM storage will fail with
  the following messages when CSM ECSA has been exhausted.
    IST154I EXPANSION FAILED FOR TI BUFFER POOL - CODE 4
    IVT5506I  IVTSMCEX STORAGE ALLOCATION FAILED IN CSM
  
  A display of VTAM's buffer usage will show a very large number
  for the current total of IO00 and TI00 buffers.
  +-------------------------------------------------------------+
  + Please check our Communications Server for OS/390 homepages +
  + for common networking tips and fixes.  The URL for these    +
  + homepages can be found in Informational APAR II11334.       +
  +-------------------------------------------------------------+
  

Problem conclusion

• EZBIFIUT has been changed to set the dlrumsg@ once when the
  segment is not packed.  EZBIFOUT has been amended to increment
  the number of messages when the packed segment is being sent
  over a claw device.
  
  * Cross Reference between External and Internal Names
  

Temporary fix

• *********
  * HIPER *
  *********
  

Comments

APAR Information

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

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

  UQ69039 UQ69040

Modules/Macros

• EZBIFIF1 EZBIFIUT EZBIFOUT EZBUDBYP
  

Fix information

• Fixed component name

  TCP/IP V3 MVS

• Fixed component ID

  5655HAL00

Applicable component levels

• R120 PSY UQ69039

     UP02/08/29 P F208 «

• R50A PSY UQ69040

     UP02/08/29 P F208 «

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.