idnits 2.17.1 draft-ietf-mpls-bgp-mpls-restart-01.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** Looks like you're using RFC 2026 boilerplate. This must be updated to follow RFC 3978/3979, as updated by RFC 4748. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- ** The document seems to lack a 1id_guidelines paragraph about 6 months document validity -- however, there's a paragraph with a matching beginning. Boilerplate error? == No 'Intended status' indicated for this document; assuming Proposed Standard == The page length should not exceed 58 lines per page, but there was 9 longer pages, the longest (page 2) being 61 lines Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an Introduction section. ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) ** The document seems to lack an Authors' Addresses Section. ** The abstract seems to contain references ([RFC2119], [1]), which it shouldn't. Please replace those with straight textual mentions of the documents in question. Miscellaneous warnings: ---------------------------------------------------------------------------- -- The document seems to lack a disclaimer for pre-RFC5378 work, but may have content which was first submitted before 10 November 2008. If you have contacted all the original authors and they are all willing to grant the BCP78 rights to the IETF Trust, then this is fine, and you can ignore this comment. If not, you may need to add the pre-RFC5378 disclaimer. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- Couldn't find a document date in the document -- date freshness check skipped. Checking references for intended status: Proposed Standard ---------------------------------------------------------------------------- (See RFCs 3967 and 4897 for information about using normative references to lower-maturity documents in RFCs) == Outdated reference: A later version (-13) exists of draft-ietf-idr-restart-01 ** Obsolete normative reference: RFC 3107 (ref. '2') (Obsoleted by RFC 8277) ** Obsolete normative reference: RFC 2385 (Obsoleted by RFC 5925) Summary: 8 errors (**), 0 flaws (~~), 3 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group Yakov Rekhter (Juniper Networks) 3 Internet Draft Rahul Aggarwal (Redback Networks) 4 Expiration Date: March 2003 6 Graceful Restart Mechanism for BGP with MPLS 8 draft-ietf-mpls-bgp-mpls-restart-01.txt 10 Status of this Memo 12 This document is an Internet-Draft and is in full conformance with 13 all provisions of Section 10 of RFC2026. 15 Internet-Drafts are working documents of the Internet Engineering 16 Task Force (IETF), its areas, and its working groups. Note that other 17 groups may also distribute working documents as Internet-Drafts. 19 Internet-Drafts are draft documents valid for a maximum of six months 20 and may be updated, replaced, or obsoleted by other documents at any 21 time. It is inappropriate to use Internet-Drafts as reference 22 material or to cite them other than as ``work in progress.'' 24 The list of current Internet-Drafts can be accessed at 25 http://www.ietf.org/ietf/1id-abstracts.txt 27 The list of Internet-Draft Shadow Directories can be accessed at 28 http://www.ietf.org/shadow.html. 30 Abstract 32 A mechanism for BGP that would help minimize the negative effects on 33 routing caused by BGP restart is described in "Graceful Restart 34 Mechanism for BGP" (see [1]). This document extends this mechanism to 35 also minimize the negative effects on MPLS forwarding caused by the 36 Label Switching Router's (LSR's) control plane restart, and 37 specifically by the restart of its BGP component when BGP is used to 38 carry MPLS labels and the LSR is capable of preserving the MPLS 39 forwarding state across the restart. 41 The mechanism described in this document is agnostic with respect to 42 the types of the addresses carried in the BGP Network Layer 43 Reachability Information (NLRI) field. As such it works in 44 conjunction with any of the address famililies that could be carried 45 in BGP (e.g., IPv4, IPv6, etc...) 46 The mechanism described in this document is applicable to all LSRs, 47 both those with the ability to preserve their forwarding state during 48 BGP restart and those without (although the latter need to implement 49 only a subset of the mechanism described in this document). 50 Supporting (a subset of) the mechanism described here by the LSRs 51 that can not preserve their MPLS forwarding state across the restart 52 would not reduce the negative impact on MPLS traffic caused by their 53 control plane restart, but it would minimize the impact if their 54 neighbor(s) are capable of preserving the forwarding state across the 55 restart of their control plane and implement the mechanism described 56 here. 58 The mechanism makes minimalistic assumptions on what has to be 59 preserved across restart - the mechanism assumes that only the actual 60 MPLS forwarding state has to be preserved; the mechanism does not 61 require any of the BGP-related state to be preserved across the 62 restart. 64 Specification of Requirements 66 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 67 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 68 document are to be interpreted as described in RFC 2119 [RFC2119]. 