3.802.1ag CFM
Last updated
Was this helpful?
Last updated
Was this helpful?
OAM is generally utilized for detecting and monitoring network faults and examining network status ,etc. think of ping and traceroute in Linux tools which are oam tools for Layer 2,and you may not know yet any L2 OAM tools .
Until recently Ethernet lacked OAM(Operation, Administration and Management) functionality which like that found in connection-oriented network(ATM),now it has two kinds of OAM, Ethernet link-layer OAM which is specified in 802.3ah by IEEE 802.3 work group ,and Ethernet service OAM by Y.1731(ITU)802.1ag(IEEE).note functionalities in Y.1731 and 802.1ag are not contradictory ,nor overlapped ,they are complementary.
why OAM? service provider needs it,and even user needs it.as you know ,Ethernet is that popular nowadays, service provider is involved into service provision to their customers ,both provider and customer should be confident about the Ethernet network status quo,but originally Ethernet is bare packet switching and connection-less ,I guess there is no such mechanism to support OAM functionality at all.as (carrier class)Ethernet evolves ,OAM for Ethernet appears .
802.3ah is for 'Ethernet in the first mile(EFM)' ,it operates on a single link, we will not tell any more here about link-layer OAM.
it comes to use when the provider needs to monitor end-to-end Ethernet service for its customer,you may remember Ethernet Service we introduced in Chapter 2 where SPB natively support EVPN(E-Line/E-LAN/E-Tree) ,what we do when we find a defect in the provider network,we could identify what customer is affected and try to reroute the traffic even.
Ethernet OAM is level hierarchical ,it supports eight levels ,it allows users ,service provider and operator to run OAM on their own level ,it means that OAM belongs to higher level will be forwarded transparently by lower level bridges.
the devices we monitor here we call them Maintenance Entity , it 's supposed to grouped into ME groups (MEG), the MEG end point which is located at the end of managed entities is MEG endpoint(MEP),while those in the middle are called MEG Intermediate Point(MIP) .they are relatively meaningful ,and MIP at high level may be MEP at the lower one.
the following figure show the layered end points.
for customer ,the provider must transparently forward the OAM packets which are originated from user end station.
802.1ag support three types of OAM message,continuity check message ,loopback message and link trace message ,continuity check is proactive ,once it's started ,it will periodically generates cc messages ,loopback message is similar to ping tool in Linux,and link trace is similar to traceroute in Linux.
802.1ag run on layer 2 directly with Ethertype 0x8902, then follows a 84-byte header plus variable length TLV tuples as shown below:
for level :
0-3 for operator
3-4 for provider
5-7 for customer
version is supposed to be zero. opcode is the indicator which indicates what type of message it is,it's enumerated as below:
0 reserved
1 continuity check message
2 loopback reply
3 loopback message
4 linktrace reply
5 linktrace message
... ...
flag indicate whether additional feature is enabled,TLV offset indicates the offset value from which the TLVs start.
here is a git reference which implements 802.1ag LB and LT function:https://github.com/chillancezen/dot1ag-utils, it's based on libpcap ,daemon listens on a specific interface and filter CFM pakcet only and then respond CFM formatted reply.
by listening on a interface:
from other end we start to L2 ping:
note that vethXYKV3Q and eth0 in mycentos1 is a linux veth pair,so the CFM packets will be directed to the other end in whatever cases.
last we inspect the packet by capturing it:
