MPLS Working Group H. Zhang
Internet Draft J. He
Intended status: Standards Track Huawei
H. Li
China Mobile
March 8, 2010
Expires: September 2010
SD-Triggered Protection Switching in MPLS-TP
draft-zhl-mpls-tp-sd-02.txt
Abstract
In MPLS-TP survivability framework, a fault condition includes both
Signal Failure (SF) and Signal Degrade (SD) that can be used to
trigger protection switching. This document describes Signal Degrade
(SD) detection and the consequent protection switching mechanism.
Status of this Memo
This Internet-Draft is submitted to IETF in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt
The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html
This Internet-Draft will expire on April 8, 2010.
Copyright Notice
Copyright (c) 2010 IETF Trust and the persons identified as the
document authors. All rights reserved.
Zhang and He Expires September 8, 2010 [Page 1]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(http://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with
respect to this document.
Table of Contents
1. Introduction.....................................................2
2. Terminology......................................................3
3. SD-Triggered Protection Switching Architecture...................4
4. SD-Triggered Protection Solution Analysis........................4
4.1. SD-Triggered Protection Based on OAM packets................5
4.1.1. Detection of SD........................................5
4.1.2. APS Protocol...........................................5
4.2. Broadcast Bridge with Special APS Request Priority..........6
4.2.1. Protection Switching...................................6
4.2.2. APS Protocol...........................................7
4.3. Broadcast Bridge with Extra Protection Switching Coordination
................................................................7
4.3.1. Protection Switching...................................8
4.3.2. APS Protocol...........................................8
4.4. Analysis....................................................8
5. Security Considerations..........................................8
6. IANA Considerations..............................................8
7. Acknowledgments..................................................9
8. References......................................................10
8.1. Normative References.......................................10
8.2. Informative References.....................................10
Author's Addresses.................................................10
1. Introduction
In packet transport network, protection switching can be triggered by
fault conditions and external manual commands. The fault conditions
include Signal Failure (SF) and Signal Degrade (SD). SF on a
transport entity can be detected by fault management OAM functions;
SD on a transport entity can be detected by performance monitoring
OAM functions.
The SD condition used for protection switching mostly depends on
packet loss ratio. For some services that are sensitive to time and
synchronization, the SD condition may depend on packet loss ratio,
packet delay and delay variation.
Zhang and He Expires September 8, 2010 [Page 2]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
In MPLS-TP OAM tools, the packet loss measurement (LM) is used to
measure the amount of the lost service packets and compute the loss
ratio of service packet. The packet Delay Measurement (DM) is used to
measure the packet delay and delay variation by sending periodic DM
packets during the diagnostic interval. The LM and DM mechanisms are
out of scope of this document.
The detection of SD (e.g. packet loss) must be based on service
packets. But in some situation, there is no normal traffic on
working/protection transport entity, which makes the detection of SD
based on service packets impossible and causes switch flapping.
This document describes the mechanism for SD detection and consequent
protection switching in 1:1 and 1:n protection architecture.
2. Terminology
The reader is assumed to be familiar with the terminology in MPLS-TP.
The relationship between ITU-T and IETF terminologies on MPLS-TP can
be found in [Rosetta stone].
MPLS-TP: MPLS Transport Profile
SF: Signal Failure
SF-W: SF for working entity
SF-P: SF for protection entity
SD: Signal Degrade
SD-W: SD for working entity
SD-P: SD for protection entity
APS: Automatic Protection Switching
OAM: Operations, Administration and Maintenance
CC/CV: Continuity Check/Connectivity Verification
LM: Loss Measurement
DM: Delay Measurement
Zhang and He Expires September 8, 2010 [Page 3]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
3. SD-Triggered Protection Switching Architecture
The SD-triggered protection switching complies with general
protection architecture.
In case of MPLS-TP 1+1 protection architecture, the normal traffic is
permanently sent on both working and protection entities by the
permanent bridge at the source. Therefore, the detection of SD or the
clearing of SD on both working and protection entities can be based
on the characteristics of the normal traffic.
In case of MPLS-TP 1:1 and 1:n protection architecture, the selector
bridge is usually adopted at the source end. The normal traffic is
transported on the working entity together with the OAM of the
working entity, while on the protection entity only the OAM of the
protection entity is transported alone so that it may not be possible
to detect SD. If SD is detected on a working entity, protection
switching is occurred, and the traffic is transported on the
protection entity together with OAM. On the working entity OAM
packets is transported alone, and the detected SD may be cleared even
if the working entity is still degraded. This will result in another
switch of traffic back to the working entity, and SD may be detected
again, etc.
