Network Working Group Chandra Ramachandran (Ed) Internet Draft Yakov Rekhter Intended status: Standards Track Juniper Networks Ina Minei Google, Inc Ebben Aries Facebook Dante Pacella Verizon Expires: September 06, 2015 March 06, 2015 Refresh Interval Independent FRR Facility Protection draft-chandra-mpls-enhanced-frr-bypass-01 Status of this Memo 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), 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 September 06, 2015. Copyright Notice Copyright (c) 2015 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 Chandra, et al Expires September 06, 2015 [Page 1] Internet-Draft Enhanced FRR bypass March 2015 (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. Abstract This document defines RSVP-TE extensions to facilitate refresh- interval independent FRR facility protection. Conventions used in this document 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]. Table of Contents 1. Introduction...................................................3 2. Motivation.....................................................3 3. Problem Description............................................4 4. Solution Aspects...............................................6 4.1. Signaling Protection availability in Path RRO Flags.......7 4.1.1. PLR Behavior.........................................7 4.1.2. Remote Signaling Adjacency...........................8 4.1.3. PATH RRO flags Propagation...........................9 4.1.4. MP Behavior..........................................9 4.1.5. "Remote" state on MP.................................9 4.2. Impact of Failures on LSP State..........................10 4.2.1. Non-MP Behavior on Phop Link/Node Failure...........10 4.2.2. LP-MP Behavior on Phop Link Failure.................11 4.2.3. LP-MP Behavior on Phop Node Failure.................11 4.2.4. NP-MP Behavior on Phop Link/Node Failure............11 4.2.5. NP-MP Behavior on PLR Link Failure..................11 4.2.6. Phop Link Failure on a Node that is LP-MP and NP-MP.12 4.2.7. Phop Node Failure on Node that is LP-MP and NP-MP...13 4.3. Conditional Path Tear....................................13 4.3.1. Sending Conditional Path Tear.......................13 4.3.2. Processing Conditional Path Tear....................13 4.3.3. CONDITIONS object...................................14 4.4. Remote State Teardown....................................15 4.4.1. PLR Behavior on Local Repair Failure................15 Chandra, et al Expires September 06, 2015 [Page 2] Internet-Draft Enhanced FRR bypass March 2015 4.4.2. PLR Behavior on Resv RRO Change.....................15 4.4.3. LSP Preemption during Local Repair..................16 4.4.3.1. Preemption on LP-MP after Phop Link failure....16 4.4.3.2. Preemption on NP-MP after Phop Link failure....16 4.5. Backward Compatibility Procedures........................17 4.5.1. Detecting Support for Enhanced FRR Facility Protection ...........................................................18 4.5.2. Procedures for backward compatibility...............19 4.5.2.1. Lack of support on Downstream Node.............19 4.5.2.2. Lack of support on Upstream Node...............19 4.5.2.3. Incremental Deployment.........................20 5. Security Considerations.......................................21 6. IANA Considerations...........................................21 6.1. New Object - CONDITIONS..................................21 6.2. New CAPABILITY Object value..............................21 7. Normative References..........................................21 8. Acknowledgments...............................................22 9. Authors' Addresses............................................22 1. Introduction The facility backup protection mechanism is one of two methods discussed in [RFC4090] for enabling the fast reroute of traffic onto backup LSP tunnels in 10s of milliseconds, in the event of a failure. This document discusses a few shortcomings with some of the refresh-interval reliant procedures proposed for this method in [RFC4090]. These shortcomings come to the fore under scaled conditions and get highlighted even further when large RSVP-TE refresh intervals are used. The RSVP-TE extensions defined in this document will enhance the facility backup protection mechanism by making the corresponding procedures refresh-interval independent. 2. Motivation Standard RSVP [RFC2205] maintains state via the generation of RSVP Path/Resv refresh messages. Refresh messages are used to both synchronize state between RSVP neighbors and to recover from lost RSVP messages. The use of Refresh messages to cover many possible failures has resulted in a number of operational problems. One problem relates to RSVP control plane scaling due to periodic refreshes of Path and Resv messages, another relates to the reliability and latency of RSVP signaling. An additional problem is the time to clean up the stale state after a tear message is lost. For more on these problems see Section 1 of [RFC2961]. All these problems adversely affect RSVP control plane scalability. RSVP-TE inherited all these problems from standard RSVP. Chandra, et al Expires September 06, 2015 [Page 3] Internet-Draft Enhanced FRR