G.

8032
Ethernet Ring Protection Overview

March, 2008
ITU-T Q9 SG 15
Agenda
G.8032 Recommendation Introduction

G.8032 Objectives and Principles

G.8032 Concepts

G.8032 Protection Switching

G.8032 R-APS Messages

G.8032 Items Under Study

2
 ITU-T SG15/Q9 G.8032 Recommendation
G.8032 Recommendation consented February 2008
y This Recommendation defines the APS protocol and protection
switching mechanisms for ETH layer ring topologies. Included are
details pertaining to bridged ring protection characteristics, architectures
and the ring APS protocol.

G.8032 Proposals G.8032 Recommendation

and Contributions Consent

G.8032 Objective,
Principles, and G.8032 Converged Proposals
Requirements Setting

G.8032 ERP Work

Item Approved in
ITU-
ITU-T

J F M A M J J A S O N D J F M A M J J A S O N D J F M
A E A P A U U U E C O E A E A P A U U U E C O E A E A
N B R R Y N L G P T V C N B R R Y N L G P T V C N B R

2006 2007 2008

3
 G.8032 Objectives and Principles
Use of standard 802 MAC and OAM frames around the ring.
Uses standard 802.1Q (and amended Q bridges), but with
xSTP disabled.
Ring nodes supports standard FDB MAC learning, forwarding,
flush behaviour and port blocking/unblocking mechanisms.
Prevents loops within the ring by blocking one of the links
(either a pre-determined link or a failed link).
Monitoring of the ETH layer for discovery and identification of
Signal Failure (SF) conditions.
Protection and recovery switching within 50 ms for typical
rings.
Total communication for the protection mechanism should
consume a very small percentage of total available bandwidth.
4
G.8032 Terms and Concepts
Ring Protection Link (RPL) Link designated by mechanism that is
blocked during Idle state to prevent loop on Bridged ring
RPL Owner Node connected to RPL that blocks traffic on RPL during Idle
state and unblocks during Protected state
Link Monitoring Links of ring are monitored using standard ETH CC OAM
messages (CFM)
Signal Fail (SF) Signal Fail is declared when ETH trail signal fail condition
is detected
No Request (NR) No Request is declared when there are no outstanding
conditions (e.g., SF, etc.) on the node
Ring APS (R-APS) Messages Protocol messages defined in Y.1731 and
G.8032
Automatic Protection Switching (APS) Channel - Ring-wide VLAN
used exclusively for transmission of OAM messages including R-APS
messages
5
G.8032 Timers
G.8032 specifies the use of different timers to avoid race
conditions and unnecessary switching operations
WTR (Wait to Restore) Timer Used by the RPL Owner to
verify that the ring has stabilized before blocking the RPL after SF
Recovery
Hold-off Timers Used by underlying ETH layer to filter out
intermittent link faults
y Faults will only be reported to the ring protection mechanism if this timer
expires

6
Controlling the Protection Mechanism
Protection switching triggered by
Detection/clearing of Signal Failure (SF) by ETH CC OAM
Remote requests over R-APS channel (Y.1731)
Expiration of G.8032 timers
R-APS requests control the communication and states of the
ring nodes
Two basic R-APS messages specified - R-APS(SF) and R-APS(NR)
RPL Owner may modify the R-APS(NR) indicating the RPL is blocked:
R-APS(NR,RB)
Ring nodes may be in one of two states
Idle normal operation, no link/node faults detected in ring
Protecting Protection switching in effect after identifying a signal fault

7
Ring Idle State
A. Physical topology has all nodes ETH-CC ETH-CC
RPL RPL
connected in a ring ETH-CC ETH-CC Owner

B. ERP guarantees lack of loop by

ETH-CC
ETH-CC
ETH-CC

ETH-CC
blocking the RPL (link between 6 & 1
in figure)
ETH-CC ETH-CC
C. Logical topology has all nodes ETH-CC ETH-CC
connected without a loop.
D. Each link is monitored by its two
adjacent nodes using ETH CC OAM 2 1 6
RPL
messages
3 4 5
E. Signal Failure as defined in Y.1731, is
trigger to ring protection Physical topology
Loss of Continuity 2 1 6
Server layer failure (e.g. Phy Link
Down) 3 4 5
Logical topology
8
Protection Switching Link Failure
A. Link/node failure is detected by the
RPL RPL
nodes adjacent to the failure. Owner

B. The nodes adjacent to the failure,

block the failed link and report this

R-APS(SF)
failure to the ring using R-APS (SF) R-APS(SF) R-APS(SF)
message
C. R-APS (SF) message triggers R-APS(SF)

RPL Owner unblocks the RPL

All nodes perform FDB flushing
D. Ring is in protection state 2 1 6 2 1 6
RPL RPL
E. All nodes remain connected in the
3 4 5 3 4 5
logical topology. Physical topology

2 1 6 2 1 6

3 4 5 3 4 5
9
Logical topology
Protection Switching Failure Recovery
A. When the failed link recovers, the R-APS(NR, RB)
traffic is kept blocked on the nodes RPL RPL
Owner
adjacent to the recovered link

