Insight interface traffic management and QoS
An Insight-enabled port channel may be
the destination endpoint of multiple mirroring rules or forwarded traffic, resulting in a
very high rate of traffic. The maximum Insight Interface bandwidth on all platforms is
10Gbps, and although the TVPM runs on two vCPUs, the cumulative traffic may result in the
fast plane ASIC dropping some egress traffic at the Insight Interface.
From a traffic management perspective, QoS
for an Insight Interface is similar to QoS for a regular port. If required, SLX
conventional egress traffic rate-limiting or typical QoS features may be applied to the
Port Channel with an enabled Insight Interface. The difference is the QoS configuration
is applied to an Insight Interface LAG (port-channel). Consider the following when you
configure this feature:
- The TM supports egress scheduling,
rate shaping, WRED, and ingress buffer management for insight interface.
- TM egress scheduling and shaping
for the insight interface must be configured under a port-channel interface.
- The QoS configuration is
automatically applied to all ports in the port-channel.
- Follow the same configuration
procedures for ingress buffer management and WRED as you would with a
standard port.
For more information, see
Configuring QoS egress scheduling.