70 Summary for Sub-IP Area 72 Summary 74 This document describes a mechanism that helps to minimize the 75 negative effects on MPLS forwarding caused by LSR's control plane 76 restart, and specifically by the restart of its BGP component in the 77 case where BGP is used to carry MPLS labels and LSR is capable of 78 preserving its MPLS forwarding state across the restart. 80 Related documents 82 See the Reference Section 84 Where does it fit in the Picture of the Sub-IP Work 86 This work fits squarely in MPLS box. 88 Why is it Targeted at this WG 90 The specifications on carrying MPLS Labels in BGP is a product of the 91 MPLS WG. This document specifies procedures to minimize the negative 92 effects on MPLS forwarding caused by the restart of the control plane 93 BGP module in the case where BGP is used to carry MPLS labels. Since 94 the procedures described in this document are directly related to 95 MPLS forwarding and carrying MPLS labels in BGP, it would be logical 96 to target this document at the MPLS WG. 98 Justification 100 The WG should consider this document, as it allows to minimize the 101 negative effects on MPLS forwarding caused by the restart of the 102 control plane BGP module in the case where BGP is used to carry MPLS 103 labels. 105 1. Motivation 107 For the sake of brevity in the context of this document by "MPLS 108 forwarding state" we mean either (outgoing label, 109 next hop)>, or
(outgoing label, next hop)> 110 mapping. In the context of this document the forwarding state that is 111 referred to in [1] means MPLS forwarding state. 113 In the case where a Label Switching Router (LSR) could preserve its 114 MPLS forwarding state across restart of its control plane, and 115 specifically its BGP component, and BGP is used to carry MPLS labels 116 (as specified in [2]), it may be desirable not to perturb the LSPs 117 going through that LSR (and specifically, the LSPs established by 118 BGP). In this document, we describe a mechanism that allows to 119 accomplish this goal. The mechanism described in this document works 120 in conjunction with the mechanism specified in [1]. The mechanism 121 described in this document places no restrictions on the types of 122 addresses (address families) that it can support. 124 The mechanism described in this document is applicable to all LSRs, 125 both those with the ability to preserve forwarding state during BGP 126 restart and those without (although the latter need to implement only 127 a subset of the mechanism described in this document). Supporting (a 128 subset of) the mechanism described here by the LSRs that can not 129 preserve their MPLS forwarding state across the restart would not 130 reduce the negative impact on MPLS traffic caused by their control 131 plane restart, but it would minimize the impact if their neighbor(s) 132 are capable of preserving the forwarding state across the restart of 133 their control plane and implement the mechanism described here. 135 2. Assumptions 137 First of all we assume that an LSR implements the Graceful Restart 138 Mechanism for BGP, as specified in [1]. Second, we assume that the 139 LSR is capable of preserving its MPLS forwarding state across the 140 restart of its control plane (including the restart of BGP). 142 The mechanism makes minimalistic assumptions on what has to be 143 preserved across restart - the mechanism assumes that only the actual 144 MPLS forwarding state has to be preserved; the mechanism does not 145 require any of the BGP-related state to be preserved across the 146 restart. 148 In the scenario where label binding on an LSR is created/maintained 149 not just by the BGP component of the control plane, but by other 150 protocol components as well (e.g., LDP, RSVP-TE), and the LSR 151 supports restart of the individual components of the control plane 152 that create/maintain label binding (e.g., restart of BGP, but no 153 restart of LDP) the LSR needs to preserve across the restart the 154 information about which protocol has assigned which labels. 156 3. Capability Advertisement 158 An LSR that supports the mechanism described in this document 159 advertises this to its peer by using the Graceful Restart Capability, 160 as specified in [1]. The Subsequent Address Family Identifier (SAFI) 161 in the advertised capability MUST indicate that the Network Layer 162 Reachability Information (NLRI) field carries not just addressing 163 information but labels as well (see [2]). 165 4. Procedures for the restarting LSR 167 After the LSR restarts, it follows the procedures as specified in 168 [1]. In addition, if the LSR is able to preserve its MPLS forwarding 169 state across the restart, the LSR advertises this to its neighbors by 170 appropriately setting the Flag field in the Graceful Restart 171 Capability for all applicable AFI/SAFI pairs. 