Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

Navigation
Guide That Contains This Content
[+] Expand All
[-] Collapse All

    Failover Policy

    The failover policy manages how the SIC sends messages over multiple paths to upstream and downstream network elements, and how it responds when it does not receive a response from a target in the network element within a specified amount of time. You configure a failover policy for the following targets:

    • Accounting targets in downstream network elements
    • Authentication targets in downstream network elements
    • Dynamic authorization targets in upstream network elements

    When multiple paths are configured to a network element, you need to specify the order in which the SIC uses the paths to send messages to the target. When the SIC sends a message to one of these targets, it expects to receive a reply within a certain amount of time as specified by the fast fail policy. If it does not receive the reply in the specified time, it places the target into fast fail mode and rejects the request. You configure the failover policy by specifying the fast fail and retry parameters, which control such options as minimum number of times the server retransmits messages to the accounting, delay between sending retransmissions, as well as various timeout and delay settings that control how fast the server goes in and out of fast fail mode.

    When the SIC has messages to send to a target, it first examines what failover mode is configured—either round-robin or primary or backup. It then examines whether all paths to the target are operational. It then sends the messages accordingly (over whatever paths are operating). As such, these features inherently manage communication failures by adjusting what paths are used when one or more paths are not working.

    Failover Mode

    Failover mode manages how the SIC sends messages over multiple paths to a network element target. You can configure failover mode for either round-robin or primary or backup. When the server has a message to send, it first examines whether failover mode is set for round-robin or primary or backup. Next, it examines whether all paths to the network element where the target resides are operational. It then sends the message over whatever path is operational based on failover mode.

    Round-Robin

    When failover mode is set to round-robin, the SIC alternates the path it uses to send messages to the target.

    Figure 1 illustrates how the Round-Robin feature operates when all paths are working properly (top portion), and how it operates when one of the paths has failed (bottom portion).

    With all three paths operating properly, if the SIC received three messages, the first message would be sent over path 1, the second message would be sent over path 2, and the third message would be sent over path 3. The next message received would be sent over path 1, and so on. However, if the server received three messages and path 2 had failed, the first message would be sent over path 1, the second message would be sent over path 3, and the third message would be sent over path 1.

    Figure 1: Round-Robin

    Round-Robin

    Primary or Backup

    When failover mode is set to primary or backup, the SIC sends all messages over the first path defined in the ordered list. If the first path fails, all messages are sent over the next path in the ordered list. When the first path becomes operational again, all messages are again sent over it.

    Published: 2014-06-19