| MPLS Working Group | J. Ryoo, Ed. |
| Internet-Draft | ETRI |
| Updates: 6378 (if approved) | E. Gray, Ed. |
| Intended status: Standards Track | Ericsson |
| Expires: September 28, 2014 | H. van Helvoort |
| Huawei Technologies | |
| A. D'Alessandro | |
| Telecom Italia | |
| T. Cheung | |
| ETRI | |
| E. Osborne | |
| March 27, 2014 |
MPLS Transport Profile (MPLS-TP) Linear Protection to Match the Operational Expectations of SDH, OTN and Ethernet Transport Network Operators
draft-ietf-mpls-tp-psc-itu-04.txt
This document describes alternate mechanisms to perform some of the functions of MPLS Transport Profile (MPLS-TP) linear protection defined in RFC 6378, and also defines additional mechanisms. The purpose of these alternate and additional mechanisms is to provide operator control and experience that more closely models the behavior of linear protection seen in other transport networks.
This document also introduces capabilities and modes for linear protection. A capability is an individual behavior, and a mode is a particular combination of capabilities. Two modes are defined in this document: Protection State Coordination (PSC) mode and Automatic Protection Switching (APS) mode.
This document describes the behavior of the PSC protocol including priority logic and state machine when all the capabilities associated with the APS mode are enabled.
This document updates RFC 6378 in that the capability advertisement method defined here is an addition to that document.
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at http://datatracker.ietf.org/drafts/current/.
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."
This Internet-Draft will expire on September 28, 2014.
Copyright (c) 2014 IETF Trust and the persons identified as the document authors. All rights reserved.
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. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.
Linear protection mechanisms for the MPLS Transport Profile (MPLS-TP) are described in RFC 6378 [RFC6378] to meet the requirements described in RFC 5654 [RFC5654].
This document describes alternate mechanisms to perform some of the functions of linear protection, and also defines additional mechanisms. The purpose of these alternate and additional mechanisms is to provide operator control and experience that more closely models the behavior of linear protection seen in other transport networks, such as Synchronous Digital Hierarchy (SDH), Optical Transport Network (OTN) and Ethernet transport networks. Linear protection for SDH, OTN, and Ethernet transport networks are defined in ITU-T Recommendations G.841 [G841], G.873.1 [G873.1] and G.8031 [G8031], respectively.
The reader of this document is assumed to be familiar with [RFC6378].
The alternative mechanisms described in this document are for the following capabilities:
and the following capabilities have been added to define additional mechanisms:
The priority modification includes raising the priority of Signal Fail on Protection path (SF-P) relative to Forced Switch (FS), and raising the priority level of Clear Signal Fail (SFc) above SF-P.
Non-revertive behavior is modified to align with the behavior defined in RFC 4427 [RFC4427] as well as to follow the behavior of linear protection seen in other transport networks.
Support of MS-W command to revert traffic to the working path in non-revertive operation is covered in this document.
Support of protection switching protocol against SD is covered in this document. The specifics for the method of identifying SD is out of the scope for this document and is treated similarly to Signal Fail (SF) in [RFC6378].
Support of EXER command to test if the Protection State Coordination (PSC) communication is operating correctly is also covered in this document. EXER command tests and validates the linear protection mechanism and PSC protocol including the aliveness of the priority logic, the PSC state machine and the PSC message generation and reception, and the integrity of the protection path, without triggering the actual traffic switching.
This document introduces capabilities and modes. A capability is an individual behavior. The capabilities of a node are advertised using the method given in this document. A mode is a particular combination of capabilities. Two modes are defined in this document: PSC mode and Automatic Protection Switching (APS) mode.
Other modes may be defined as new combinations of the capabilities defined in this document or through the definition of additional capabilities. In either case, the specification defining a new mode will be responsible for documenting the behavior, the priority logic, and the state machine of the PSC protocol when the set of capabilities in the new mode are enabled.
This document describes the behavior, the priority logic, and the state machine of the PSC protocol when all the capabilities associated with the APS mode are enabled. The PSC protocol behavior for the PSC mode is as defined in [RFC6378].
This document updates [RFC6378] by adding a capability advertisement mechanism. It is recommended that existing implementations of the PSC protocol be updated to support this capability. Backward compatibility with existing implementations that do not support this mechanism is described in Section 9.2.1.
Implementations are expected to be configured to support a specific set of capabilities (a mode) and to reject messages that indicate the use of a different set of capabilities (a different mode). Thus, the capabilities advertisement is not a negotiation, but a verification that peers are using the same mode.
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119].
This document uses the following acronyms:
| APS | Automatic Protection Switching |
| DNR | Do-not-Revert |
| EXER | Exercise |
| FS | Forced Switch |
| LO | Lockout of protection |
| MS | Manual Switch |
| MS-P | Manual Switch to Protection path |
| MS-W | Manual Switch to Working path |
| MPLS-TP | MPLS Transport Profile |
| NR | No Request |
| OC | Operator Clear |
| OTN | Optical Transport Network |
| PSC | Protection State Coordination |
| RR | Reverse Request |
| SD | Signal Degrade |
| SDH | Synchronous Digital Hierarchy |
| SD-P | Signal Degrade on Protection path |
| SD-W | Signal Degrade on Working path |
| SF | Signal Fail |
| SFc | Clear Signal Fail |
| SFDc | Clear Signal Fail or Degrade |
| SF-P | Signal Fail on Protection path |
| SF-W | Signal Fail on Working path |
| WTR | Wait-to-Restore |
[RFC6378] defines the priority of FS to be higher than that of SF-P. That document also defines the priority of Clear SF (SFc) to be low. This document defines the priority modification capability whereby the relative priorities of FS and SF-P are swapped and the priority of Clear SF (SFc) is raised. In addition, this capability introduces the Freeze command as described in Appendix C. The rationale for these changes is detailed in the following sub-sections from both the technical and network operational aspects.
Defining the priority of FS higher than that of SF-P can result in a situation where the protected traffic is taken out-of-service. When the protection path fails PSC communication may stop as a result. In this case, if any input that is supposed to be signaled to the other end has a higher priority than SF-P then this can result in unpredictable protection switching state. An example scenario that may result in an out-of-service situation is presented in Appendix A of this document.
According to Section 2.4 of [RFC5654] it MUST be possible to operate an MPLS-TP network without using a control plane. This means that the PSC communication channel is very important for the transfer of external switching commands (e.g., FS), and these commands should not rely on the presence of a control plane. In consequence, the failure of the PSC communication channel has higher priority than FS.
In other transport networks (such as SDH, OTN, and Ethernet transport networks) the priority of SF-P has been higher than that of FS. It is therefore important to offer network operators the option of having the same behavior in their MPLS-TP networks so that they can have the same operational protection switching behavior to which they have become accustomed. Typically, FS command is issued before network maintenance jobs, (e.g., replacing optical cables or other network components). When an operator pulls out a cable on the protection path, by mistake, the traffic should continue to be protected and the operator expects this behavior based on his/her experience on the traditional transport network operations.
The priority level of SFc defined in [RFC6378] can cause traffic disruption when a node that has experienced local signal fails on both the working and the protection paths is recovering from these failures.
An example of sequence diagram highlighting the problem with the priority level of SFc as defined in [RFC6378] is presented in Appendix B.
With the priority swapping between FS and SF-P, the traffic is always moved back to the working path when SF-P occurs in Protecting Administrative state. In case network operators need an option to control their networks so that the traffic can remain on the protection path even when the PSC communication channel is broken, the Freeze command can be used. Freeze is defined to be a "local" command that is not signaled to the remote node. The use of the Freeze command is described in Appendix C.
When the modified priority order specified in this document is in use, the list of local requests in order of priority SHALL be as follows:
This requires modification to the PSC Control Logic (including the state machine) relative to that described in [RFC6378]. Sections 10 and 11 present the PSC Control Logic when all capabilities of APS mode are enabled.
Non-revertive operation of protection switching is defined in [RFC4427]. In this operation, the traffic does not return to the working path when switch-over requests are terminated.
However, the PSC protocol defined in [RFC6378] supports this operation only when recovering from a defect condition: it does not support the non-revertive function when an operator's switch-over command, such as FS or Manual Switch (MS), is cleared. To be aligned with the behavior in other transport networks and to be consistent with [RFC4427], a node should go into the Do-not-Revert (DNR) state not only when a failure condition on the working path is cleared, but also when an operator command that requested switch-over is cleared.
This requires modification to the PSC Control Logic (including the state machine) relative to that described in [RFC6378]. Sections 10 and 11 present the PSC Control Logic when all capabilities of APS mode are enabled.
Changing the non-revertive operation as described in Section 5 introduces necessity of a new operator command to revert traffic to the working path in the DNR state. When the traffic is on the protection path in the DNR state, a Manual Switch to Working (MS-W) command is issued to switch the normal traffic back to the working path. According to Section 4.3.3.6 (Do-not-Revert State) in [RFC6378], "to revert back to the Normal state, the administrator SHALL issue a Lockout of protection (LO) command followed by a Clear command." However, using LO command introduces the potential risk of an unprotected situation while the LO is in effect.
Manual Switch-over for recovery LSP/span command is defined in [RFC4427]. Requirement 83 in [RFC5654] states that the external commands defined in [RFC4427] MUST be supported. Since there is no support for this external command in [RFC6378], this functionality should be added to PSC. This support is provided by introducing the MS-W command. The MS-W command, as described here, corresponds to the "Manual Switch-over for recovery LSP/span" command.
[RFC6378] uses the term "Manual Switch" and its acronym "MS". This document uses the term "Manual Switch to Protection path" and "MS-P" to have the same meaning, while avoiding confusion with "Manual Switch to Working path" and its acronym "MS-W".
Similarly, we modify the name of "Protecting Administrative" state (as defined in [RFC6378]) to be "Switching Administrative" state to include the case where traffic is switched to the working path as a result of the external MS-W command.
MS-P and MS-W SHALL have the same priority. We consider different instances of determining the priority of the commands when they are received either in succession or simultaneously.
Two commands, MS-P and MS-W are transmitted using the same Request field value, but SHALL indicate in the Fault Path (FPath) value the path that the traffic is being diverted from. When traffic is switched to the protection path, the FPath field value SHALL be set to 1, indicating that traffic is being diverted from the working path. When traffic is switched to the working path, the FPath field value SHALL be set to 0, indicating that traffic is being diverted from the protection path. The Data Path (Path) field SHALL indicate where user data traffic is being transported (i.e., if the working path is selected, then Path is set to 0; if the protection path is selected, then Path is set to 1).
When an MS command is in effect at a node, any subsequent MS or EXER command and any other lower priority requests SHALL be ignored.
[RFC6378] defines only one rule for equal priority condition in Section 4.3.2 as "The remote message from the remote LER is assigned a priority just below the similar local input." In order to support the manual switch behavior described in Section 6.3, additional rules for equal priority resolution are required. Since the support of protection against signal degrade also requires a similar equal priority resolution, the rules are described in Section 7.4.
Support of this function requires changes to the PSC Control Logic (including the state machine) relative to that shown in [RFC6378]. Sections 10 and 11 present the PSC Control Logic when all capabilities of APS mode are enabled.
In the MPLS-TP Survivability Framework [RFC6372], both SF and SD fault conditions can be used to trigger protection switching.
[RFC6378], which defines the protection switching protocol for MPLS-TP, does not specify how the SF and SD are detected, and specifies the protection switching protocol associated with SF only.
