Latency-based optimization
Link rerouting
After injecting latency on a particular link beyond the configured threshold (i.e. 90,000 microseconds in this example), we can see that the LSPs that were using the link have been re-routed, bypassing the link with the high latency value.
Note: To demonstrate this particular use case on a lab setup, the link latency was manually configured on a link via API, overriding the value obtained from the OAM tests. In a real network, if there is congestion or network-related performance problems, the latency measurements coming from the OAM source are expected to fluctuate, sometimes exceeding the configured link threshold values.
Example, before optimization:
The latency measurement from the API, 120000 microseconds, is taking precedence over the OAM measurement of 1608 microseconds.
Example, after optimization:
The LSPs are no longer using this link to reach the destination.
View the LSP in the Path Control, LSPs view to see the new set of hops. For reference, the previous hops were:
