Tuning Parameters

With the Tuning Parameters you can set maximum delivery characteristics.

Setting
Description
Delivery Mode
The delivery mode feature allows the Unica Discover Network Capture Application to deliver traffic through different methods to its delivery peers (css boxes). Traffic can be delivered by the following methods:
  • Even Distribution - Traffic is distributed evenly among peers. For example, if four peers are configured, then each receives approximately 25% of total traffic.
  • Failover - This method requires two peers, a primary and secondary. Primary receives 100% of the total traffic while the secondary idles. If connection to the primary peer is lost or too many hits are being dropped, the DNCA closes the connection to the primary peer and fails over to the secondary. If the secondary peer fails, the DNCA attempts to fail back to the primary peer. If the DNCA is unable to establish a healthy connection to both the primary and secondary peer, it switches between the two until a healthy connection can be established.
    Note: This delivery mode is likely to be deprecated in a future release.
  • Clone - This option was the previous default behavior. Every peer gets 100% of the total traffic.
  • None - This option replaces the 'deliver to null' option. If this option is selected, traffic is dropped before being sent to any delivery peers, which are useful for debugging purposes only.
Max Delivery Wait
Sets the duration of the timer that is used to send packaged hits. When this timer expires, the device attempts to send all packaged hits to the target recipients.
Polling Interval
This setting is not currently used.
Watchdog Timer
The maximum time (in seconds) allowed to make a connection to the Unica Discover Server. If the timeout is exceeded, the connection is marked as disconnected. The default value is 30 seconds.
Max Queue Length
Sets the maximum length of the delivery queue. If the queue reaches this length, the system begins discarding newly packaged hits until the queue shortens. By enforcing this limit, the system prevents the backlog from causing a crash.