bypass March 2015 Procedures specified in [RFC2961] address the above mentioned problems by eliminating dependency on refreshes for state synchronization and for recovering from lost RSVP messages, and by eliminating dependency on refresh timeout for stale state cleanup. Implementing these procedures allows to improve RSVP-TE control plane scalability. However, the procedures specified in [RFC2961] do not fully address stale state cleanup for facility backup protection [RFC4090], as facility backup protection still depends on refresh timeouts for stale state cleanup. Thus [RFC2961] is insufficient to address the problem of stale state cleanup when facility backup protection is used. The procedures specified in this document, in combination with [RFC2961], eliminate facility backup protection dependency on refresh timeouts for stale state cleanup. These procedures, in combination with [RFC2961], fully address the above mentioned problem of RSVP-TE stale state cleanup, including the cleanup for facility backup protection. The procedures specified in this document assume reliable delivery of RSVP messages, as specified in [RFC2961]. Therefore this document makes support for [RFC2961] a pre-requisite. 3. Problem Description [E] / \ / \ / \ / \ / \ / \ [A]-----[B]-----[C]-----[D] \ / \ / \ / \ / \ / \ / [F] Figure 1: Example Topology Chandra, et al Expires September 06, 2015 [Page 4] Internet-Draft Enhanced FRR bypass March 2015 In the topology illustrated in Figure 1, consider a large number of LSPs from A to D transiting B and C. Assume that refresh interval has been configured to be large of the order of minutes and refresh reduction extensions are enabled on all routers. Also assume that node protection has been configured for the LSPs and the LSPs are protected by each router in the following way - A has made node protection available using bypass LSP A -> E -> C; A is the Point of Local Repair (PLR) and C is Node Protecting Merge Point (NP-MP) - B has made node protection available using bypass LSP B -> F -> D; B is the PLR and D is the NP-MP - C has made link protection available using bypass LSP C -> B -> F -> D; C is the PLR and D is the Link Protecting Merge Point (LP- MP) In the above condition, assume that B-C link fails. The following is the sequence of events that is expected to occur for all protected LSPs under normal conditions. 1.B performs local repair and re-directs LSP traffic over the bypass LSP B -> F -> D. 2.B also creates backup state for the LSP and triggers sending of backup LSP state to D over the bypass LSP B -> F -> D. 3.D receives backup LSP states and merges the backups with the protected LSPs. 4.As the link on C over which the LSP states are refreshed has failed, C will no longer receive state refreshes. Consequently the protected LSP states on C will time out and C will send tear down message for all LSPs. While the above sequence of events has been described in [RFC4090], there are a few problems for which no mechanism has been specified explicitly. - If the protected LSP on C times out before D receives signaling for the backup LSP, then D would receive PathTear from C prior to receiving signaling for the backup LSP, thus resulting in deleting the LSP state. This would be possible at scale even with default refresh time. Chandra, et al Expires September 06, 2015 [Page 5] Internet-Draft Enhanced FRR bypass March 2015 - If upon the link failure C is to keep state until its timeout, then with long refresh interval this may result in a large amount of stale state on C. Alternatively, if upon the link failure C is to delete the state and send PathTear to D, this would result in deleting the state on D, thus deleting the LSP. D needs a reliable mechanism to determine whether it is MP or not to overcome this problem. - If head-end A attempts to tear down LSP after step 1 but before step 2 of the above sequence, then B may receive the tear down message before step 2 and delete the LSP state from its state database. If B deletes its state without informing D, with long refresh interval this could cause (large) buildup of stale state on D. - If B fails to perform local repair in step 1, then B will delete the LSP state from its state database without informing D. As B deletes its state without informing D, with long refresh interval this could cause (large) buildup of stale state on D. The purpose of this document is to provide solutions to the above problems which will then make it practical to scale up to a large number of protected LSPs in the network. 