173 Once the restarting LSR completes its route selection (as specified 174 in Section "Procedures for the Restarting Speaker" of [1]), then in 175 addition to the procedures specified in [1], the restarting LSR 176 performs one of the following: 178 4.1. Case 1 180 The following applies when (a) the best route selected by the 181 restarting LSR was received with a label, (b) that label is not an 182 Implicit NULL, and (c) the LSR advertises this route with itself as 183 the next hop. 185 In this case the restarting LSR searches its MPLS forwarding state 186 (the one preserved across the restart) for an entry with equal to the one in the received route. If such an 188 entry is found, the LSR no longer marks the entry as stale. In 189 addition if the entry is of type rather than , the LSR 191 uses the incoming label from the entry when advertising the route to 192 its neighbors. If the found entry has no incoming label, or if no 193 such entry is found, the LSR just picks up some unused label when 194 advertising the route to its neighbors (assuming that there are 195 neighbors to which the LSR has to advertise the route with a label). 197 4.2. Case 2 199 The following applies when (a) the best route selected by the 200 restarting LSR was received either without a label, or with an 201 Implicit NULL label, or the route is originated by the restarting 202 LSR, (b) the LSR advertises this route with itself as the next hop, 203 and (c) the LSR has to generate a (non Implicit NULL) label for the 204 route. 206 In this case the LSR searches its MPLS forwarding state for an entry 207 that indicates that the LSR has to perform label pop, and the next 208 hop equal to the next hop of the route in consideration. If such an 209 entry is found, then the LSR uses the incoming label from the entry 210 when advertising the route to its neighbors. If no such entry is 211 found, the LSR just picks up some unused label when advertising the 212 route to its neighbors. 214 The description in the above paragraph assumes that the restarting 215 LSR generates the same label for all the routes with the same next 216 hop. If this is not the case, and the restarting LSR generates a 217 unique label per each such route, then the LSR needs to preserve 218 across the restart not just mapping, but also the prefix associated with this mapping. In 220 such case the LSR would search its MPLS forwarding state for an entry 221 that (a) indicates Label pop (means no outgoing label), (b) the next 222 hop equal to the next hop of the route and (c) has the same prefix as 223 the route. If such an entry is found, then the LSR uses the incoming 224 label from the entry when advertising the route to its neighbors. If 225 no such entry is found, the LSR just picks up some unused label when 226 advertising the route to its neighbors. 228 4.3. Case 3 230 The following applies when the restarting LSR does not set BGP Next 231 Hop to self. 233 In this case the restarting LSR, when advertising its best route for 234 a particular NLRI just uses the label that was received with that 235 route. And if the route was received with no label, the LSR 236 advertises the route with no label as well. 238 5. Alternative procedures for the restarting LSR 240 In this section we describe an alternative to the procedures 241 described in Section "Procedures for the restarting LSR". 243 The procedures described in this section assume that the restarting 244 LSR has (at least) as many unallocated as allocated labels. The 245 latter forms the MPLS forwarding state that the LSR managed to 246 preserve across the restart. The former is used for allocating labels 247 after the restart. 249 After the LSR restarts, it follows the procedures as specified in 250 [1]. In addition, if the LSR is able to preserve its MPLS forwarding 251 state across the restart, the LSR advertises this to its neighbors by 252 appropriately setting the Flag field in the Graceful Restart 253 Capability. 255 To create local label bindings the LSR uses unallocated labels (this 256 is pretty much the normal procedure). That means that as long as the 257 LSR retains the MPLS forwarding state that the LSR preserved across 258 the restart, the labels from that state are not used for creating 259 local label bindings. 261 The restarting LSR SHOULD retain the MPLS forwarding state that the 262 LSR preserved across the restart at least until the LSR sends End-of- 263 RIB marker to all of its neighbors (by that time the LSR already 264 completed its route selection process, and also advertised its Adj- 265 RIB-Out to its neighbors). The restarting LSR MAY retain the 266 forwarding state even a bit longer, as to allow the neighbors to 267 receive and process the routes that have been advertised by the 268 restarting LSR. After that, the restarting LSR MAY delete the MPLS 269 forwarding state that it preserved across the restart. 271 Note that while an LSR is in the process of restarting, the LSR may 272 have not one, but two local label bindings for a given BGP route - 273 one that was retained from prior to restart, and another that was 274 created after the restart. Once the LSR completes its restart, the 275 former will be deleted. Both of these bindings though would have the 276 same outgoing label (and the same next hop). 278 6. Procedures for a neighbor of a restarting LSR 280 The neighbor of a restarting LSR (the receiving router in terminology 281 used in [1]) follows the procedures specified in [1]. In addition, 282 the neighbor treats the MPLS labels received from the restarting LSR 283 the same way as it treats the routes received from the restarting LSR 284 (both prior and after the restart). 