IT35599: Standby instance of a Managed File Transfer highly available agent shuts down after being disconnected from its queue manager

APAR status

• Closed as program error.

Error description

• An instance of an MQ Managed File Transfer highly available
  agent starts up, connects to its agent queue manager and becomes
  the standby instance. Shortly after the instance starts, the
  agent queue manager is stopped. As a result, the standby
  instance writes the following messages to its event log and
  shuts down:
  
  BFGAG0200E: An error occurred while determining the version of
  active instance. The error is 'cc=2 rc=2009 op=_get - MQGET'.
  BFGAG0071I: The agent has suspended its current transfers and is
  now stopping.
  

Local fix

Problem summary

• ****************************************************************
  USERS AFFECTED:
  This issue affects all users of MQ Managed File Transfer highly
  available (HA) agents.
  
  
  Platforms affected:
  MultiPlatform
  
  ****************************************************************
  PROBLEM DESCRIPTION:
  When using highly available Managed File Transfer agents, one of
  the requirements is that the active instance and all of the
  standby instances are running the same version of the product.
  The following mechanism is used to perform this version check:
  
  ------------------------------
  Step 1:
  --------
  After a standby instance of an agent has started up and
  connected to its agent queue manager, it will
  - Create a temporary queue on the agent queue manager.
  - Construct a message containing various pieces of information,
  including the version of the product that it is using
  - Put the message to the SYSTEM.FTE.COMMAND.<agent_name> on the
  queue manager.
  - And wait for a reply to arrive on the temporary queue.
  
  Step 2:
  --------
  The active instance of the agent picks up the message from the
  SYSTEM.FTE.COMMAND.<agent_name> queue, extracts the version
  information and compares it to the version of the product that
  it is running.
  
  If it determines that it is using the same version as the
  standby instance, then it constructs a message indicating that
  the standby instance can remain active and puts it onto the
  temporary queue.
  
  However, if the active instance finds that it is running a
  different version to the standby instance, then it constructs a
  message indicating that the standby instance should shut down.
  This message is then put to the temporary reply queue.
  
  Step 3:
  --------.
  When the standby instance gets the message from the reply queue,
  it checks the contents
  
  If the message indicates that the standby instance can remain
  active, then the instance will start trying to open the
  SYSTEM.FTE.HA.<agent_name> queue on the agent queue manager at
  regular intervals, so that it can take over if the active
  instance stops for some reason.
  
  If the reply message indicates that it needs to stop, then the
  standby instance writes the message:
  
  BFGAG0199E: The version of IBM MQ Managed File Transfer
  installed on standby machine does not match with the version on
  active instance running elsewhere. This instance of agent will
  stop."
  
  to its event log and shuts itself down.
  &#09;
  ------------------------------
  &#09;
  Now, if the standby instance encountered an issue while
  performing the version check, it would write the messages:
  
  BFGAG0200E: An error occurred while determining the version of
  active instance. The error is 'cc=<completion code> rc=<reason
  code> op=<operation>
  BFGAG0071I: The agent has suspended its current transfers and is
  now stopping.
  
  to its event log (output0.log) and shut itself down, regardless
  of whether the issue was a recoverable error (such as the
  instance becoming disconnected from its agent queue manager due
  to a network error) or a more permanent failure. When this
  happened, the instance had to be manually restarted.
  

Problem conclusion

• To resolve this issue, MQ Managed File Transfer highly available
  agents have been updated so that:
  
  - If a standby instance encounters an issue while performing the
  version check.
  - And that issue represents a recoverable error (such as
  MQRC_CONNECTION_BROKEN)
  
  then the instance will write the message:
  
  BFGAG0183I: The agent received MQI reason code <reason_code>.
  Agent recovery will be initiated.
  
  to its event log and then attempt to reconnect to the agent
  queue manager. This ensures that the instance remains active,
  and does not require a manual restart.
  
  ---------------------------------------------------------------
  The fix is targeted for delivery in the following PTFs:
  
  Version    Maintenance Level
  v9.2 LTS   9.2.0.3
  v9.x CD    9.2.3
  
  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