The PSC protocol associated with SD is covered in this document, but the specifics for the method of identifying SD is out of scope for the protection protocol in the same way that SF detection and MS or FS command initiation are out of scope.
In this document the term Clear Signal Fail or Degrade (SFDc) is used to indicate the clearance of either a degraded condition or a failure condition.
The second paragraph of Section 4.3.3.2 Unavailable state in [RFC6378] shows the intention of including Signal Degrade on Protection path (SD-P) in the Unavailable state. Even though the protection path can be partially available under the condition of SD-P, this document follows the same state grouping as [RFC6378] for SD-P.
The bullet item on the Protecting Failure state in Section 3.6 of [RFC6378] includes the degraded condition in the Protecting Failure state. This document follows the same state grouping as [RFC6378] for Signal Degrade on Working path (SD-W).
To better align the behavior of MPLS-TP networks with that of other transport networks (such as SDH, OTN, and Ethernet transport networks) we define the followings:
The SD message indicates that the transmitting node has identified degradation of the signal or integrity of the packet received on either the working path or the protection path. The FPath field SHALL identify the path that is reporting the degraded condition (i.e., if the protection path, then FPath is set to 0; if the working path, then FPath is set to 1), and the Path field SHALL indicate where the data traffic is being transported (i.e., if the working path is selected, then Path is set to 0; if the protection path is selected, then Path is set to 1).
When the SD condition is cleared and the protected domain is recovering from the situation, the Wait-to-Restore (WTR) timer SHALL be used if the protected domain is configured for revertive behavior. The WTR timer SHALL be started at the node that recovers from a local degraded condition on the working path.
Protection switching against SD is always provided by a selector bridge duplicating user data traffic and feeding it to both the working path and the protection path under SD condition. When a local or remote SD occurs on either the working path or the protection path, the node SHALL duplicate user data traffic and SHALL feed to both the working path and the protection path. The packet duplication SHALL continue as long as any SD condition exists in the protected domain. When the SD condition is cleared, in revertive operation, the packet duplication SHALL continue in the WTR state and SHALL stop when the node leaves the WTR state; while in non-revertive operation, the packet duplication SHALL stop immediately.
The selector bridge with the packet duplication under SD condition, which is a non-permanent bridge, is considered to be a 1:1 protection architecture.
Protection switching against SD does not introduce any modification to the operation of the selector at the sink node described in [RFC6378]. The selector chooses either the working or protection path from which to receive the normal traffic in both 1:1 and 1+1 architectures. The position of the selector, i.e., which path to receive the traffic, is determined by the PSC protocol in bidirectional switching or by the local input in unidirectional switching.
In order to support the MS behavior described in Section 6.3 and the protection against SD described in Section 7.3, it is necessary to expand rules for treating equal priority inputs.
For equal priority local inputs, such as MS and SD, apply a simple first-come, first-served rule. Once a local input is determined as the highest priority local input, then a subsequent equal priority local input requesting a different action, i.e., the action results in the same PSC Request field but different FPath value, will not be presented to the PSC Control Logic as the highest local request. Furthermore, in the case of MS command, the subsequent local MS command requesting a different action will be cancelled.
If a node is in a remote state due to a remote SD (or MS) message, a subsequent local input having the same priority but requesting a different action to the PSC Control Logic, will be considered as having lower priority than the remote message, and will be ignored. For examples, if a node is in remote Switching Administrative state due to a remote MS-P, then any subsequent local MS-W SHALL be ignored and automatically cancelled. If a node is in remote Unavailable state due to a remote SD-P, then any subsequent local SD-W input will be ignored. However, the local SD-W SHALL continue to appear in the Local Request Logic as long as the SD condition exists, but SHALL NOT be the top priority global request, which determines the state transition at the PSC Control Logic.
Cases where two end-points of the protected domain simultaneously receive local triggers of the same priority that request different actions (for example, one node receives SD-P and the other receives SD-W) may occur. Subsequently, each node will receive a remote message with the opposing action indication. To address these cases, we define the following priority resolution rules:
In the preceding paragraphs, the "simultaneously" refers to the case a sent SD (or MS) request has not been confirmed by the remote end in bidirectional protection switching. When a local node that has transmitted a SD message receives a SD (or MS) message that indicates a different value of Path field from the value of Path field in the transmitted SD (or MS) message, both the local and remote SD requests are considered to occur simultaneously.
The addition of support for protection against SD requires modification to the PSC Control Logic (including the state machine) relative to that described in [RFC6378]. Sections 10 and 11 present the PSC Control Logic when all capabilities of APS mode are enabled.
The EXER command is used to verify the correct operation of the PSC communication, such as the aliveness of the Local Request Logic, the integrity of the PSC Control Logic, the PSC message generation and reception mechanism, and the integrity of the protection path. EXER does not trigger any actual traffic switching.
The command is only relevant for bidirectional protection switching, since it is dependent upon receiving a response from the remote node. The EXER command is assigned lower priority than any switching message. It may be used regardless of the traffic usage of the working path.
When a node receives a remote EXER message, it SHOULD respond with a Reverse Request (RR) message with the FPath and Path fields set according to the current condition of the node. The RR message SHALL be generated only in response to a remote EXER message.
This command is documented in R84 of [RFC5654].
If EXER commands are input at both ends, then a race condition may arise. This is resolved as follows:
The following PSC Requests are added to the PSC Request field to support the Exercise command (see also Section 14.1):
The relative priorities of EXER and RR are defined in Section 10.2.
A Capability is an individual behavior whose use is signaled in a Capabilities TLV, which is placed in Optional TLVs field inside the PSC message shown in Figure 2 of [RFC6378]. The format of the Capabilities TLV is:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type = Capabilities | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Value = Flags |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 1: Format of Capabilities TLV
The value of the Type field is TBD pending IANA allocation.