4. Solution Aspects The solution consists of five parts. - Enhance the facility protection method defined in [RFC4090] by introducing an MP determination mechanism that enables PLR to signal availability of link or node protection to the MP. See section 4.1 for more details. - Handle upstream link or node failures by cleaning up LSP states if the node has not found itself as MP through the MP determination mechanism. See section 4.2 for more details. The combination of "path state" maintained as Path State Block (PSB) and "reservation state" maintained as Reservation State Block (RSB) forms an individual LSP state on an RSVP-TE speaker. - Introduce extensions to enable a router to send tear down message to downstream router that enables the receiving router to conditionally delete its local state. See section 4.3 for more details. Chandra, et al Expires September 06, 2015 [Page 6] Internet-Draft Enhanced FRR bypass March 2015 - Enhance facility protection by allowing a PLR to directly send tear down message to MP without requiring the PLR to either have a working bypass LSP or have already signaled backup LSP state. See section 4.4 for more details. - Introduce extensions to enable the above procedures to be backward compatible with routers along the LSP path running implementation that do not support these procedures. See section 4.5 for more details. 4.1. Signaling Protection availability in Path RRO Flags This section specifies a mechanism to allow the PLR to inform the MP if local protection is available. This mechanism relies on a combination of rules around the propagation of RRO flags carried in PATH messages (Section 4.1.2) and a targeted Node-ID Hello session (Section 4.1.3). 4.1.1. PLR Behavior As per the procedures specified in RFC 4090, when a protected LSP comes up and if the "local protection desired" flag is set in the SESSION_ATTRIBUTE object, each node along the LSP path attempts to make local protection available for the LSP. - If the "node protection desired" flag is set, then the node tries to become a PLR by attempting to create a NP-bypass LSP to the NNhop node avoiding the Nhop node on protected LSP path. In case node protection could not be made available after some time out, the node attempts to create a LP-bypass LSP to Nhop node avoiding only the link that protected LSP takes to reach Nhop - If the "node protection desired" flag is not set, then the PLR attempts to create a LP-bypass LSP to Nhop node avoiding the link that the protected LSP takes to reach Nhop With regard to the PLR procedures described above and that are specified in RFC 4090, this document specifies the following recommendations involving addresses selection, and additional PLR procedures involving RRO flags carried in PATH message as well as the initiation of Node-ID based Hello sessions. - While selecting the destination address of the bypass LSP, the PLR SHOULD attempt to select the router ID of the NNhop or Nhop node. If the PLR and the MP are in same area, then the PLR may utilize the TED to determine the router ID from the interface Chandra, et al Expires September 06, 2015 [Page 7] Internet-Draft Enhanced FRR bypass March 2015 address in RRO (if NodeID is not included in RRO). If the PLR and the MP are in different IGP areas, then the PLR SHOULD use the NodeID address of NNhop MP if included in the RRO of RESV. If the NP-MP in a different area has not included NodeID in RRO, then the PLR SHOULD use NP-MP's interface address present in the RRO. The PLR SHOULD use its router ID as the source address of the bypass LSP. The PLR SHOULD also include its router ID as the NodeID in PATH RRO unless configured explicitly not to include NodeID. In parallel to the attempt made to create NP-bypass or LP-bypass, the PLR SHOULD initiate a Node-ID based Hello session to the NNhop or Nhop node respectively to establish the RSVP-TE signaling adjacency. This Hello session is used to track the state of the adjacency, including detection of adjacency failure. - If the NP-bypass LSP comes up, then the PLR SHOULD set the "local protection available" and "NP available" RRO flags and triggers PATH to be sent. - If the LP-bypass LSP comes up, then the PLR SHOULD set the "local protection available" RRO flag and triggers PATH to be sent. - After signaling protection availability, if the PLR finds that the protection becomes unavailable then it SHOULD attempt to make protection available. The PLR SHOULD wait for a time out before resetting RRO flags relating to protection availability and triggering PATH downstream. On the other hand, the PLR need not wait for a time out to set RRO flags relating to protection availability and immediately trigger PATH downstream. 4.1.2. Remote Signaling Adjacency A NodeID based RSVP-TE Hello session is one in which NodeID is used in source and destination address fields in RSVP Hello. [RFC4558] formalizes NodeID based Hello messages between two routers. This document extends NodeID based RSVP Hello session to track the state of RSVP-TE neighbor that is not directly connected by at least one interface. In order to apply NodeID based RSVP-TE Hello session between any two routers that are not immediate neighbors, the router that supports the extensions defined in the document SHOULD set TTL to 255 in the NodeID based Hello messages exchanged between PLR and MP. In the rest of the document the term "signaling adjacency", or "remote signaling adjacency" refers specifically to the RSVP-TE signaling adjacency. Chandra, et al Expires September 06, 2015 [Page 8] Internet-Draft Enhanced FRR bypass March 2015 4.1.3. PATH RRO flags Propagation As each node along the LSP path can make protection available, propagating PATH immediately due to change in RRO flags on any upstream node would increase control plane message load. So whenever a node receives PATH, it SHOULD check if the only change is in RRO flags. If the change is only in PATH RRO flags, then the node SHOULD decide whether to propagate the PATH based on the following rule. - If "NP desired" flag is set and "NP available" flag has changed in Phop's RRO flags, then PATH is triggered. - In all other cases the change is not propagated. 4.1.4. MP Behavior When the NNhop or Nhop node receives the triggered PATH with RRO flag(s) set, the node SHOULD check the presence of remote signaling adjacency with PLR (this check is needed to detect network being partitioned). If the flags are set and the RSVP-TE signaling adjacency is present, the node concludes that protection has been made available at the PLR. If the PLR has included NodeID in PATH RRO, then that NodeID is the remote neighbor address. Otherwise, the PLR's interface address in RRO will be the remote neighbor address. If the "NP available" flag is set by PPhop node, then it is NP-MP. Otherwise, it concludes it is LP-MP. 4.1.5. "Remote" state on MP Once a router concludes it is MP, it SHOULD create a remote path state for the LSP. The "remote" state is identical to the protected LSP path state except for the difference in HOP object. The HOP object corresponding to the "remote" path state contains the address of remote node signaling adjacency with PLR. The MP SHOULD consider the "remote" path state automatically deleted if: - NP-MP later receives a PATH with "NP available" flag reset in PLR's RRO flags, or - LP-MP later receives PATH with "local protection available" flag reset in PLR's RRO flags, or - Node signaling adjacency with PLR goes down, or Chandra, et al Expires September 06, 2015 [Page 9] Internet-Draft Enhanced FRR bypass March 2015 - MP receives backup LSP signaling from PLR or - MP receives PathTear, or - MP deletes the LSP state on local policy or exception event Unlike the normal path state that is either locally generated on Ingress or created from PATH message from Phop node, the "remote" path state is not signaled explicitly form PLR. The purpose of "remote" path state is to enable the PLR to explicitly tear down path and reservation states corresponding to the LSP by sending tear message for the "remote" path state. Such message tearing down "remote" path state is called "Remote PathTear. The scenarios in which "Remote" PathTear is applied are described in Section 4.4 - Remote State Teardown. 4.2. Impact of Failures on LSP State This section describes the procedures for routers on the LSP path for different kinds of failures. The procedures described on detecting RSVP control plane adjacency failures do not impact the RSVP-TE graceful restart mechanisms ([RFC3473], [RFC5063]). If the router executing these procedures act as helper for neighboring router, then the control plane adjacency will be declared as having failed after taking into account the grace period extended for neighbor by the helper. It should be noted that even though this section and the subsequent sections of the document mention "link failure" and "node failure" separately involving upstream or downstream of a protected LSP, a router implementing the procedures specified in the document need not have a mechanism to distinguish between these two types of failures. Optionally, a router MAY run Node-ID based RSVP-TE signaling adjacency with immediate neighbors to distinguish between these two types of failures. 4.2.1. Non-MP Behavior on Phop Link/Node Failure When a router detects Phop link or Phop node failure and the router is not an MP for the LSP, then it SHOULD send Conditional PathTear (refer to Section "Conditional PathTear" below) and delete PSB and RSB states corresponding to the LSP. Chandra, et al Expires September 06, 2015 [Page 10] Internet-Draft Enhanced FRR bypass March 2015 4.2.2. LP-MP Behavior on Phop Link Failure When the Phop link for an LSP fails on a router that is LP-MP for the LSP, the LP-MP SHOULD retain PSB and RSB states corresponding to the LSP till the occurrence of any of the following events. - Node-ID signaling adjacency with Phop PLR goes down, or - MP receives normal or "Remote" PathTear for PSB, or - MP receives ResvTear RSB. 4.2.3. LP-MP Behavior on Phop Node Failure When a router that is LP-MP for an LSP detects Phop node failure from Node-ID signaling adjacency state, the LP-MP SHOULD send normal PathTear and delete PSB and RSB states corresponding to the LSP. 4.2.4. NP-MP Behavior on Phop Link/Node Failure When a router that is NP-MP for an LSP detects Phop link failure, or Phop node failure from Node-ID signaling adjacency, the router SHOULD retain PSB and RSB states corresponding to the LSP till the occurrence of any of the following events. - Remote Node-ID signaling adjacency with PPhop PLR goes down, or - MP receives normal or "Remote" PathTear for PSB, or - MP receives ResvTear for RSB. 4.2.5. NP-MP Behavior on PLR Link Failure If the PLR link that is not attached to NP-MP fails and if NP-MP receives Conditional PathTear from the Phop node, then the MP SHOULD retain PSB and RSB states corresponding to the LSP till the occurrence of any of the following events. - Remote Node-ID signaling adjacency with PPhop PLR goes down, or - MP receives normal or "Remote" PathTear for PSB, or - MP receives ResvTear for RSB. Chandra, et al Expires September 06, 2015 [Page 11] Internet-Draft Enhanced FRR bypass March 2015 Receiving Conditional PathTear from the Phop node will not impact the "remote" state from the PLR. Note that Phop node would send Conditional PathTear if it was not an MP. In the example topology in Figure 1, assume C & D are NP-MP for PLRs A & B respectively. Now when A-B link fails, as B is not MP and its Phop link signaling adjacency has failed, B will delete LSP state (this behavior is required for unprotected LSPs - Section 4.2.1). In the data plane, that would require B to delete the label forwarding entry corresponding to the LSP. So if B's downstream nodes C and D continue to retain state, it would not be correct for D to continue to assume itself as NP-MP for PLR B. The mechanism that enables D to stop considering itself as NP-MP and delete "remote" path state is given below. 1. When C receives Conditional PathTear from B, it decides to retain LSP state as it is NP-MP of PLR A. C also SHOULD check whether Phop B had previously signaled availability of node protection. As B had previously signaled NP availability in its PATH RRO flags, C SHOULD reset "local protection available" and "NP available" on RRO flags corresponding to B and trigger PATH to D. 2. When D receives triggered PATH, it realizes that it is no longer NP-MP and so deletes the "remote" path state. D does not propagate PATH further down because the only change is in PATH RRO flags of B. 4.2.6. Phop Link Failure on a Node that is LP-MP and NP-MP A router may be both LP-MP as well as NP-MP at the same time for Phop and PPhop nodes respectively of an LSP. If Phop link fails on such node, the node SHOULD retain PSB and RSB states corresponding to the LSP till the occurrence of any of the following events. - Both Node-ID signaling adjacencies with Phop and PPhop nodes go down, or - MP receives normal or "Remote" PathTear for PSB, or - MP receives ResvTear for RSB. Chandra, et al Expires September 06, 2015 [Page 12] Internet-Draft Enhanced FRR bypass March 2015 4.2.7. Phop Node Failure on Node that is LP-MP and NP-MP If a router that is both LP-MP and NP-MP detects Phop node failure, then the node SHOULD retain PSB and RSB states corresponding to the LSP till the occurrence of any of the following events. - Remote Node-ID signaling adjacency with PPhop PLR goes down, or - MP receives normal or "Remote" PathTear for PSB, or - MP receives ResvTear for RSB. 4.3. Conditional Path Tear In the example provided in the Section 4.2.5 "NP-MP Behavior on PLR link failure", B deletes PSB and RSB states corresponding to the LSP once B detects its link to Phop went down as B is not MP. If B were to send PathTear normally, then C would delete LSP state immediately. In order to avoid this, there should be some mechanism by which B can indicate to C that B does not require the receiving node to unconditionally delete the LSP state immediately. For this, B SHOULD add a new optional object called CONDITIONS object in PathTear. The new optional object is defined in Section 4.3.3. If node C also understands the new object, then C SHOULD delete LSP state only if it is not an NP-MP - in other words C SHOULD delete LSP state if there is no "remote" PLR state on C. 4.3.1. Sending Conditional Path Tear A router that is not an MP for an LSP SHOULD delete PSB and RSB states corresponding to the LSP if Phop link or Phop Node-ID signaling adjacency goes down (Section 4.2.1). The router SHOULD send Conditional PathTear if the following are also true. - Ingress has requested node protection for the LSP, and - PathTear is not received from upstream node 4.3.2. Processing Conditional Path Tear When a router that is not an NP-MP receives Conditional PathTear, the node SHOULD delete PSB and RSB states corresponding to the LSP, and process Conditional PathTear by considering it as normal PathTear. Specifically, the node SHOULD NOT propagate Conditional PathTear downstream but remove the optional object and send normal PathTear downstream. Chandra, et al Expires September 06, 2015 [Page 13] Internet-Draft Enhanced FRR bypass March 2015 When a node that is an NP-MP receives Conditional PathTear, it SHOULD NOT delete LSP state. The node SHOULD check whether the Phop node previously set "NP available" flag in PATH RRO flags. If the flag had been set previously by Phop, then the node SHOULD clear "local protection available" and "NP available" flags in Phop's RRO flags and trigger PATH downstream. If Conditional PathTear is received from a neighbor that has not advertised support (refer to Section 4.5) for the new procedures defined in this document, then the node SHOULD consider the message as normal PathTear. The node SHOULD propagate normal PathTear downstream and delete LSP state. 4.3.3. CONDITIONS object As any implementation that does not support Conditional PathTear SHOULD ignore the new object but process the message as normal PathTear without generating any error, the Class-Num of the new object SHOULD be 10bbbbbb where 'b' represents a bit (from Section 3.10 of [RFC2205]). The new object is called as "CONDITIONS" object that will specify the conditions under which default processing rules of the RSVP-TE message SHOULD be invoked. The object has the following format: +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Length | Class | C-type | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Reserved |M| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Length This contains the size of the object in bytes and should be set to eight. Class TBD C-type 1 Chandra, et al Expires September 06, 2015 [Page 14] Internet-Draft Enhanced FRR bypass March 2015 M bit This bit indicates that the message SHOULD be processed based on the condition whether the receiving node is Merge Point or not. 4.4. Remote State Teardown If the Ingress wants to tear down the LSP because of a management event while the LSP is being locally repaired at a transit PLR, it would not be desirable to wait till backup LSP signaling to perform state cleanup. To enable LSP state cleanup when the LSP is being locally repaired, the PLR SHOULD send "remote" PathTear message instructing the MP to delete PSB and RSB states corresponding to the LSP. Consider node C in example topology (Figure 1) has gone down and B locally repairs the LSP. 1. Ingress A receives a management event to tear down the LSP. 2. A sends normal PathTear to B. 3. To enable LSP state cleanup, B SHOULD send "remote" PathTear with destination IP address set to that of D used in Node-ID signaling adjacency with D, and HOP object containing local address used in Node-ID signaling adjacency. 4. B then deletes PSB and RSB states corresponding to the LSP. 5. On D there would be a remote signaling adjacency with B and so D SHOULD accept the remote PathTear and delete PSB and RSB states corresponding to the LSP. 4.4.1. PLR Behavior on Local Repair Failure If local repair fails on the PLR after a failure, then this should be considered as a case for cleaning up LSP state from PLR to the Egress. PLR would achieve this using "remote" PathTear to clean up state from MP. If MP has retained state, then it would propagate PathTear downstream thereby achieving state cleanup. Note that in the case of link protection, the PathTear would be directed to LP-MP node IP address rather than the Nhop interface address. 4.4.2. PLR Behavior on Resv RRO Change When a router that has already made NP available detects a change in the RRO carried in RESV message, and if the RRO change indicates Chandra, et al Expires September 06, 2015 [Page 15] Internet-Draft Enhanced FRR bypass March 2015 that the router's former NP-MP is no longer present in the LSP path, then the router SHOULD send "Remote" PathTear directly to its former NP-MP. In the example topology in Figure 1, assume A has made node protection available and C has concluded it is NP-MP. When the B-C link fails then implementing the procedure specified in Section 4.2.4 of this document, C will retain state till: remote NodeID control plane adjacency with A goes down, or PathTear or ResvTear is received for PSB or RSB respectively. If B also has made node protection available, B will eventually complete backup LSP signaling with its NP-MP D and trigger RESV to A with RRO changed. The new RRO of the LSP carried in RESV will not contain C. When A processes the RESV with a new RRO not containing C - its former NP- MP, A SHOULD send "Remote" PathTear to C. When C receives a "Remote" PathTear for its PSB state, C will send normal PathTear downstream to D and delete both PSB and RSB states corresponding to the LSP. As D has already received backup LSP signaling from B, D will retain control plane and forwarding states corresponding to the LSP. 4.4.3. LSP Preemption during Local Repair If an LSP is preempted when there is no failure along the path of the LSP, the node on which preemption occurs would send PathErr and ResvTear upstream and only delete the forwarding state and RSB state corresponding to the LSP. But if the LSP is being locally repaired upstream of the node on which the LSP is preempted, then the node SHOULD delete both PSB and RSB states corresponding to the LSP and send normal PathTear downstream. 4.4.3.1. Preemption on LP-MP after Phop Link failure If an LSP is preempted on LP-MP after its Phop or incoming link has already failed but the backup LSP has not been signaled yet, then the node SHOULD send normal PathTear and delete both PSB and RSB states corresponding to the LSP. As the LP-MP has retained LSP state because the PLR would signal the LSP through backup LSP signaling, preemption would bring down the LSP and the node would not be LP-MP any more requiring the node to clean up LSP state. 4.4.3.2. Preemption on NP-MP after Phop Link failure If an LSP is preempted on NP-MP after its Phop link has already failed but the backup LSP has not been signaled yet, then the node SHOULD send normal PathTear and delete PSB and RSB states corresponding to the LSP. As the NP-MP has retained LSP state Chandra, et al Expires September 06, 2015 [Page 16] Internet-Draft Enhanced FRR bypass March 2015 because the PLR would signal the LSP through backup LSP signaling, preemption would bring down the LSP and the node would not be NP-MP any more requiring the node to clean up LSP state. Consider B-C link goes down on the same example topology (Figure 1). As C is NP-MP for PLR A, C will retain LSP state. 1. The LSP is preempted on C. 2. C will delete RSB state corresponding to the LSP. But C cannot send PathErr or ResvTear to PLR A because backup LSP has not been signaled yet. 3. As the only reason for C having retained state after Phop node failure was that it was NP-MP, C SHOULD send normal PathTear to D and delete PSB state also. D would also delete PSB and RSB states on receiving PathTear from C. 4. B starts backup LSP signaling to D. But as D does not have the LSP state, it will reject backup LSP PATH and send PathErr to B. 5. B will delete its reservation and send ResvTear to A. 4.5. Backward Compatibility Procedures The "Enhanced FRR facility protection" referred below in this section refers to the set of changes that have been proposed in previous sections. Any implementation that does not support them has been termed as "existing implementation". Of the proposed extensions, signaling protection using RRO flags is expected to be backward compatible and can work safely irrespective of whether the refresh time is small or arbitrarily long. This is because the existing implementations would not send error or tear down message in response to the flags in PATH RRO but would simply ignore and propagate them. On the other hand, changes proposed relating to LSP state cleanup namely Conditional and remote PathTear require support from other nodes along the LSP path. So procedures that fall under LSP state cleanup category SHOULD be turned on only if all nodes involved in the node protection FRR i.e. PLR, MP and intermediate node in the case of NP, support the extensions. Note that for LSPs requesting only link protection, the PLR and the LP-MP should support the extensions. Chandra, et al Expires September 06, 2015 [Page 17] Internet-Draft Enhanced FRR bypass March 2015 4.5.1. Detecting Support for Enhanced FRR Facility Protection An implementation supporting the FRR facility protection extensions specified in previous sections SHOULD set a new flag "Enhanced facility protection" in CAPABILITY object in Hello messages. - As nodes supporting the extensions SHOULD initiate Node Hellos with adjacent nodes, a node on the path of protected LSP can determine whether its Phop or Nhop neighbor supports FRR enhancements from the Hello messages sent by the neighbor. - If a node attempts to make node protection available, then the PLR SHOULD initiate remote Node-ID signaling adjacency with NNhop. If the NNhop (a) does not reply to remote node Hello message or (b) does not set "Enhanced facility protection" flag in CAPABILITY object in the reply, then the PLR can conclude that NNhop does not support FRR extensions. - If node protection is requested for an LSP and if (a) PPhop node has not set "local protection available" and "NP available" flags in its RRO flags or (b) PPhop node has not initiated remote node Hello messages, then the node SHOULD conclude that PLR does not support FRR extensions. The details are described in the "Procedures for backward compatibility" section below. The new flag that will be introduced to CAPABILITY object is specified below. 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 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Length | Class-Num(134)| C-Type (1) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Reserved |E|T|R|S| +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ E bit Indicates that the sender supports Enhanced FRR facility protection Any node that sets the new E-bit is set in its CAPABILITY object MUST also set Refresh-Reduction-Capable bit in common header of all RSVP-TE messages. Chandra, et al Expires September 06, 2015 [Page 18] Internet-Draft Enhanced FRR bypass March 2015 4.5.2. Procedures for backward compatibility The procedures defined hereafter are performed on a subset of LSPs that traverse a node, rather than on all LSPs that traverse a node. This behavior is required to support backward compatibility for a subset of LSPs traversing nodes running existing implementations. 4.5.2.1. Lack of support on Downstream Node - If the Nhop does not support enhanced facility protection FRR, then the node SHOULD reduce the "refresh period" in TIME_VALUES object carried in PATH to default small refresh default value. - If node protection is requested and the NNhop node does not support the enhancements, then the node SHOULD reduce the "refresh period" in TIME_VALUES object carried in PATH to a small refresh default value. If the node reduces the refresh time from the above procedures, it SHOULD also not send remote PathTear or Conditional PathTear messages. Consider the example topology in Figure 1. If C does not support scalability improvements, then: - A and B SHOULD reduce the refresh time to default value of 30 seconds and trigger PATH - If B is not an MP and if Phop link of B fails, B cannot send Conditional PathTear to C but SHOULD time out PSB state from A normally. This would be accomplished if A would also reduce the refresh time to default value. So if C does not support enhanced facility protection, then Phop B and PPhop A SHOULD reduce refresh time to a small default value. 