With SPD, the
behavior of normal IPv6 packets is not changed. However, routing protocol
packets are given higher priority, because SPD recognizes routing protocol
packets by the IPv6 precedence field. Therefore, if the IPv6 precedence is set
to 6, then the packet is given priority.
SPD prioritizes IPv6
packets with a precedence of 6 by allowing the Cisco IOS software to queue them
into the process level input queue above the normal input queue limit. The
number of packets allowed in excess of the normal limit is called the SPD
headroom. The SPD headroom default is 100, which means that a high precedence
packet is not dropped if the size of the input hold queue is lower than 175
(which is the input queue default size + SPD headroom size).
Because Interior
Gateway Protocols (IGPs) and link stability are tenuous and crucial, such
packets are given the highest priority and are given extended SPD headroom with
a default of 10 packets. These packets are not dropped if the size of the input
hold queue is lower than 185 (input queue default size + SPD headroom size +
SPD extended headroom).
Non-IPv6 packets such
as Connectionless Network Service Intermediate System-to-Intermediate System
(CLNS IS-IS) packets, PPP packets, and High-Level Data Link Control (HDLC)
keepalives are treated as normal priority as a result of being Layer 2 instead
of Layer 3. In addition, IGPs operating at Layer 3 or higher are given priority
over normal IPv6 packets, but are given the same priority as Border Gateway
Protocol (BGP) packets. Therefore, during BGP convergence or during times of
very high BGP activity, IGP hellos and keepalives often are dropped, causing
IGP adjacencies to fail.