Faultisolation, verification, and recovery(isolation and verification are providedby a
combinationof protocols, whilerecovery is the function of protocols such as spanning
tree)
Theloopback protocol used in Ethernet OAM is modeledon the standard IP ping. After
afault is detected, the loopbackprotocol performs fault verification and isolationunder
thedirection of a networkoperator. Theloopback is performed using request and response
messagepairs. A unicast loopbackmessage is generated by a MEP and a loopback reply
isgenerated by the destination MIP or MEP.The target MAC address is learnedby the
continuitycheck protocolor linktrace protocol. The loopbackmessage's packet is always
forwardedto a unique port by the originating MEP,as determined by a MAC tablelookup
orthe MEP interface MAC address. The targetMIP or MEP generates a unicast loopback
replyin response to the received loopbackmessage. The loopback message followsthe
samepath as a data packet, and intermediatebridges simply forward the packet to the
destinationMIP or MEP.
Related
Documentation
MXSeries Ethernet Services Routers SolutionsPage
EthernetOAM Connectivity Fault Managementon page 104
Example:Configuring Ethernet CFM over VPLS on page105
Example:Configuring Ethernet CFM on Bridge Connections on page112
Example:Configuring Ethernet CFM on Physical Interfaceson page 116
EthernetOAM Connectivity Fault Management
Themost complete connectivity faultmanagement (CFM) is defined in IEEE 802.1ag.
Thistopic emphasizes the use of CFM in a Metro Ethernet environment.
Themajor features of CFM are:
Faultmonitoring using the continuity check protocol.This is a neighbor discovery and
healthcheck protocol which discoversand maintains adjacencies at the VLAN or link
level.
Pathdiscoveryand fault verificationusing the linktrace protocol. Similar toIP traceroute,
thisprotocol maps the path taken to a destinationMAC address through one or more
bridgednetworks between the source and destination.
Faultisolation using the loopback protocol.Similar to IP ping, this protocol works with
thecontinuity check protocol during troubleshooting.
CFMpartitions the service network into various administrativedomains. For example,
operators,providers, and customersmay be part of different administrative domains.
Eachadministrative domain is mapped into one maintenancedomain providing enough
informationtoperform its own management, thus avoiding securitybreaches and making
end-to-endmonitoring possible. Each maintenance domain is associatedwith a
maintenancedomain level from 0 through 7.Level allocation is based on the network
hierarchy,where outermost domains are assigneda higher level than the innermost
domains.Customer end points have to highest maintenance domain level.In a CFM
Copyright© 2010, Juniper Networks,Inc.104
Junos10.4 MX Series Ethernet ServicesRouters Solutions Guide