Pseudowire OAM Because pseudowires are carried over MPLS LSPs, the pseudowire
architecture assumes that network faults will be corrected at the MPLS layer. However,
it is still essential for a PE to be able to inform a remote PE if its local attachment
circuit goes down or if it becomes aware that the LSP to or from that remote PE has
failed (the remote PE may be unaware of the LSP failure). In the general case, the PE
also informs the device connected to the local attachment circuit that there is a fault;
however, in the Ethernet VLAN case, there is no mechanism by which a PE may notify
a CE of a fault (when doing a port-mode Ethernet pseudowire, the PE may disable the
link), though the MEF is currently working on an Ethernet LMI to address this.
The initial versions of draft-martini relied on the PE sending a label withdraw for the
pseudowire to the remote PE to notify it of failures. Thus, there were only two states
that could be distinguished: pseudowire up and pseudowire down. There was no means
of identifying whether a pseudowire was down due to administrative purposes or due
to a fault.
The next step in pseudowire OAM was the addition of the status TLV.
Pages:
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992