Siebel Universal Queuing Administration Guide > Installing and Running Siebel Universal Queuing > Implementing Failure Recovery >

Siebel Communications Server Failure

When the Siebel Universal Queuing routing engine detects that it has failed to receive a heartbeat message from any active Siebel Communications Server connection, the following actions are taken:

• The routing engine attempts to send a critical alarm. However, this may not be possible if no other client asynchronous inputs are available.
• The routing engine logs out all agents that logged in through the failed Siebel Communications Server.

The scenarios to consider for normal connection failures are as follows:

• The connection fails because Siebel Communications Server has failed—Siebel Communications Server does not recover, and agents fail over to another Siebel Communications Server.

  When the agents log in, they are using a different Communications Server and, therefore, a different connection ID, from the one used before the failure.

  When the routing engine detects that it no longer receives the heartbeat message within the time specified by the Ping Interval parameter, as a result of a Siebel Communications Server failure, but the agents have logged back in within the disconnect retry time, the following actions will occur:

  • Interactive items assigned to agents when Siebel Communications Server failed will be completed by the routing engine with a reason code of abnormal termination.
  • Noninteractive items will be requeued. There will be an agent journal entry created to account for the time period of the failure.
  • Agent journal entries will be generated for affected agents to indicate that the agent logged in again after the failure. If, after the disconnect retry time, the agent has not logged back in, then the work items that were assigned to that agent when the failure was detected will be requeued.
• The connection fails because Siebel Communications Server has failed—agents do not fail over and do not promptly log in again.

  When the routing engine detects that it no longer receives the heartbeat message within the time specified by the Ping Interval parameter, as a result of a Siebel Communications Server failure, and the agents have not logged back in within the disconnect retry time, the following actions will occur:

  • Interactive items assigned to agents when Siebel Communications Server failed will be completed by the routing engine with a reason code of abnormal termination.
  • Noninteractive items will be placed back in queues.
  • Agent journals will be generated for all agents logged in on the failed connection, with a logout reason code of connection failure.