IT38745: Highly available Managed File Transfer agent uses 100% CPU while trying to reconnect to its agent queue manager.

APAR status

• Closed as program error.

Error description

• An instance of an MQ Managed File Transfer highly available (MFT
  HA) agent is running on Linux.
  
  While the agent instance is running, it becomes disconnected
  from its agent queue manager. When this happens, the CPU usage
  of the agent instance process remains at 100% until the instance
  reconnects.
  

Local fix

Problem summary

• ****************************************************************
  USERS AFFECTED:
  This issue affects all users of MQ Managed File Transfer highly
  available (MFT HA) agents.
  
  
  Platforms affected:
  MultiPlatform
  
  ****************************************************************
  PROBLEM DESCRIPTION:
  A highly available Managed File Transfer agent consists of:
  
  - One active instance.
  - One or more standby instances.
  
  The first instance of the agent that starts up locks a shared
  resource (the SYSTEM.FTE.HA.agent_name queue on the agent queue
  manager). When the other instances start, they fail to obtain
  the lock and become a standby instance. The standby instances
  will then attempt to take the lock at regular intervals, as
  specified by the agent property standbyPollInterval - once a
  standby instance obtains the lock, then it becomes the active
  instance.
  
  After the active instance has locked the shared resource, it
  performs its normal startup operations and then starts
  processing managed transfers.
  
  Now, if the agent queue manager was stopped while the highly
  available agent was running, then the following sequence of
  events would occur:
  
  - The agent instances became disconnected and immediately tried
  to reconnect.
  - These reconnection attempts failed, because the queue manager
  was not running.
  - The agent instances immediately tried to reconnect again.
  - Once again, the reconnection attempts failed because the queue
  manager was unavailable, and so the instances tried to reconnect
  again straight away.
  - These reconnection attempts also failed.
  - The instances immediately tried to reconnect for a third time.
  
  And so on. Because the instances were trying to reconnect to the
  agent queue manager in a tight loop, they ended up consuming a
  lot of CPU.
  

Problem conclusion

• To resolve this issue, MQ Managed File Transfer highly available
  agents have been updated to wait for the period of time
  specified by the agent property:
  
  standbyPollInterval
  
  before trying to reconnect to the agent queue manager (the
  default value of this property is 5 seconds, which means that
  agents will wait 5 seconds in between reconnection attempts).
  This ensures that the agent does not perform reconnection
  attempts in a tight loop, and so reduces the CPU usage of the
  agent instance processes.
  
  ---------------------------------------------------------------
  The fix is targeted for delivery in the following PTFs:
  
  Version    Maintenance Level
  v9.2 LTS   9.2.0.5
  v9.x CD    9.2.5
  
  The latest available maintenance can be obtained from
  'WebSphere MQ Recommended Fixes'
  http://www-1.ibm.com/support/docview.wss?rs=171&uid=swg27006037
  
  If the maintenance level is not yet available information on
  its planned availability can be found in 'WebSphere MQ
  Planned Maintenance Release Dates'
  http://www-1.ibm.com/support/docview.wss?rs=171&uid=swg27006309
  ---------------------------------------------------------------
  

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:

Fix information

• Fixed component name

  MQ BASE V9.2

• Fixed component ID

  5724H7281

Applicable component levels