Configuring the ANCP Agent to Dampen the Effects of Short-Term Adjacency Losses
By default, the ANCP agent treats a loss of adjacency as if it has received a Port Down message for every access loop that is represented by the adjacency. All Layer 2 wholesale sessions are logged out and cleaned up. If the associated physical interface is in the Down state, then pending sessions cannot be reestablished when the interface transitions back to the Up state.
You can configure the ANCP agent to maintain the corresponding ANCP-triggered Layer 2 wholesale sessions for a configurable period in the event that an ANCP adjacency is lost. If the adjacency is restored before the timer expires, the session continues. If the timer expires before the adjacency is restored, then the session is logged out and cleaned up. This behavior dampens the effect of unstable ANCP connections. The hold timer can also detect when an access line is unconfigured on a neighbor and trigger logout and cleanup of the related sessions.
The default value of the timer is 0, which means that the loss of neighbor adjacency immediately triggers a logout of all corresponding Layer 2 wholesale sessions.
To configure how long the ANCP agent maintains sessions in the event of an adjacency loss for any neighbor:
Specify the hold timer duration in seconds.
[edit protocols ancp] user@host# set adjacency-loss-hold-time seconds
To configure how long the ANCP agent maintains sessions in the event of an adjacency loss for a specific neighbor:
Specify the hold timer duration in seconds.
[edit protocols ancp neighbor ip-address] user@host# set adjacency-loss-hold-time seconds