In the MS-PW
case, however, the MPLS layer now only provides connectivity from T-PE to S-PE and
from one S-PE to another. This creates two challenges:
?– The service provider must define a-priori the set of S-PEs through which a pseudowire
will flow from T-PE to T-PE and must configure each one to switch the
pseudowire.
?– Failure of an S-PE now implies failure of the pseudowire.
To address these challenges, work is ongoing in the IETF on ???dynamic placement??? of
MS-PWs. In this model, globally unique FEC 129 Type 2 AIIs are constructed, consisting
of a 32-bit global ID, a 32-bit prefix (typically, the IP address of the T-PE where the
attachment circuit is located) and a 32-bit Attachment Circuit ID. When setting up a
dynamic MS-PW, a source T-PE performs a longest prefix match (similar to the longest
MPLS 465
prefix match used in IP forwarding) on its local Layer 2 AII PW routing table. Each
entry in the table consists of a 96-bit Type 2 AII with a mask; for example, an entry with
a /64 mask (i.e., 64 ones and 32 zeroes) would match all Attachment Circuit IDs for a
given global ID and prefix, and a matching next hop (i.e.
Pages:
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048