286 Replacing the stale routes by the routing updates received from the 287 restarting LSR involves replacing/updating the appropriate MPLS 288 labels. 290 In addition, if the Flags in the Graceful Restart Capability received 291 from the restarting LSR indicate that the LSR wasn't able to retain 292 its MPLS state across the restart, the neighbor SHOULD immediately 293 remove all the NLRI and the associated MPLS labels that it previously 294 acquired via BGP from the restarting LSR. 296 An LSR, once it creates a binding between a label and a Forwarding 297 Equivalence Class (FEC), SHOULD keep the value of the label in this 298 binding for as long as the LSR has a route to the FEC in the binding. 299 If the route to the FEC disappears, and then re-appears again later, 300 then this may result in using a different label value, as when the 301 route re-appears, the LSR would create a new binding. 303 To minimize the potential mis-routing caused by the label change, 304 when creating a new binding the LSR SHOULD pick up the 305 least recently used label. Once an LSR releases a label, the LSR 306 SHALL NOT re-use this label for advertising a binding to 307 a neighbor that supports graceful restart for at least the Restart 308 Time, as advertised by the neighbor to the LSR. 310 7. Security Consideration 312 The security considerations pertaining to the original BGP protocol 313 remain relevant. 315 In addition, the mechanism described here renders LSRs that implement 316 it to additional denial-of-service attacks as follows: 318 An intruder may impersonate a BGP peer in order to force a failure 319 and reconnection of the TCP connection, but where the intruder 320 sets the Forwarding State (F) bit (as defined in [1]) to 0 on 321 reconnection. This forces all labels received from the peer to be 322 released. 324 An intruder could intercept the traffic between BGP peers and 325 override the setting of the Forwarding State (F) bit to be set to 326 0. This forces all labels received from the peer to be released. 328 All of these attacks may be countered by use of an authentication 329 scheme between BGP peers, such as the scheme outlined in [RFC2385]. 331 As with BGP carrying labels, a security issue may exist if a BGP 332 implementation continues to use labels after expiration of the BGP 333 session that first caused them to be used. This may arise if the 334 upstream LSR detects the session failure after the downstream LSR has 335 released and re-used the label. The problem is most obvious with the 336 platform-wide label space and could result in mis-routing of data to 337 other than intended destinations and it is conceivable that these 338 behaviors may be deliberately exploited to either obtain services 339 without authorization or to deny services to others. 341 In this document, the validity of the BGP session may be extended by 342 the Restart Time, and the session may be re-established in this 343 period. After the expiry of the Restart Time the session must be 344 considered to have failed and the same security issue applies as 345 described above. 347 However, the downstream LSR may declare the session as failed before 348 the expiration of its Restart Time. This increases the period during 349 which the downstream LSR might reallocate the label while the 350 upstream LSR continues to transmit data using the old usage of the 351 label. To reduce this issue, this document requires that labels are 352 not re-used until for at least the Restart Time. 354 8. Intellectual Property Considerations 356 Juniper Networks, Inc. is seeking patent protection on some or all of 357 the technology described in this Internet-Draft. If technology in 358 this document is adopted as a standard, Juniper Networks agrees to 359 license, on reasonable and non-discriminatory terms, any patent 360 rights it obtains covering such technology to the extent necessary to 361 comply with the standard. 363 Redback Networks, Inc. is seeking patent protection on some of the 364 technology described in this Internet-Draft. If technology in this 365 document is adopted as a standard, Redback Networks agrees to 366 license, on reasonable and non-discriminatory terms, any patent 367 rights it obtains covering such technology to the extent necessary to 368 comply with the standard. 370 9. Acknowledgments 372 We would like to thank Chaitanya Kodeboyina and Loa Andersson for 373 their review and comments. The approach described in Section 374 "Alternative procedures for the restarting LSR" is based on the idea 375 suggested by Manoj Leelanivas. 377 10. Normative References 379 [1] "Graceful Restart Mechanism for BGP", draft-ietf-idr- 380 restart-01.txt 382 [2] Rekhter, Y., Rosen, E., "Carrying Label Information in BGP-4", 383 RFC3107 385 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 386 Requirement Levels", BCP 14, RFC 2119 388 [RFC2385] Heffernan, A., "Protection of BGP Sessions via the TCP MD5 389 Signature Option", RFC2385 390 11. Author Information 392 Yakov Rekhter 393 Juniper Networks 394 1194 N.Mathilda Ave 395 Sunnyvale, CA 94089 396 e-mail: yakov@juniper.net 398 Rahul Aggarwal 399 Redback Networks 400 350 Holger Way 401 San Jose, CA 95134 402 e-mail: rahul@redback.com