4.5.2.2. Lack of support on Upstream Node - If Phop node does not support enhanced facility protection, then the node SHOULD reduce the "refresh period" in TIME_VALUES object carried in RESV to default small refresh time value. - If node protection is requested and the Phop node does not support the enhancements, then the node SHOULD reduce the "refresh period" in TIME_VALUES object carried in PATH to default value. Chandra, et al Expires September 06, 2015 [Page 19] Internet-Draft Enhanced FRR bypass March 2015 - If node protection is requested and PPhop node does not support the enhancements, then the node SHOULD reduce the "refresh period" in TIME_VALUES object carried in RESV to default value. - If the node reduces the refresh time from the above procedures, it SHOULD also not execute MP determination procedures. 4.5.2.3. Incremental Deployment The backward compatibility procedures described in the previous sub- sections imply that a router supporting the FRR extensions specified in this document can apply the procedures specified in the document either in the downstream or upstream direction of an LSP, depending on the capability of the routers downstream or upstream in the LSP path. - FRR extensions and procedures are enabled for downstream Path, PathTear and ResvErr messages corresponding to an LSP if link protection is requested for the LSP and the Nhop node supports the extensions - FRR extensions and procedures are enabled for downstream Path, PathTear and ResvErr messages corresponding to an LSP if node protection is requested for the LSP and both Nhop & NNhop nodes support the extensions - FRR extensions and procedures are enabled for upstream PathErr, Resv and ResvTear messages corresponding to an LSP if link protection is requested for the LSP and the Phop node supports the extensions - FRR extensions and procedures are enabled for upstream PathErr, Resv and ResvTear messages corresponding to an LSP if node protection is requested for the LSP and both Phop and PPhop nodes support the extensions For example, if implementation supporting the FRR extensions specified in this document is deployed on all routers in particular region of the network and if all the LSPs in the network request node protection, then the FRR extensions will only be applied for the LSP segments that traverse the particular region. This will aid incremental deployment of these extensions and also allow reaping the benefits of the extensions in portions of the network where it is supported. Chandra, et al Expires September 06, 2015 [Page 20] Internet-Draft Enhanced FRR bypass March 2015 5. Security Considerations This document extends the applicability of Node-ID based Hello session between immediate neighbors. The Node-ID based Hello session between PLR and NP-MP may require the two routers to exchange Hello messages with non-immediate neighbor. So, the implementations SHOULD provide the option to configure Node-ID neighbor specific or global authentication key to authentication messages received from Node-ID neighbors. The network administrator MAY utilize this option to enable RSVP-TE routers to authenticate Node-ID Hello messages received with TTL greater than 1. 6. IANA Considerations 6.1. New Object - CONDITIONS [RFC2205] defines the Class-Number name space for RSVP objects. The name space is managed by IANA. IANA registry: RSVP Parameters Subsection: Class Names, Class Numbers, and Class Types A new RSVP object using a Class-Number of form 10bbbbbb called the "CONDITIONS" object is defined in Section 4.3 of this document. The Class-Number is TBD. 6.2. New CAPABILITY Object value [RFC5063] defines the name space for RSVP Capability Object Values. The name space is managed by IANA. IANA registry: RSVP PARAMETERS Subsection: Capability Object Values A new Capability flag called "Enhanced FRR facility protection" is defined in Section 4.5 of this document. The bit number for this flag is TBD. 7. Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [RFC4090] Pan, P., "Fast Reroute Extensions to RSVP-TE for LSP Tunnels", RFC 4090, May 2005. Chandra, et al Expires September 06, 2015 [Page 21] Internet-Draft Enhanced FRR bypass March 2015 [RFC2961] Berger, L., "RSVP Refresh Overhead Reduction Extensions", RFC 2961, April 2001. [RFC3209] Awduche, D., "RSVP-TE: Extensions to RSVP for LSP Tunnels", RFC 3209, December 2001. [RFC2205] Braden, R., "Resource Reservation Protocol (RSVP)", RFC 2205, September 1997. [RFC4558] Ali, Z., "Node-ID Based Resource Reservation (RSVP) Hello: A Clarification Statement", RFC 4558, June 2006. 8. Acknowledgments Thanks to Raveendra Torvi and Yimin Shen for their comments and inputs. 9. Authors' Addresses Chandra Ramachandran Juniper Networks Email: csekar@juniper.net Yakov Rekhter Juniper Networks Email: yakov@juniper.net Ina Minei Google, Inc inaminei@google.com Ebben Aries Facebook Email: exa@fb.com Dante Pacella Verizon Email: dante.j.pacella@verizon.com Markus Jork Juniper Networks Email: mjork@juniper.net Harish Sitaraman Chandra, et al Expires September 06, 2015 [Page 22] Internet-Draft Enhanced FRR bypass March 2015 Juniper Networks Email: hsitaraman@juniper.net Vishnu Pavan Beeram Juniper Networks Email: vbeeram@juniper.net Chandra, et al Expires September 06, 2015 [Page 23]