RFC 8563
Bidirectional Forwarding Detection (BFD) Multipoint Active Tails
Bidirectional Forwarding Detection (BFD) Multipoint Active Tails ✨
Internet Engineering Task Force (IETF) 🎀 · Request for Comments: 8563 🦋 · ISSN: 2070-1721
Okay so this is RFC 8563 💖, and it's a Standards Track document from April 2019 🌸. The authors are D. Katz (Juniper Networks), D. Ward (Cisco Systems), S. Pallagatti, Ed. (VMware), and G. Mirsky, Ed. (ZTE Corp.) 👑. It doesn't obsolete or update anything, so it's a brand-new extension 🧚. It's bestie-approved as an official Internet Standard, fr fr 💅.
Abstract ✨
This document describes active tail extensions to the Bidirectional Forwarding Detection (BFD) protocol for multipoint networks 🦋.
Status of This Memo 📜
This is an Internet Standards Track document 🌈.
This document is a product of the Internet Engineering Task Force (IETF) 💜. It represents the consensus of the IETF community 🫶. It has received public review and has been approved for publication by the Internet Engineering Steering Group (IESG) ✨. Further information on Internet Standards is available in Section 2 of RFC 7841 🎀.
Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at https://www.rfc-editor.org/info/rfc8563 🔥.
Copyright Notice ©️
Copyright (c) 2019 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 (https://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 💜.
Table of Contents 🍓
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Terminology and Acronyms . . . . . . . . . . . . . . . . . . 3
3. Keywords . . . . . . . . . . . . . . . . . . . . . . . . . . 3
4. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . 4
5. Operational Scenarios . . . . . . . . . . . . . . . . . . . . 5
5.1. No Head Notification . . . . . . . . . . . . . . . . . . 5
5.2. Head Notification . . . . . . . . . . . . . . . . . . . . 5
5.2.1. Head Notification without Polling . . . . . . . . . . 5
5.2.2. Head Notification and Tail Solicitation with
Multipoint Polling . . . . . . . . . . . . . . . . . 6
5.2.3. Head Notification with Composite Polling . . . . . . 6
6. Protocol Details . . . . . . . . . . . . . . . . . . . . . . 7
6.1. Multipoint Client Session . . . . . . . . . . . . . . . . 8
6.2. Multipoint Client Session Failure . . . . . . . . . . . . 8
6.3. State Variables . . . . . . . . . . . . . . . . . . . . . 8
6.3.1. New State Variables . . . . . . . . . . . . . . . . . 8
6.3.2. New State Variable Value . . . . . . . . . . . . . . 9
6.3.3. State Variable Initialization and Maintenance . . . . 10
6.4. Controlling Multipoint BFD Options . . . . . . . . . . . 11
6.5. State Machine . . . . . . . . . . . . . . . . . . . . . . 11
6.6. Session Establishment . . . . . . . . . . . . . . . . . . 12
6.7. Discriminators and Packet Demultiplexing . . . . . . . . 12
6.8. Controlling Tail Packet Transmission . . . . . . . . . . 12
6.9. Soliciting the Tails . . . . . . . . . . . . . . . . . . 13
6.10. Verifying Connectivity to Specific Tails . . . . . . . . 13
6.11. Detection Times . . . . . . . . . . . . . . . . . . . . . 14
6.12. MultipointClient Down/AdminDown Sessions . . . . . . . . 15
6.13. Base BFD for Multipoint Networks Specification Text
Replacement . . . . . . . . . . . . . . . . . . . . . . . 15
6.13.1. Reception of BFD Control Packets . . . . . . . . . . 15
6.13.2. Demultiplexing BFD Control Packets . . . . . . . . . 16
6.13.3. Transmitting BFD Control Packets . . . . . . . . . . 16
7. Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . 17
8. Operational Considerations . . . . . . . . . . . . . . . . . 18
9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 18
10. Security Considerations . . . . . . . . . . . . . . . . . . . 18
11. Normative References . . . . . . . . . . . . . . . . . . . . 19
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . 19
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 201. Introduction 🦋
Okay so this is an extension to Multipoint BFD [RFC8562] ✨, and it lets tails tell the head when they've lost multipoint connectivity 💅. As an extra bonus, heads can ask the tails for a notification by using a polling mechanism 🫶. Notification can be turned on for all tails, or just for a subset of them 🌈.
The goal here is for the head to have reasonably rapid knowledge of tails that have lost connectivity from the head 🔥.
Scaling is a primary concern 😭, especially state explosion toward the head 💀. So it is required that the head be in control of all timing aspects of the mechanism 👑, and that BFD packets from the tails to the head not be synchronized 🧚.
Throughout this document, the term "multipoint" is defined as a mechanism by which one or more systems receive packets sent by a single sender 🌸. This specifically includes things like IP multicast and point-to-multipoint MPLS 💜.
The term "connectivity" in this document is not being used in the context of connectivity verification in a transport network 🙈. Instead, it's an alternative to "continuity", i.e., the existence of a path between the sender and the receiver 🍓.
This document effectively modifies and adds to Sections 5.12 and 5.13 of the base BFD multipoint networks specification [RFC8562] 🎀.
2. Terminology and Acronyms 📖
- BFD: Bidirectional Forwarding Detection 💖
- c-poll: Composite Poll 🦋
- m-poll: Multipoint Poll 🌸
3. Keywords 🔑
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all capitals, as shown here 💜.
4. Overview 🌈
Okay so a head might want to know about the tails' connectivity (or the lack of it), and there are a bunch of ways to do that ✨. First, if all that's needed is a best-effort failure notification, as discussed in Section 5.2.1, the tails can send unsolicited unicast BFD Control packets to the head when the path fails, as described in Section 6.4 💅.
If the head wants to know which tails are active on the multipoint path, it can send a multipoint BFD Control packet with the Poll (P) bit set 🫶. That makes the tails send back a unicast BFD Control packet with the Final (F) bit set (see a detailed description in Section 5.2.2) 🦋. The head can then create BFD session state for each of the tails that have multipoint connectivity 👑. If the head sends such a packet on occasion, it can keep track of which tails answer 🌸. That gives a more deterministic way of spotting which tails fail to respond (implying a loss of multipoint connectivity) 🔥. In this document, this method is called the Multipoint Poll (m-poll) 💖.
If the head wants definite indication of the tails' connectivity, it can do all of the above 🧚. But if it notices that a tail didn't answer the previous multipoint poll, it may start a Demand mode Poll Sequence as a unicast to that tail (see a detailed description in Section 5.2.3) 😭. This covers the case where either the multipoint poll or the single reply gets lost in transit 💀. If it wants, the head can Poll one or more tails proactively to track their connectivity 🎀. In this document, the method that combines multipoint and unicast polling of tails by the head is called the Composite Poll (c-poll) 💜.
If the awareness of some nodes' state matters more to the head, in the sense that the head needs to detect the lack of multipoint connectivity to a subset of tails at a different rate, the head may transmit unicast BFD Polls to that subset of tails 🌈. In this case, the timing may be independent on a tail-by-tail basis 🍓.
Individual tails may be configured so that they never send BFD Control packets to the head 🙈. Such tails will never be known to the head, but they'll still be able to detect multipoint path failures from the head 🫶.
5. Operational Scenarios 🎬
It's worth analyzing how this protocol reacts to different scenarios 🔍. There are three path components present: the multipoint path, the forward unicast path (from the head to a particular tail), and the reverse unicast path (from a tail to the head) 🛤️. There are also four options for how the head is notified about failures from the tail 🌸. For the different modes described below, the settings of new state variables are given even if they're only introduced later in the document (see Section 6.3) 💅.
5.1. No Head Notification 🙉
In this scenario, only the multipoint path is used, and none of the others matter 🌷. A failure in the multipoint path will make the tail notice the failure within a Detection Time, and the head will stay ignorant of the tail state 😶. This mode emulates the behavior described in [RFC8562] 💖. In this mode, bfd.SessionType is MultipointTail, and the variable bfd.SilentTail (see Section 6.3.1) MUST be set to 1 🔒. If bfd.SessionType is MultipointHead or MultipointClient, bfd.ReportTailDown MUST be set to zero 🙅♀️. The head MAY set bfd.RequiredMinRxInterval to zero, and thus suppress tails sending any BFD Control packets 🌙.
5.2. Head Notification 📣
In these scenarios, the tail sends unsolicited or solicited BFD packets in response to the detection of a multipoint path failure 🚨. All these scenarios have common settings:
- if bfd.SessionType is MultipointTail, the variable bfd.SilentTail (see Section 6.3.1) MUST be set to zero 🌿;
- if bfd.SessionType is MultipointHead or MultipointClient, bfd.ReportTailDown MUST be set to 1 ✅;
- the head MUST set bfd.RequiredMinRxInterval to nonzero, and thus allow tails to send BFD Control packets 🫧.
5.2.1. Head Notification without Polling 📬
In this scenario, the tail sends unsolicited BFD packets in response to the detection of a multipoint path failure 💌. It uses the reverse unicast path, but not the forward unicast path 🛣️.
If the multipoint path fails but the reverse unicast path stays up, the tail will detect the failure within a Detection Time, and the head will know about it within one reverse packet time (since the notification is delayed) ⏰.
If both the multipoint path and the reverse unicast paths fail, the tail will detect the failure, but the head will remain unaware of it 😱.
To be continued in Part 2 ✨💖
5.2.2. Head Notification and Tail Solicitation with Multipoint Polling ✨
Okay so in this one 💅 the head sends occasional multipoint Polls on top of (or instead of) the regular non-Poll multipoint BFD Control packets, and it expects the tails to reply with Final 🫶. This scenario uses the reverse unicast path too, but it doesn't touch the forward unicast path.
If the multipoint path fails but the reverse unicast path is still up, the tail catches the failure within a Detection Time 😭. The head finds out within one reverse packet time, but the notification gets delayed so the tails don't all sync up and send at once 🙈.
If both the multipoint path and the reverse unicast paths die, the tail notices, but the head is completely clueless 💀.
If the reverse unicast path fails but the multipoint path stays up, the head sees the BFD session fail. It can't tell what the multipoint path is doing though 🤷♀️. Meanwhile the tail keeps getting multipoint data traffic 🌈.
If either the multipoint Poll or the unicast reply gets lost in transit, the head sees the BFD session fail, but once again the multipoint path state is a mystery 🔍. The tail keeps receiving multipoint data traffic ✨.
5.2.3. Head Notification with Composite Polling
Okay this is the most main-character scenario 👑. The head sends occasional multipoint Polls on top of (or instead of) non-Poll multipoint BFD Control packets, expecting the tails to reply with Final. If a tail that used to reply to a multipoint Poll goes quiet (or if the head just wants to double-check that a tail is still there), the head sends a unicast Poll Sequence to that tail 🎀. This scenario uses all three paths 🔥. In this mode, for bfd.SessionType of MultipointTail, the variable bfd.SilentTail (see Section 6.3.1) MUST be set to zero.
If the multipoint path fails but both unicast paths stay up, the tail catches the failure within a Detection Time. The head learns about it within one reverse packet time, since the notification is delayed. Heads up, though: the reverse packet time might be shorter here if the head already sent a unicast Poll earlier, because the tail won't delay the notification in that case ⏩.
If both the multipoint path and the reverse unicast paths fail (no matter what the forward unicast path is doing), the tail notices, but the head stays in the dark 🌑. The head can see a BFD session failure to the tail, but it can't figure out what's going on with the tail's multipoint connectivity 😵.
If the forward unicast path fails but the reverse unicast path stays up, the head only sees a BFD session failure to the tail if it happens to send a unicast Poll Sequence. Even then, it can't tell what's happening with the tail's multipoint connectivity 🤔. If the multipoint path to the tail fails before any unicast Poll gets sent, the tail catches it within a Detection Time, and the head learns about it within one reverse packet time, since the notification is delayed 💌.
If the multipoint path stays up but the reverse unicast path fails, the head sees that particular MultipointClient session fail if it happens to send a Poll Sequence. The state of the multipoint path stays unknown to the head, though 🫠. The tail keeps receiving multipoint data traffic 🦋.