R-APS(NR,R-APS(NR)
B. The nodes adjacent to the
recovered link transmit R-

RB)
APS(NR) message indicating they R-APS(NR) R-APS(NR)

have no local request present

C. When the RPL Owner receives R- R-APS(NR)

APS(NR) message it Starts WTR

timer
2 1 6 2 1 6
D. Once WTR timer expires, RPL RPL RPL
Owner blocks RPL and transmits
R-APS (NR, RB) message 3 4 5 3 4 5
E. Nodes receiving the message Physical topology
perform a FDB Flush and unblock 2 1 6 2 1 6
their previously blocked ports
F. Ring is now returned to Idle state 3 4 5 3 4 5
10 Logical topology
 Signaling Channel Information
ERP uses R-APS messages to manage and coordinate the protection
switching
R-APS defined in Y.1731 - OAM common fields are defined in Y.1731.
Version 00000 for this version of Recommendation
OpCode defined to be 40 in Y.1731
Flags 00000000 should be ignored by ERP

1 2 3 4

8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1

1 MEL Version (0) OpCode (R-APS = 40) Flags (0) TLV Offset (32)
5 R-APS Specific Information (32 octets)
..

37 [optional TLV starts here; otherwise End TLV]

last End TLV (0)

Defined by Y.1731 Defined by G.8032 Non-specified content

11
 R-APS Specific Information
Specific information (32octets) defined by G.8032
Request/Status(4bits) 1011 = SF | 0000 = NR | Other = Future
Status RB (1bit) Set when RPL is blocked (used by RPL Owner in NR)
Status DNF (1bit) Set when FDB Flush is not necessary (Future)
NodeID (6octets) MAC address of message source node (Informational)
Reserved1(4bits), Status Reserved(6bits), Reserved2(24octets) - Future
development

1 2 3 4
8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1 8 7 6 5 4 3 2 1
Request /State Reserved 1 Status Node ID (6 octets)

R D Status Reserved
B N
F
(Node ID)

Reserved 2 (24 octets)

12
 Items Under Study
G.8032 is currently an initial recommendation that will continue to be
enhanced. The following topics are under study for future versions of the
recommendation:
a) Interconnected rings scenarios: shared node, shared links
b) RPL blocked at both ends configuration of the ring where both nodes connected
to the RPL control the protection mechanism
c) Support for Manual Switch administrative decision to close down a link and force
a recovery situation are necessary for network maintenance
d) Support for Signal Degrade scenarios SD situations need special consideration
for any protection mechanism
e) Non-revertive mode Allows the network to remain in recovery configuration
either until a new signal failure or administrative switching
f) RPL Displacement Displacement of the role of the RPL to another ring link
flexibly in the normal (idle) condition
g) In-depth analysis of different optimizations (e.g., FDB flushing)
h) Etc.

13
Thank You

14
Backup

15
G.8032 Basic Functions

Robustness: Source Steering Protection Connectivity: Unicast, MC and BC Data Transfers

Failure Unicast
Flow
Multicast
Flow

50ms Switch Times

16
ERP Functionality Decomposition
SF / clear SF Local Defect ETH_CI_SSF (E)
Top Priority Logic (E)
Local Local
Local Request
Requests Priority
(operator) Logic SF / Clear SF Local Defect ETH_CI_SSF (W)
Logic (W)

Validity Request/State Request/State

ETH_CI_RAPS (E) Guard
Check + Status + Status Priority WTR Running WTR
Timer Logic WTR Expire Timer
ETH_CI_RAPS (W)
Top Priority Request

Start/ Stop
Start/Stop
R-APS
R-APS Request/State + Status ETH_CI_RAPS (E)
R-APS
Request Message ETH_CI_RAPS (W)
MI_RPL_Owner Processing Stop Tx R-APS
Transmission

Block/Unblock ring ports

Flush FDB

Inputs Local node requests, ETH signals, ERP Processing R-APS Request
R-APS messages, WTR Timer. Processing processes filtered message
Filters Local Priority Logic, Validity Check, according to State Machine with
Guard Timer, Priority Logic guarantee that knowledge of current state and RPL-
highest priority message arrives for Owner indication
processing. Outputs Node functions (FDB
Flushing, Block/Unblock port), sending
17
(or stopping) R-APS messages
State Machine
Outputs
Inputs
Three node states
Node High Priority Row Actions Next
y Initialization when first defining the ring State request # node
state
y A (Idle) the normal state of the ring nodes
Stop guard timer
with RPL blocked and all nodes/ports working * Initialization 0
Stop WTR timer

y B (Protecting) when protection switching is If RPL Owner:

Block RPL port
in effect, RPL unblocked, other (usually fault) Unblock non-RPL port
Tx R-APS (NR, RB) B
link is blocked. Else:
Block both ports
Different input messages assigned priority Stop Tx R-APS

as appears in the State Machine

description. Local SF 1 Block failed port;
Unblock non Failed port;
B
Priority mechanism to allow proper
Tx R-APS(SF);
Flush FDB;

reaction to faults Local Clear SF 2 No action

A
State Machine describes the actions to be R-APS(SF) 3 Unblock non Failed port;

taken by the node dependent on current A (Idle)

