JR35910: AFTER APPLYING FP1, SOME JOBS HAVE INTERMITTENTLY THE FOLLOWING FATAL ERRORS.

Download Fix Pack 2 for InfoSphere Information Server Version 8.1

APAR status

• Closed as program error.

Error description

• [Problem]
  After applying FP1, some jobs have intermittently the following
  FATAL errors. (since I translated the Japanese message to
  English, the wording might not be accurate...)
  Jobs were running successfully before applying FP1.
  
  
     Item #: 22
     Event ID: 153
     Timestamp: 2010-01-26 17:32:26
     Type:FATAL
     Username: dsadm
     Message ID: IIS-DSEE-TFIO-00231
     Message: /gpf/data/mid/GPF_DS_CD/GPFDSU1200_2.ds,1:
  Configured timeout of 600 seconds reached for accepting player
  connections for pid 13,636. Pending fifo count: 0. Pending
  shared memory count: 1.  This is most likely due to the failure
  of an upstream operator.
  
     Item #: 23
     Event ID: 154
     Timestamp: 2010-01-26 17:32:26
     Type: FATAL
     Username: dsadm
     Message ID: IIS-DSEE-TFPM-00123
     Message: /gpf/data/mid/GPF_DS_CD/GPFDSU1200_2.ds,1: Fatal
  Error: Cannot start  ORCHESTRATE network connection on Node
  node2 (gpfds). APT_PMConnectionSetup::acceptConnection: Cannot
  accept the connection.
  
   [Additional info.]
  - The same issue happens on several jobs.
  - This happens intermittently. Some times the job aborts but
  some times the job finishes without any problem even though the
  same job and same data is used.
  - The error message shows 600sec Timeout, but it does not take
  600 sec. when the issue happens.
  - If the number of node is 1, the issue does not happen even if
  he tries to test 10 times. But the issue happens when the number
  of node is more than 2.
  - now I'm confirming if there is any change on the system around
  when applying FP1.
  - I'm requesting the job design by using which it is possible to
  reproduce the issue.
  

Local fix

Problem summary

• When using multi-node APT_CONFIG_FILE, a job or jobs may abort
  with following error even the time interval is much less than 10
  minute (600 seconds.)
  
  Message ID: IIS-DSEE-TFIO-00231
  Message: <the-stage-name with node-number>: Configured timeout
  of 600 seconds reached for accepting player connections for pid
  <the-pid>. Pending fifo count: 0. Pending shared memory count:
  1.  This is most likely due to the failur of an upstream
  operator.
  

Problem conclusion

• Install the patch.
  

Temporary fix

• Using 1 node configuration file.
  

Comments

APAR Information

• 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

  WIS DATASTAGE

• Fixed component ID

  5724Q36DS

Applicable component levels

• R810 PSY

     UP

• R850 PSY

     UP