IY95372: AGENT SHOWED OFFLINE WHILE ACTUALLY IT WAS STILL ONLINE.

APAR status

• Closed as program error.

Error description

• Large amounts of ITM agents were shown offline in Managed System
  Status workspace.  However, the agents were still actively
  connect to the remote HUB system.
  
  The problem was caused by agent switching mechanism.  This was a
  design problem.  An agent was originally connected to remote
  TEMS.  However, due to agent switching, it will be reconnected
  to another remote TEMS.  At this time, the new remote TEMS
  will send a Y record to HUB.  Then the original remote TEMS
  heartbeat expired and it will sent out a N record to HUB.  So
  the status will be showing "offline" for the agents.
  

Local fix

• Recycle the agent and the status on the TEPS will show "online"
  again.
  

Problem summary

• Delayed node status events are overwriting accurate node status
  states at the HUB after an agent switches to a different REMOTE
  TEMS.
  
  When an agent switches to a different REMOTE TEMS the status for
  that agent will be sent to the HUB TEMS.  The HUB TEMS will
  reflect the agent's status and the new thrunode (the REMOTE
  TEMS) that it is reporting through to the HUB.  If the REMOTE
  TEMS comes back online that the agent had been reporting
  through, it will send a status record for this agent has it does
  not know that the agent has switched.  The agent's status will
  reflect a offline state which is incorrect.  The code has been
  modified to ignore the offline status record if the agent is
  online.  However, a dignostic trace message is included for this
  to indicate the event was discarded.  Additionally at the HUB,
  this diagnostic trace message is also published whenever a
  REMOTE TEMS goes offline.  This is a know problem and will be
  fixed in a future release.
  
  An example of this message text is:
     ('462FA008.01BC-BD0:kfastins.c,1764,"Process") Node Status
     Update is ignored - OFFLINE notification is ignored
     since the thrunode <REMOTE_NEBULA2>  is not the
     current thrunode <HUB_NEBULA>'
  

Problem conclusion

• The delayed node status event is ignored if the thrunode is
  different and the status reflects an OFFLINE status the agent is
  currently ONLINE.  Additionally a trace statement is issued each
  time this occurs.
  
  The fix for this APAR is included in the following maintenance
  vehicle:
  
  | fix pack | 6.1.0-TIV-ITM-FP0005
  

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:

  OA20874

Fix information

• Fixed component name

  TEMS

• Fixed component ID

  5724C04MS

Applicable component levels

• R610 PSY

     UP