IBM Tivoli Monitoring > Version 6.3 Fix Pack 2 > Installation Guides > High Availability Guide for Distributed Systems > Autonomous mode and autonomous agents > Autonomous mode agent switch from a secondary monitoring server back to the primary hub monitoring server

IBM Tivoli Monitoring, Version 6.3 Fix Pack 2


Switchback processing

When an agent loses its connection with its Tivoli Enterprise Monitoring Server, it switches to a secondary server if one has been defined with the CT_CMSLIST environment variable. The agent must switch back to the primary monitoring server as soon as possible and without restarting.

When an agent running in autonomous mode detects that the primary hub has come back online, it needs to switch from the secondary hub monitoring server back to the primary. The following steps show how the switchback proceeds:

  1. The agent detects that it has connected to a secondary monitoring server.

  2. The switch back to the primary server occurs when the following conditions are true:

    • There are multiple monitoring servers defined by the CT_CMSLIST parameter.

    • An agent is actively connected to a secondary monitoring server.

    • That agent is not in shutdown mode.

  3. The agent starts a background periodic lookup task.

  4. The agent calculates the periodic lookup interval from the CTIRA_PRIMARY_FALLBACK_INTERVAL customized value and enforces the minimum value limit.

  5. When a periodic lookup interval expires, the lookup task performs Local Location Broker lookup of the primary server. This periodic lookup continues until the primary Tivoli Enterprise Monitoring Server is found or the agent shuts down.

  6. On a successful lookup of the primary monitoring server, the agent initiates server-switch processing:

    1. Notifies the acting monitoring server that the agent is now disconnecting.

    2. Performs the agent shutdown procedure as if it has lost monitoring server connectivity.

    3. Sets the primary monitoring server as the preferred switch-to target.

    4. Begins the server connection procedure.

    5. Exits periodic lookup unconditionally: If the reconnection is successful, there is no need for the periodic lookup task. Otherwise, the standard agent reconnection procedure takes effect.


Parent topic:

Autonomous mode agent switch from a secondary monitoring server back to the primary hub monitoring server

+

Search Tips   |   Advanced Search