The value of the Length field is the length of the Flags field in octets. The length of the Flags field MUST be a multiple of 4 octets and MUST be the minimum required to signal all the required capabilities.
Section 4 to Section 8 discuss five capabilities that are signaled using the five most significant bits; if a node wishes to signal these five capabilities, it MUST send a Flags field of 4 octets. A node would send a Flags field greater than 4 octets only if it had more than 32 Capabilities to indicate. All unused bits MUST be set to zero.
If the bit assigned for an individual capability is set to 1, it indicates the sending node's intent to use that capability in the protected domain. If a bit is set to 0, the sending node does not intend to use the indicated capability in the protected domain. Note that it is not possible to distinguish between the intent not to use a capability and a node's complete non-support (i.e., lack of implementation) of a given capability.
This document defines five specific capabilities that are described in Section 4 to Section 8. Each capability is assigned bit as follows:
If all the five capabilities should be used, a node SHALL set the Flags field to 0xF8000000.
A node MUST include its Capabilities TLV in every PSC message that it transmits. The transmission and acceptance of the PSC message is described in Section 4.1 of [RFC6378].
When a node receives a Capabilities TLV it MUST compare the Flags value to its most recent Flags value transmitted by the node. If the two are equal, the protected domain is said to be running in the mode indicated by that set of capabilities (see Section 9.2). If the sent and received Capabilities TLVs are not equal, this indicates a capabilities TLV mismatch. When this happens, the node MUST alert the operator and MUST NOT perform any protection switching until the operator resolves the mismatch between the two end-points.
A mode is a given set of Capabilities. Modes are shorthand; referring to a set of capabilities by their individual values or by the name of their mode does not change the protocol behavior. This document defines two modes - PSC and APS. Capability TLVs with other combinations than the one specified by a mode are not supported in this specification.
PSC mode is defined as the lack of support for any of the additional capabilities defined in this document - that is, a Capabilities set of 0x0. It is the behavior specified in [RFC6378].
There are two ways to declare PSC mode. A node can send no Capabilities TLV at all since there are no TLV units defined in [RFC6378], or it can send a Capabilities TLV with Flags value set to 0x0. In order to allow backward compatibility between two end-points - one which supports sending the Capabilities TLV, and one which does not, the node that has the ability to send and process the PSC mode Capabilities TLV MUST be able to both send the PSC mode Capabilities TLV and send no Capabilities TLV at all. An implementation MUST be configurable between these two options.
APS mode is defined as the use of all the five specific capabilities, which are described in Section 4 to Section 8 in this document. APS mode is indicated with the Flags value of 0xF8000000.
This section and the following section define the behavior of PSC protocol when all of the aforementioned capabilities are enabled, i.e., APS mode.
This document defines two new values for the "Request" field in the PSC protocol message that is shown in Figure 2 of [RFC6378] as follows:
See also Section 14.1 of this document.
Based on the description in Sections 3 and 4.3.2 in [RFC6378], the priorities of multiple outstanding local inputs are evaluated in the Local Request Logic, where the highest priority local input (highest local request) is determined. This highest local request is passed to the PSC Control Logic, that will determine the higher priority input (top priority global request) between the highest local request and the last received remote message. When a remote message comes to the PSC Control Logic, the top priority global request is determined between this remote message and the highest local request which is present. The top priority global request is used to determine the state transition, which is described in Section 11. In this document, in order to simplify the description on the PSC Control Logic, we strictly decouple the priority evaluation from the state transition table lookup.
The priorities for both local and remote requests are defined as follows from highest to lowest:
Note that the "Local only" requests are not tranmitted to the remote node. Likewise, the "Remote only" requests do not exist in the Local Request Logic as local inputs. For example, the priority of WTR only applies to the received WTR message, which is generated from the remote node. The remote node that is running the WTR timer in the WTR state has no local request.
The remote SF and SD on either the working path or the protection path and the remote MS to either the working path or the protection path are indicated by the values of the Request and FPath fields in the PSC message.
The remote request from the remote node is assigned a priority just below the same local request except NR and equal priority requests, such as SD and MS. Since a received NR message needs to be used in the state transition table lookup when there is no outstanding local request, the remote NR request SHALL have a higher priority than the local NR. For the equal priority requests, see Section 10.2.1.
As stated in Section 10.2, the remote request from the remote node is assigned a priority just below the same local request. However, for equal priority requests, such as SD and MS, the priority SHALL be evaluated as described in this section.
For equal priority local requests, first-come, first-served rule SHALL be applied. Once a local request appears in the Local Request Logic, a subsequent equal priority local request requesting a different action, i.e., the action results in the same Request value but a different FPath value, SHALL be considered to have a lower priority. Furthermore, in the case of MS command, the subsequent local MS command requesting a different action SHALL be rejected and cleared.
When the priority is evaluated in the PSC Control Logic between the highest local request and a remote request, the following equal priority resolution rules SHALL be applied:
A local input indicating a defect, such as SF-P, SF-W, SD-P and SD-W, SHALL be accepted and retained persistently in the Local Request Logic as long as the defect condition exists. If there is any higher priority local input than the local defect input, the higher priority local input is passed to the PSC Control Logic as the highest local request, but the local defect input cannot be removed but remains in the Local Request Logic. When the higher priority local input is cleared, the local defect will become the highest local request if the defect condition still exists.
Operator Clear (OC) command, SFDc and WTR Timer Expiry are not persistent. Once they appear to the Local Request Logic and complete all the operations in the protection switching control, they SHALL disappear.
LO, FS, MS, and EXER commands SHALL be rejected if there is any higher priority local input in the Local Request Logic. If a new higher-priority local request (including an operator command) is accepted, any previous lower-priority local operator command SHALL be cancelled. When any higher-priority remote request is received, a lower-priority local operator command SHALL be cancelled. The cancelled operator command is cleared. If the operators wish to renew the cancelled command then they should reissue the command.
When there is a change in the highest local request or in remote PSC messages, the top priority global request SHALL be evaluated and the state transition tables SHALL be looked up in the PSC Control Logic. The following rules are applied to the operation related to the state transition table lookup.
The extended states, as they appear in the table, are as follows:
N Normal state
UA:LO:L Unavailable state due to local LO command
UA:P:L Unavailable state due to local SF-P
UA:DP:L Unavailable state due to local SD-P
UA:LO:R Unavailable state due to remote LO message
UA:P:R Unavailable state due to remote SF-P message
UA:DP:R Unavailable state due to remote SD-P message
PF:W:L Protecting Failure state due to local SF-W
PF:DW:L Protecting Failure state due to local SD-W
PF:W:R Protecting Failure state due to remote SF-W message
PF:DW:R Protecting Failure state due to remote SD-W message
SA:F:L Switching Administrative state due to local FS command
SA:MW:L Switching Administrative state due to local MS-W command
SA:MP:L Switching Administrative state due to local MS-P command
SA:F:R Switching Administrative state due to remote FS message
SA:MW:R Switching Administrative state due to remote MS-W message
SA:MP:R Switching Administrative state due to remote MS-P message
WTR Wait-to-Restore state
DNR Do-not-Revert state
E::L Exercise state due to local EXER command
E::R Exercise state due to remote EXER message
Each state corresponds to the transmission of a particular set of Request, FPath and Path fields. The table below lists the message that is generally sent in each particular state. If the message to be sent in a particular state deviates from the table below, it is noted in the footnotes to the state transition tables.
State Request(FPath,Path)
------- ------------------------------------
N NR(0,0)
UA:LO:L LO(0,0)
UA:P:L SF(0,0)
UA:DP:L SD(0,0)
UA:LO:R highest local request(local FPath,0)
UA:P:R highest local request(local FPath,0)
UA:DP:R highest local request(local FPath,0)
PF:W:L SF(1,1)
PF:DW:L SD(1,1)
PF:W:R highest local request(local FPath,1)
PF:DW:R highest local request(local FPath,1)
SA:F:L FS(1,1)
SA:MW:L MS(0,0)
SA:MP:L MS(1,1)
SA:F:R highest local request(local FPath,1)
SA:MW:R NR(0,0)
SA:MP:R NR(0,1)
WTR WTR(0,1)
DNR DNR(0,1)
E::L EXER(0,x), where x is the existing Path value
when Exercise command is issued.
E::R RR(0,x), where x is the existing Path value
when RR message is generated.
Some operation examples of APS mode are shown in Appendix D.
In the state transition tables below, the letter 'i' stands for "ignore", and is an indication to remain in the current state and continue transmitting the current PSC message
| OC | LO | SFDc | SF-P | FS | SF-W |
--------+-----+---------+------+--------+--------+--------+
N | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
UA:LO:L | (1) | i | i | i | i | i |
UA:P:L | i | UA:LO:L | (1) | i | i | i |
UA:DP:L | i | UA:LO:L | (1) | UA:P:L | SA:F:L | PF:W:L |
UA:LO:R | i | UA:LO:L | i | UA:P:L | i | PF:W:L |
UA:P:R | i | UA:LO:L | i | UA:P:L | i | PF:W:L |
UA:DP:R | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
PF:W:L | i | UA:LO:L | (2) | UA:P:L | SA:F:L | i |
PF:DW:L | i | UA:LO:L | (2) | UA:P:L | SA:F:L | PF:W:L |
PF:W:R | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
PF:DW:R | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
SA:F:L | (3) | UA:LO:L | i | UA:P:L | i | i |
SA:MW:L | (1) | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
SA:MP:L | (3) | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
SA:F:R | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
SA:MW:R | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
SA:MP:R | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
WTR | (4) | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
DNR | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
E::L | (5) | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
E::R | i | UA:LO:L | i | UA:P:L | SA:F:L | PF:W:L |
| SD-P | SD-W | MS-W | MS-P | WTRExp | EXER
--------+---------+---------+---------+---------+--------+------
N | UA:DP:L | PF:DW:L | SA:MW:L | SA:MP:L | i | E::L
UA:LO:L | i | i | i | i | i | i
UA:P:L | i | i | i | i | i | i
UA:DP:L | i | i | i | i | i | i
UA:LO:R | UA:DP:L | PF:DW:L | i | i | i | i
UA:P:R | UA:DP:L | PF:DW:L | i | i | i | i
UA:DP:R | UA:DP:L | PF:DW:L | i | i | i | i
PF:W:L | i | i | i | i | i | i
PF:DW:L | i | i | i | i | i | i
PF:W:R | UA:DP:L | PF:DW:L | i | i | i | i
PF:DW:R | UA:DP:L | PF:DW:L | i | i | i | i
SA:F:L | i | i | i | i | i | i
SA:MW:L | UA:DP:L | PF:DW:L | i | i | i | i
SA:MP:L | UA:DP:L | PF:DW:L | i | i | i | i
SA:F:R | UA:DP:L | PF:DW:L | i | i | i | i
SA:MW:R | UA:DP:L | PF:DW:L | SA:MW:L | i | i | i
SA:MP:R | UA:DP:L | PF:DW:L | i | SA:MP:L | i | i
WTR | UA:DP:L | PF:DW:L | SA:MW:L | SA:MP:L | (6) | i
DNR | UA:DP:L | PF:DW:L | SA:MW:L | SA:MP:L | i | E::L
E::L | UA:DP:L | PF:DW:L | SA:MW:L | SA:MP:L | i | i
E::R | UA:DP:L | PF:DW:L | SA:MW:L | SA:MP:L | i | E::L
NOTES:
| LO | SF-P | FS | SF-W | SD-P | SD-W |
--------+---------+--------+--------+--------+---------+---------+
N | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
UA:LO:L | i | i | i | i | i | i |
UA:P:L | UA:LO:R | i | i | i | i | i |
UA:DP:L | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | i | (7) |
UA:LO:R | i | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
UA:P:R | UA:LO:R | i | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
UA:DP:R | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | i | PF:DW:R |
PF:W:L | UA:LO:R | UA:P:R | SA:F:R | i | i | i |
PF:DW:L | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | (8) | i |
PF:W:R | UA:LO:R | UA:P:R | SA:F:R | i | UA:DP:R | PF:DW:R |
PF:DW:R | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | i |
SA:F:L | UA:LO:R | UA:P:R | i | i | i | i |
SA:MW:L | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
SA:MP:L | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
SA:F:R | UA:LO:R | UA:P:R | i | PF:W:R | UA:DP:R | PF:DW:R |
SA:MW:R | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
SA:MP:R | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
WTR | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
DNR | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
E::L | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
E::R | UA:LO:R | UA:P:R | SA:F:R | PF:W:R | UA:DP:R | PF:DW:R |
| MS-W | MS-P | WTR | EXER | RR | DNR | NR
--------+---------+---------+-----+------+----+------+----
N | SA:MW:R | SA:MP:R | i | E::R | i | i | i
UA:LO:L | i | i | i | i | i | i | i
UA:P:L | i | i | i | i | i | i | i
UA:DP:L | i | i | i | i | i | i | i
UA:LO:R | SA:MW:R | SA:MP:R | i | E::R | i | i | N
UA:P:R | SA:MW:R | SA:MP:R | i | E::R | i | i | N
UA:DP:R | SA:MW:R | SA:MP:R | i | E::R | i | i | N
PF:W:L | i | i | i | i | i | i | i
PF:DW:L | i | i | i | i | i | i | i
PF:W:R | SA:MW:R | SA:MP:R | (9) | E::R | i | (10) | (11)
PF:DW:R | SA:MW:R | SA:MP:R | (9) | E::R | i | (10) | (11)
SA:F:L | i | i | i | i | i | i | i
SA:MW:L | i | i | i | i | i | i | i
SA:MP:L | i | i | i | i | i | i | i
SA:F:R | SA:MW:R | SA:MP:R | i | E::R | i | DNR | N
SA:MW:R | i | SA:MP:R | i | E::R | i | i | N
SA:MP:R | SA:MW:R | i | i | E::R | i | DNR | N
WTR | SA:MW:R | SA:MP:R | i | i | i | i | (12)
DNR | SA:MW:R | SA:MP:R | (13)| E::R | i | i | i
E::L | SA:MW:R | SA:MP:R | i | i | i | i | i
E::R | SA:MW:R | SA:MP:R | i | i | i | DNR | N
NOTES:
The state transition tables given in Sections 11.1 and 11.2 are for bidirectional protection switching, where remote PSC protocol messages are used to determine the protection switching actions. 1+1 unidirectional protection switching does not require the remote information in PSC protocol message and acts upon local inputs only. The state transition by local inputs in Section 11.1 SHALL be reused for 1+1 unidirectional protection under the following conditions:
The remote PSC message that is received from the remote node is subject to the detection of provisioning mismatch and protocol failure conditions. In APS mode, provisioning mismatches are handled as follows:
The followings are the protocol failure situations and the actions to be taken:
This document introduces no new security risks. [RFC6378] points out that MPLS relies on assumptions about traffic injection difficulty and assumes that the control plane does not have end-to-end security. [RFC5920] describes MPLS security issues and generic methods for securing traffic privacy and integrity. MPLS use should conform such advice.
In the "Generic Associated Channel (G-ACh) Parameters" registry, IANA maintains the "MPLS PSC Request Registry".
IANA is requested to assign two new code points from this registry. The values shall be allocated as follows:
Value Description Reference
----- --------------------- ---------------
2 Reverse Request (this document)
3 Exercise (this document)
In the "Generic Associated Channel (G-ACh) Parameters" registry, IANA maintains the "MPLS PSC TLV Registry".
This document defines a new value for the Capabilities TLV type in the "MPLS PSC TLV Registry".
Value Description Reference
------ --------------------- ---------------
TBD Capabilities (this document)
IANA is requested to create and maintain a new registry within the "Generic Associated Channel (G-ACh) Parameters" registry called "MPLS PSC Capability Flag Registry". All flags within this registry SHALL be allocated according to the "Standards Action" procedures as specified in RFC 5226 [RFC5226].
The length of the flags MUST be a multiple of 4 octets. This document defines 4 octet flags. Flags greater than 4 octets SHALL be used only if more than 32 Capabilities need to be defined. Flags defined in this document are:
Bit Hex Value Capability Reference
---- ---------- ----------------------------------- ---------------
0 0x80000000 priority modification (this document)
1 0x40000000 non-revertive behavior modification (this document)
2 0x20000000 support of MS-W command (this document)
3 0x10000000 support of protection against SD (this document)
4 0x08000000 support of EXER command (this document)
5-31 Unassigned (this document)
The authors would like to thank Yaacov Weingarten, Yuji Tochio, Malcolm Betts, Ross Callon, Qin Wu and Xian Zhang for their valuable comments and suggestions on this document.
We would also like to acknowledge explicit text provided by Loa Andersson and Adrian Farrel.
| [RFC2119] | Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. |
| [RFC5226] | Narten, T. and H. Alvestrand, "Guidelines for Writing an IANA Considerations Section in RFCs", BCP 26, RFC 5226, May 2008. |
| [RFC5654] | Niven-Jenkins, B., Brungard, D., Betts, M., Sprecher, N. and S. Ueno, "Requirements of an MPLS Transport Profile", RFC 5654, September 2009. |
| [RFC6378] | Weingarten, Y., Bryant, S., Osborne, E., Sprecher, N. and A. Fulignoli, "MPLS Transport Profile (MPLS-TP) Linear Protection", RFC 6378, October 2011. |
| [RFC4427] | Mannie, E. and D. Papadimitriou, "Recovery (Protection and Restoration) Terminology for Generalized Multi-Protocol Label Switching (GMPLS)", RFC 4427, March 2006. |
| [RFC5920] | Fang, L., "Security Framework for MPLS and GMPLS Networks", RFC 5920, July 2010. |
| [RFC6372] | Sprecher, N. and A. Farrel, "MPLS Transport Profile (MPLS-TP) Survivability Framework", RFC 6372, September 2011. |
| [G841] | International Telecommunications Union, "Types and characteristics of SDH network protection architectures ", ITU-T Recommendation G.841, October 1998. |
| [G873.1] | International Telecommunications Union, "Optical Transport Network (OTN): Linear protection ", ITU-T Recommendation G.873.1, July 2011. |
| [G8031] | International Telecommunications Union, "Ethernet Linear Protection Switching ", ITU-T Recommendation G.8031/Y.1342, June 2011. |
The sequence diagram shown is an example of the out-of-service scenarios based on the priority level defined in [RFC6378]. The first PSC message which differs from the previous PSC message is shown.
A Z
| |
(1) |-- NR(0,0) ------>| (1)
|<----- NR(0,0) ---|
| |
| |
| (FS issued at Z) | (2)
(3) |<------ FS(1,1) --|
|-- NR(0,1) ------>|
| |
| |
(4) | (SF on P(A<-Z)) |
| |
| |
| (Clear FS at Z) | (5)
(6) | X <- NR(0,0) --|
| |
| |
(1) Each end is in the Normal state, and transmits NR(0,0) messages.
(2) When a FS command is issued at node Z, node Z goes into local Protecting Administrative state (PA:F:L) and begins transmission of an FS(1,1) messages.
(3) A remote FS message causes node A to go into remote Protecting Administrative state (PA:F:R), and node A begins transmitting NR(0,1) messages.
(4) When node A detects a unidirectional SF-P, node A keeps sending NR(0,1) message because SF-P is ignored under the PA:F:R state.
(5) When a Clear command is issued at node Z, node Z goes into the Normal state and begins transmission of NR(0,0) messages.
(6) But, node A cannot receive PSC message because of local unidirectional SF-P. Because no valid PSC message is received, over a period of several successive message intervals, the last valid received message remains applicable and the node A continue to transmit an NR(0,1) message in the PA:F:R state.
Now, there exists a mismatch between the bridge/selector positions of node A (transmitting an NR(0,1)) and node Z (transmitting an NR(0,0)). It results in out-of-service even when there is neither SF-W nor FS.
An example of sequence diagram showing the problem with the priority level of SFc defined in [RFC6378] is given below. The following sequence diagram is depicted for the case of bidirectional signal fails. However, other cases with unidirectional signal fails can result in the same problem. The first PSC message which differs from the previous PSC message is shown.
A Z
| |
(1) |-- NR(0,0) ------>| (1)
|<----- NR(0,0) ---|
| |
| |
(2) | (SF on P(A<->Z)) | (2)
|-- SF(0,0) ------>|
|<------ SF(0,0) --|
| |
| |
(3) | (SF on W(A<->Z)) | (3)
| |
| |
(4) | (Clear SF-P) | (4)
| |
| |
(5) | (Clear SF-W) | (5)
| |
| |
(1) Each end is in the Normal state, and transmits NR(0,0) messages.
(2) When SF-P occurs, each node enters into the UA:P:L state and transmits SF(0,0) messages. Traffic remains on the working path.
(3) When SF-W occurs, each node remains in the UA:P:L state as SF-W has a lower priority than SF-P. Traffic is still on the working path. Traffic cannot be delivered as both the working path and the protection path are experiencing signal fails.
(4) When SF-P is cleared, local "Clear SF-P" request cannot be presented to the PSC Control Logic, which takes the highest local request and runs PSC state machine, since the priority of "Clear SF-P" is lower than that of SF-W. Consequently, there is no change in state, and the selector and/or bridge keep pointing at the working path, which has signal fail condition.
Now, traffic cannot be delivered while the protection path is recovered and available. It should be noted that the same problem will occur in the case that the sequence of SF-P and SF-W events is changed.
If we further continue with this sequence to see what will happen after SF-W is cleared,
(5) When SF-W is cleared, local "Clear SF-W" request can be passed to the PSC Control Logic as there is no higher priority local input, but this will be ignored in the PSC Control Logic according to the state transition definition in [RFC6378]. There will be no change in state or protocol message transmitted.
As SF-W is now cleared and the selector and/or bridge are still pointing at the working path, traffic delivery is resumed. However, each node is in the UA:P:L state and transmitting SF(0,0) message, while there exists no outstanding request for protection switching. Moreover, any future legitimate protection switching requests, such as SF-W, will be rejected as each node thinks the protection path is unavailable.
The "Freeze" command applies only to the local node of the protection group and is not signaled to the remote node. This command freezes the state of the protection group. Until the Freeze is cleared, additional local commands are rejected and condition changes and received PSC information are ignored.
"Clear Freeze" command clears the local freeze. When the Freeze command is cleared, the state of the protection group is recomputed based on the persistent condition of the local triggers.
Because the freeze is local, if the freeze is issued at one end only, a failure of protocol can occur as the other end is open to accept any operator command or a fault condition.
The sequence diagrams shown in this section are only a few examples of the APS mode operations. The first PSC protocol message which differs from the previous message is shown. The operation of hold-off timer is omitted. The Request, FPath and Path fields, whose values are changed during PSC message exchange are shown. For an example, SF(1,0) represents an PSC message with the following field values: Request=SF, FPath=1 and Path=1. The values of the other fields remain unchanged from the initial configuration. W(A->Z) and P(A->Z) indicate the working path and the protection path in the direction of A to Z, respectively.
Example 1. 1:1 bidirectional protection switching (revertive operation) - Unidirectional SF case
A Z
| |
(1) |<---- NR(0,0)---->| (1)
| |
| |
(2) | (SF on W(Z->A)) |
|---- SF(1,1)----->| (3)
(4) |<----- NR(0,1)----|
| |
| |
(5) | (Clear SF-W) |
|---- WTR(0,1)---->|
/| |
| | |
WTR timer | |
| | |
\| |
(6) |---- NR(0,1)----->| (7)
(8) |<----- NR(0,0)----|
|---- NR(0,0)----->| (9)
| |
(1) The protected domain is operating without any defect, and the working path is used for delivering the traffic in the Normal state.
(2) SF-W occurs in the Z to A direction. Node A enters into the PF:W:L state and generates SF(1,1) message. Selector and bridge of node A are pointing at the protection path.
(3) Upon receiving SF(1,1), node Z sets selector and bridge to the protection path. As there is no local request in node Z, node Z generates NR(0,1) message in the PF:W:R state.
(4) Node A confirms that the remote node is also selecting the protection path.
(5) Node A detects clearing of SF condition, starts the WTR timer, and sends WTR(0,1) message in the WTR state.
(6) At expiration of the WTR timer, node A sets selector and bridge to the working path and sends NR(0,1) message.
(7) Node Z is notified that the remote request has been cleared. Node Z transits to the Normal state and sends NR(0,0) message.
(8) Upon receiving NR(0,0) message, node A transits to the Normal state and sends NR(0,0) message.
(9) It is confirmed that the remote node is also selecting the working path.
Example 2. 1:1 bidirectional protection switching (revertive operation) - Bidirectional SF case - Inconsistent WTR timers
A Z
| |
(1) |<---- NR(0,0)---->| (1)
| |
| |
(2) | (SF on W(A<->Z)) | (2)
|<---- SF(1,1)---->|
| |
| |
(3) | (Clear SF-W) | (3)
|<---- NR(0,1)---->|
(4) |<--- WTR(0,1) --->| (4)
/| |\
| | | |
WTR timer | | WTR timer
| | | |
| | |/
| |<------ NR(0,1)---| (5)
| | |
\| |
(6) |--- NR(0,1)------>|
|<------ NR(0,0)---| (7)
(8) |--- NR(0,0)------>|
| |
(1) Each end is in the Normal state, and transmits NR(0,0) messages.
(2) When SF-W occurs, each node enters into the PF:W:L state and transmits SF(1,1) messages. Traffic is switched to the protection path. Upon receiving SF(1,1), each node confirms that the remote node is also sending and receiving the traffic from the protection path.
(3) When SF-W is cleared, each node transits to the PF:W:R state and transmits NR(0,1) messages as the last received message is SF-W.
(4) Upon receiving NR(0,1) messages, each node goes into the WTR state, starts the WTR timer, and sends the WTR(0,1) messages.
(5) At expiration of the WTR timer in node Z, node Z sends NR(0,1) as the last received APS message was WTR. When NR(0,1) arrives at node A, node A maintains the WTR state and keeps sending current WTR messages as described in the state transition table.
(6) At expiration of the WTR timer in node A, node A sends NR(0,1).
(7) When the NR(0,1) message arrives at node Z, node Z moves to the Normal state, sets selector and bridge to the working path, and sends NR(0,0) message.
(8) The received NR(0,0) message causes node A to go to the Normal state. Now, the traffic is switched back to the working path.
Example 3. 1:1 bidirectional protection switching - R bit mismatch
This example shows that both sides will interwork and the traffic is protected when one side (node A) is configured as revertive operation and the other (node Z) is configured as non-revertive operation. The interworking is covered in the state transition tables.
(revertive) A Z (non-revertive)
| |
(1) |<---- NR(0,0)---->| (1)
| |
| |
(2) | (SF on W(A<->Z)) | (2)
|<---- SF(1,1)---->|
| |
| |
(3) | (Clear SF-W) | (3)
|<---- NR(0,1)---->|
(4) |<----- DNR(0,1)---| (4)
/|-- WTR(0,1)------>|
| |<----- NR(0,1)----| (5)
| | |
WTR timer | |
| | |
| | |
\| |
(6) |--- NR(0,1)------>|
|<------ NR(0,0)---| (7)
(8) |--- NR(0,0)------>|
| |
(1) Each end is in the Normal state, and transmits NR(0,0) messages.
(2) When SF-W occurs, each node enters into the PF:W:L state and transmits SF(l,l) messages. Traffic is switched to the protection path. Upon receiving SF(1,1), each node confirms that the remote node is also sending and receiving the traffic on the protection path.
(3) When SF-W is cleared, each node transits to the PF:W:R state and transmits NR(0,1) messages as the last received message is SF-W.
(4) Upon receiving NR(0,1) messages, node A goes into the WTR state, starts the WTR timer, and sends WTR(0,1) messages. At the same time, node Z transits to the DNR state and sends DNR(0,1) message.
(5) When the WTR message arrives at node Z, node Z transits to the WTR state and send NR(0,1) message according to the state transition table. At the same time, the DNR message arrived at node Z is ignored according to the state transition table. Therefore, node Z, which is configured as non-revertive operation, is operating as if in revertive operation.
(6) At expiration of the WTR timer in node A, node A sends NR(0,1).
(7) When the NR(0,1) message arrives at node Z, node Z moves to the Normal state, sets selector and bridge to the working path, and sends NR(0,0) message.
(8) The received NR(0,0) message causes node A to transits to the Normal state. Now, the traffic is switched back to the working path.