Transcript
Ethernet OAM Troubleshooting
. Ethernet originated as a Local Area Network (LAN) technology
. A small number of co-located stations, were all managed by a single entity so E2E performance was never a real concern
. “Carrier Class Ethernet” radically changed the situation;
. Networks need to be managed and monitored by service providers in
order to guarantee SLAs
. Ethernet MANs need to support automated defect detection and
performance measurement
. IP-based tools such as ping and trace-route are not suitable
. IP tools function at higher layers and do not relate to the underlying network or service
. Four Ethernet OAM protocols have emerged;
. IEEE 802.3ah EFM link-layer OAM
. IEEE 802.1ag Connectivity Fault Management
. MEF E-LMI Ethernet Local Management Interface
. ITU-T Y.1731 OAM Functions and Mechanisms for Ethernet networks
Ethernet Virtual Connection (EVC)
.
A flat network is difficult to manage and to define accountabilities
.
Hierarchical Maintenance Domains defines OAM Flows & OAM responsibilities
UNI UNI
.
Maintenance Association (MA)
. Boundaries of an Administrator’s scope of monitoring part of the network
.
Maintenance Domain (MD)
. A level of monitoring within the hierarchy
.
Maintenance End Points (MEP)
. End Points of the MA or MD
.
Maintenance Intermediate Points (MIP)
. Intermediate Points within MA or MD
. Discover an OAM peer, and automatically detect its capabilities
. Indicate whether the test set serves in active or passive mode.
. Specify the Vendor OUI (Organizationally Unique Identifier) for the instrument
. Indicate whether the instrument will advertise that it provides unidirectional support for failure detection, remote loopback, link events, and variable retrieval
. Indicate whether you want the instrument to generate link faults, dying gasps, and critical events.
. Specification of the Maintenance Domain (MD) level, Maintenance Entity Group (MEG) End Point IDs, and Maintenance Association (MA) IDs.
. Specify the Continuity Check Message (CCM) transmission rate
. Specify the CCM and LBM Multicast address
. Specify thresholds for clearing a loss of continuity (LOC) if the number of consecutive missing CCM exceeds the number of messages expected within the calculated interval.
. NOTE: Link OAM PDUs will be sent together with Throughput test traffic. When OAM is enabled we will have the option to send 0% test traffic, in order to only send out OAM traffic. Most of the time it will be sent out with test traffic.
.
Each layer supports OAM capabilities independently
.
OAMs interoperate with each other
.
Component responsibilities are complementary
IEEE 802.3ah Link OAM
. Link-layer OAM was developed for “Ethernet in the First Mile” (EFM) . Provides functions for network segments (access links & individual circuits) . Operates purely at the Ethernet layer, and does not require an IP address . Link-layer OAM messages are sent in untagged slow protocol frames called
OAM Protocol Data Units, or OAMPDUs.
. Slow protocols are restricted in the number of protocol frames transmitted per second (for OAMPDUs . no more than 10 frames per second)
. Uses Ethertype 88-09 and is differentiated by a sub-type of 03 that appears as the first byte of the MAC client payload.
. OAMPDUs contain control and status information to monitor, test and
troubleshoot OAM-enabled links.
. Provides Link Fault Detection, Monitoring and Loopback functionalities
. Monitors individual links only, not end-to-end EVC
Defined by the flag and code fields;
. Information code
. OAMPDUs are used for auto-discovery, heartbeat, and fault notification
. Event notification OAM frames
. report various link statistics e.g. symbol errors and total frames with errors since the OAM sub-layer was reset
. Variable request and variable response frames
. Service provider MIB variable requests to obtain customer configuration
. Loopback control OAM frames
. used to enable or disable intrusive loopback in a remote passive-mode device
. Organization specific frames
. Discovery . Link Performance Monitoring . Remote Loopback . Fault Detection . Collecting Performance Statistics . Organizational Specific Extensions
. Discovery
. First phase of 802.3ah OAM protocol which identifies the 802.3ah capable link partner . Necessary for link partners to communicate their OAM capabilities and configurations because 802.3ah is a multi-vendor, non-proprietary protocol
. Link performance monitoring
. Detection and notification of link performance (quality) faults . Uses the Event Notification OAM Protocol Data Units (OAMPDUs) which are sent to OAM capable devices when error traffic is detected . Ethernet data performance can deteriorate slowly before disrupting service, so the 802.3ah standard enables a user to set the level of threshold that generates an event OAMPDU
. Remote Loopback
. An 802.3ah OAM-capable device can put its remote link partner into loopback mode using a loopback control OAMPDU . Every frame received is transmitted back on the same port to ensure the quality of links during installation or troubleshooting
. Fault Detection
. An OAMPDU flag allows an 802.3ah-capable device to convey severe events to its OAM link partner. 802.3ah OAM Detects and indicates the following conditions:
. Link Fault: Loss of link is detected by receiver . Dying Gasp: Unrecoverable condition such as power failure is sent to
the remote link partner . Critical Event: Organization specific critical event or major failure . Unidirectional Fault Detection
. Link OAM -Per link (point-to-point): . Link monitoring --Remote loopback . Remote failure indication --Loopback Control
. Good for single links, but it does not monitor across EVC !
UNIUNI
802.3ah
802.3ah
802.3ah
802.3ah
802.3ah
802.3ah
802.3ah
802.3ah
802.3ah
IEEE 802.1ag -“Connectivity Fault Management” across EVC
. Continuity Check Messages (CCM)
. Loopback Message (LBM)
. Link trace Message (LTM)
ITU-T Y.1731 . “OAM Functions & Mechanisms for Ethernet networks”
. Provides all of the 802.1ag functionality with additional performance monitoring capabilities including: . Frame Loss LM), Delay (DM) and Delay Variation Measurements . Automatic Protection Switching
Continuity Check Messages (CCM)
.MEPs periodically exchange Continuity Check OAMmessages to detect loss of continuity or incorrectnetwork connections.
.Multicast to each MEP in a MA/MEG at each
administrative level.
.Can be used to perform two way dual-ended Frame Loss measurements.
.Incorporates a Flag field which includes a bit forRemote Defect Indication (RDI) and an indication ofthe period at which CC Messages are transmitted.
Loopback Message (LBM)
. MEPs send loopback messages to verify connectivity with another MEP or MIP for a specific MA/MEG.
. Acts like a ping type request/reply function. A MEP sends a loopback request message to another MEP or MIP, which generates a subsequent LRM.
. LBMs / LRMs are used to verify bidirectional connectivity. They are typically initiated by operator command.
. A MEP can be provisioned to send LBMs periodically.
. For IEEE 802.1ag, loopback is a unicast OAM message.
. Loopback can also be used as an out of service diagnostic test. (applies to unicast loopback frames only)
Link Trace Message (LTM)
.MEPs multicast LTMs on a particular MA/MEG to identify adjacency relationships with remote MEPsand MIPs at the same administrative level.
.LTM can also be used for fault isolation. .The message body includes a destination MACaddress of a target MEP that terminates the linktrace. .When a MIP or MEP receives an LTM, it generates a unicast LTR to the initiating MEP. .It also forwards the LTM to the target MEPdestination MAC address. .An LTM effectively traces the path to the target MEP.
. AIS: Alarm Indication Signal, will initiate the transmission of AIS in the presence of certain defect conditions and will take the appropriate actions on its reception.
. RDI: Remote Defect Indication, it is used both for fault management and performance monitoring. RDI is transmitted on the detection of a defect and cleared when the defect clears by a MEP.
. APS: Automatic Protection Switching
. Ethernet Locked Signal: used to indicate that data traffic is intentionally being interrupted for administrative purposes like testing. It help to differentiate between a defect condition and controlled suspension of traffic.
.Performance Management: single and dual ended measurements
.Frame Loss
.Delay
.Delay Variation
.Throughput
Thank You
Any Questions?
Confidential & Proprietary Information of VeEX Inc.