Stop Tx R-APS
If not DNF Flush FDB;
B

state and input message. No action

WTR Expires 4 A
y Only message with highest priority will pass
WTR Running 5 No action A
through the filtering.
R-APS(NR, RB) 6 Unblock non-RPL port A
y Different actions include transmission of R-
R-APS(NR) 7 No action
APS message, blocking/unblocking a port, A

flushing the FDB, switching current state of

node, and starting/stopping timers.
18
State Machine
FDB flushing will clear all of the learned Inputs
Outputs

filtering rules within the node. Node High Priority Row Actions Next
y Current Recommendation includes basic State request # node
state
optimization only flushing once on fault
Local SF 8 Block failed port;
discovery and once on recovery. Unblock non Failed port;
B
Stop WTR;
y DNF status flag is for future definition Tx R-APS(SF);

Local Clear SF 9 Start guard timer,

B
Tx R-APS(NR),

R-APS(SF) 10 Stop WTR,

Unblock non Failed port,
B
Stop Tx R-APS

B
WTR Expires 11 Block RPL port,
(Prote Unblock non-RPL port
cting) A
Tx R-APS (NR,RB)
Flush FDB,

WTR Running 12 No action B

R-APS(NR, RB) 13 If not RPL Owner:

Unblock both ports,
A
Stop Tx R-APS,
If not DNF Flush FDB

R-APS(NR) 14 If RPL Owner:

B
Start WTR

19
Switching Triggers
Detection/Clearing of SF (Signal Fail):
y SF: Loss of CCM, SF signal passed up from underlying (sub)layer
y Switching is performed when hold-off timer expires after detection of SF.
Operators request:
y Lockout, Forced switch, Manual switch, Clear (for future specification)
Remote request:
y Switching request indicated in received R-APS information from the other
node.

Protection switching algorithm is based on priorities assigned to all

triggers

20
 Ethernet Rings and the RPL
Ethernet nodes connected in a ring topology.
y Each node connected to two neighbors via links (E & W ports)
y Deterministic network topology

Avoid loops within ring ETH_FF ETH_C

ERP Control
Process

by blocking a
ring
designated link (RPL) node
ERP

cont rol
RPL
connect ion
A
y Node responsible to ERP Cnt rol
Process
point
(blocked)
ERP Cont rol
Process

control the blocking ETH_FF

RP
ETH_FF

ERP
ERP

state of RPL is RPL

cont rol
control
ring
prot
link ect i
r ing on
owner link
(RP
L)
y RPL is unblocked
r ing
during protection state link link
ring
node

node
ring

ring
and may revert to

D
B

ETH_C
ETH_C

blocked during recovery RPL Owner

ring control
node
C
ERP

ERP Cont rol

ETH_C ETH_FF Process

21
 Scenario A - Normal to Protection
RPL RPL Owner
A B C D E F G

Normal
A
B failure

C
D
Flush SF SF SF

50 ms
SF SF Flush
Flush Flush Flush Flush
E Flush

F
SF SF
Protection SF SF SF
state SF
G

A. Normal State
B. Failure Occurs E. RPL Owner receives SF message and unblocks its end
of RPL link [row 3]
C. Nodes C and D detect local Signal Failure condition F. Stable State SF messages on the ring
and after respecting hold-off time and Block failed port
[row 1] G. Further SF messages trigger no further action [row10]
D. Node C and D periodically send SF message, on both
ring ports, while SF condition persists. Each node Message source R-APS ch block
performs a FDB Flush when receiving the SF message
Client ch block

22
 Scenario B recovery
RPL Owner
RPL
A B C D E F G
failure
A
SF SF
SF SF
SF SF
B recovery
C

Confirmation time
NR NR
Protection NR NR NR
NR
state D

F
NR,
NR, RPL Blocked

50 ms
NR, RPL Blocked
RPL Blocked Flush
G Flush Flush
Flush
Normal Flush
H Flush Flush NR,
NR Blocked
RPL
NR, NR,
RPL Blocked RPL Blocked

A. Stable SF condition Nodes C and D continue to send E. When node C and D expire the guard timer, they may
SF message every 5 secs. accept the new R-APS messages that they receive. [in
B. Recovery of Link failure guard timer]
C. Nodes C and D detect clearing of SF condition , start F. At expiration of WTR timer, RPL owner blocks its end
guard timer and initiate periodical transmission of NR of the RPL, sends, NR RB message [row 11]
message on both ring ports. ( Guard timer prevents G. Each node that receives the NR,RB message it flushes
reception of R-APS messages) [row 9] its FDB. When Nodes C and D receive NR RB
D. When the RPL Owner receives NR message, it starts message , they remove block on their blocked ports.
the WTR timer [row 14] [row 13]
H. 23 Stable normal condition all nodes in Idle state