As mentioned above, in case of 1:1 and 1:n protection architecture,
the normal traffic will be either on the working transport entity or
on the protection transport entity. This makes it impossible to
detect SD on the standby entity. Consequently SD will be detected
only after a protection switch and flapping may happen.
4. SD-Triggered Protection Solution Analysis
In case of 1:1 and 1:n protection architecture, there are the
following solutions to implement the SD-triggered protection
switching without flapping.
- The detection of SD is based on OAM packets; (Section 4.1)
- The broadcast bridge has to be applied to enable SD detection in
SD-triggered protection, together with
- the special APS request priority (Section 4.2), or
- the extra protection switching coordination (Section 4.3).
The broadcasting is applied only in case protection switching is
active.
Zhang and He Expires September 8, 2010 [Page 4]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
4.1. SD-Triggered Protection Based on OAM packets
4.1.1. Detection of SD
In case of 1:1 and 1:n protection architecture, for the transport
entity on which there is no normal traffic, the detection of SD can
be based on OAM packets, e.g.
- OAM packets for Test
The Test OAM packets can be used to simulate the characteristics of
the normal traffic and replace the normal service. With the
performance monitoring OAM packets (LM) the SD condition of the
transport entity can be detected.
- OAM packets for CC/CV
The CC/CV OAM packets can be used instead of normal service in packet
loss measurement. That is, the performance monitoring OAM packets (LM)
will count CC/CV packets and the loss measurement is based on CC/CV
packets, which will be used as the input of SD condition of the
transport entity where there is no normal service.
The above-mentioned SD-triggered protection switching mechanism based
on OAM packets (test or CC/CV) is general and flexible without
constraint to the bridge type at source end, that may be selector
bridge or broadcast bridge.
The performance measurement by Test OAM packets is accurate. But the
usage of test packets on the protection entity defeats the objective
of the 1:1 and 1:n architectures, which is to have the protection
entity bandwidth available for best effort traffic during the time
there is no fault or degradation of the working transport entity. The
test packets consume now this bandwidth. For the case of the 1:1
protection, this makes the bandwidth usage of this 1:1 architecture
similar to the bandwidth usage of the 1+1 architecture.
OAM packets for CC/CV are sent in fixed transmission period (3.3ms,
10ms, 1s, etc.) which doesn't exactly reflect the condition of real
services. It is applicable only in places without strict requirement
for SD measurement. The detection of SD by CC/CV is not recommended
when the packet loss is caused by congestion.
4.1.2. APS Protocol
In APS request types, similar to the definition of SF-W (SF for
working entity) and SF-P (SF for protection entity), a definition of
Zhang and He Expires September 8, 2010 [Page 5]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
SD-W (SD for working entity) and SD-P (SD for protection entity) is
required to prevent flapping.
The priority of SD-P and SD-W shall be fixed as SD-P > SD-W similar
to SF-P > SF-W. In this case a protection switching based on SD
detection on the working entity shall not be initiated if there
exists also an SD condition on the protection transport entity.
4.2. Broadcast Bridge with Special APS Request Priority
SD-triggered protection based on normal traffic can be implemented by
adopting broadcast bridge at source end with the special APS request
priority.
4.2.1. Protection Switching
+----------+ +---------+
| | | |
---->----|---+------+------------>------------+---\ |
Source | | | Working Entity | \ | Sink
| | / | | +---|---->----
| | / | | |
| +- ---+------------>------------+--- |
| | Protection Entity | |
| | | |
+----------+ +---------+
Figure 1 Normal Condition in 1:1 Protection Architecture
In the normal state, the normal traffic is sent only on the working
transport entity and only the SD condition of the working transport
entity can be evaluated (see Figure 1).
Zhang and He Expires September 8, 2010 [Page 6]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
+----------+ +---------+
| | | |
---->----|---+------+--------SD-->------------+--- |
Source | | | Working Entity | | Sink
| | | | +---|---->----
| | | | / |
| +-+----+------------>------------+---/ |
| | Protection Entity | |
| | | |
+----------+ +---------+
Figure 2 SD on Working Entity in 1:1 Protection Architecture
When SD is detected on the working transport entity, the sink end
sends SD-W indication to the source end and the selector at the sink
end switches to the protection transport entity. The broadcast bridge
at the source end will then send the normal traffic on both working
and protection transport entities and the performance of both working
and protection transport entities can be monitored for SD (see Figure
2).
If SD is detected on the protection transport entity as well, i.e.
SD-W and SD-P exist simultaneously, normal traffic will still be
selected at the sink from the protection transport entity to avoid
flapping between protection and working states.
4.2.2. APS Protocol
In this case the priority of SD-W and SD-P in the APS protocol is
fixed as SD-W > SD-P to avoid flapping between protection entity and
working entity.
4.3. Broadcast Bridge with Extra Protection Switching Coordination
SD-triggered protection based on normal traffic can be implemented by
adopting broadcast bridge at source end with the extra protection
switching coordination.
The SD-W based protection switch action described in section 4.2.1
assumes that the probability of SD condition occurring on both
working entity and protection entity at the same time is very small.
When this assumption is not considered to be reasonable, the
operation of the selector may be modified as described hereafter.
Zhang and He Expires September 8, 2010 [Page 7]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
4.3.1. Protection Switching
+----------+ +---------+
| | | |
---->----|---+------+--------SD-->------------+---\ |
Source | | | Working Entity | \ | Sink
| | | | +---|---->----
| | | | |
| +-+----+------------>------------+--- |
| | Protection Entity | |
| | | |
+----------+ +---------+
Figure 3 SD on Working Entity in 1:1 Protection Architecture
When the sink end detects an SD condition, it does not switch to the
protection entity immediately. Instead, the broadcast bridge at the
source end will first send the normal traffic on both working and
protection transport entity after receiving SD-W indication sent by
the sink end (see Figure 3). Then sink end is able to detect the SD
condition of working and protection transport entity.
If SD is detected on the protection transport entity as well, the
normal traffic remains broadcasted to both working and protection
transport entities and is selected from the working transport entity;
if no SD is detected on the protection transport entity the normal
traffic is selected from the protection entity.
4.3.2. APS Protocol
The SD priority is the same as described in 4.1.2: SD-P > SD-W.
4.4. Analysis
In section 4.1, 4.2 and 4.3, the different mechanisms for detecting
SD are described. The mechanism described in 4.2 is preferred because
of its simplicity, accuracy and flexibility.
5. Security Considerations
To be added in a future version of the document.
6. IANA Considerations
IANA is requested to allocate two further APS request codes as
follows:
Zhang and He Expires September 8, 2010 [Page 8]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
xx Signal Degradation for Working entity (SD-W);
yy Signal Degradation for Protection entity (SD-P).
7. Acknowledgments
The authors would like to thank Huub van Helvoort for his input to
and review of the current document.
Zhang and He Expires September 8, 2010 [Page 9]
Internet-Draft draft-zhl-mpls-tp-sd-02.txt March 2010
8. References
8.1. Normative References
[RFC5654] Niven-Jenkins,B., Brungard,D., and Betts,M., "Requirements
of an MPLS Transport Profile", RFC5654, September 2009
[ITU-T Recommendation G.808.1 Amd1] ''Generic protection switching - -
Linear trail and subnetwork protection'', ITU-T G.808.1
Amendment 1, January 2009
8.2. Informative References
[MPLS-TP Framework] Bocci,M., and Bryant,S., "A Framework for MPLS in
Transport Networks", draft-ietf-mpls-tp-framework-10 (Work
in progress), February 2010
[MPLS-TP Survive Frmk] Sprecher,N., and Farrel,a., ''Multiprotocol
Label Switching Transport Profile Survivability Framework",
draft-ietf-mpls-tp-survive-fwk-03(work in progress),
November 2009
[MPLS-TP OAM Requirements] Vigoureux,M., Ward,D., and Betts,M.,
''Requirements for OAM in MPLS Transport Networks'', draft-
ietf-mpls-tp-oam-requirements-06(work in progress), March
2010
Author's Addresses
Haiyan Zhang
Huawei Technologies Co., Ltd.
Phone: +86-755-28972333
Email: zhanghaiyan@huawei.com
Jia He
Huawei Technologies Co., Ltd.
Phone: +86-755-28972333
Email: hejia@huawei.com
Han Li
China Mobile Communications Corporation
Email: lihan@chinamobile.com
Zhang and He Expires September 8, 2010 [Page 10]