Between the distribution layer and the
core, you can employ a technique such as the one described in ???Bridge Gateways??? so
that the distribution layer bridges can determine which of the links to the core are used
for which EVCs. From there on down, there is only a single path from paired virtual
bridge to paired virtual bridge available; hence no routing protocol is necessary.
Of course, to get from one distribution layer paired virtual bridge to another, the data
must pass through the core. However, this structure is extremely robust, failures and
recoveries are handled extremely quickly, and the simplicity of the network enables the
operator to configure quality of service or bandwidth guarantees easily.
Ethernet OAM and Connectivity Fault Management
There are three versions of Ethernet OAM, all of which can be very useful in conjunction
with any Ethernet-over-xyz technology, not just Ethernet bridging. IEEE Std
802.3ah-2004 defines an Ethernet OAM capability for a single point-to-point physical
IEEE 802.3 link. It provides for keep-alive pings at one-second intervals; a remotely
Figure 13.14 Paired virtual bridges
???Core??? network
A
C
B
D
E1 E2
Virtual paired bridge F
F1 F2
Virtual paired bridge K
K1 K2
Virtual paired bridge A
P1 P2
Virtual paired bridge B
Q1 Q2
Virtual paired bridge Z
Z1 Z2
.
Pages:
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921