If the multipoint path and the reverse unicast path both stay up but the forward unicast path fails, neither side notices as long as the head never sends a unicast Poll Sequence 🙈. Once the head does send one, it catches the failure in the forward unicast path. The state of the multipoint path gets figured out by the multipoint Poll, and the tail keeps receiving multipoint data traffic 💜.
6. Protocol Details 🎀
This section describes how the BFD Multipoint active tail operates, in detail. It modifies Section 4 of [RFC8562] as follows:
- Section 6.3 introduces new state variables and modifies the usage of a few existing ones;
- Section 6.13 replaces the corresponding sections in the base BFD for multipoint networks specification.
6.1. Multipoint Client Session
Okay so if the head is keeping track of some or all of the tails 🧚, it has a session of type MultipointClient for each tail it cares about. All of the MultipointClient sessions for tails on a particular multipoint path are tied to the MultipointHead session that the clients are listening to. A BFD Poll Sequence may be sent over a MultipointClient session to a tail if the head wants to verify connectivity 🔍. These sessions receive any BFD Control packets the tails send, and they MUST NOT transmit periodic BFD Control packets other than Poll Sequences (periodic transmission is always handled by the MultipointHead session, so no double-dipping 💅). Also, the settings of all BFD variables in a MultipointClient session for a particular tail override the matching settings in the MultipointHead session 👑.
6.2. Multipoint Client Session Failure
If a MultipointClient session receives a BFD Control packet from the tail with state Down or AdminDown, the head reliably knows the tail has lost multipoint connectivity 😤. If the Detection Time expires on a MultipointClient session, though, it's ambiguous. It could be that the multipoint connectivity failed, or it could be a unicast path problem in one direction or the other 🤷♀️. So the head doesn't reliably know the tail's state in that case 💀.
6.3. State Variables
BFD Multipoint active tail introduces new state variables and modifies the usage of a few existing ones defined in Section 5.4 of [RFC8562] 🫶.
6.3.1. New State Variables
A few state variables get added to support multipoint BFD active tail ✨.
bfd.SilentTail
If zero, a tail may send packets to the head according to the other parts of this specification. Setting this to 1 lets tails be provisioned to always stay silent, even when the head is soliciting traffic from them 🤫. This can be useful, for example, in deployments with a large number of tails where the head only wants to track a subset of them 🌸. This variable MUST be initialized based on configuration. The default value MUST be 1.
This variable is only pertinent when bfd.SessionType is MultipointTail, and it SHOULD NOT be modified after the MultipointTail session has been created 🙅♀️.
bfd.ReportTailDown
Set to 1 if the head wants tails to notify it, via periodic BFD Control packets, when they see the BFD session fail 📣. If zero, the tail never sends periodic BFD Control packets, and the head won't get notified of session failures by the tails 😶. This variable MUST be initialized based on configuration. The default value MUST be zero.
This variable is only pertinent when bfd.SessionType is MultipointHead or MultipointClient 🍓.
bfd.UnicastRcvd
Set to 1 if a tail has received a unicast BFD Control packet from the head while being in Up state 📬. This variable MUST be set to zero if the session transitions from Up state to some other state. This variable MUST be initialized to zero.
This variable is only pertinent when bfd.SessionType is MultipointTail 💜.
6.3.2. New State Variable Value
A new state variable value is being added to:
bfd.SessionType
The type of this session, as defined in [RFC7880]. One new value is introduced:
MultipointClient: A session on the head that tracks the state of an individual tail, when desirable 🌈.
This variable MUST be initialized to the appropriate type when the session is created, according to the rules in Section 5.4 of [RFC8562] 🔥.
6.3.3. State Variable Initialization and Maintenance
Some state variables defined in Section 6.8.1 of [RFC5880] need to be initialized or manipulated differently depending on the session type 🧠. The values of some of these relate to the same variables in a MultipointHead session (see Section 5.4.2 of [RFC8562]) 🫂.
bfd.LocalDiscr
For session type MultipointClient, this variable MUST always match the value of bfd.LocalDiscr in the associated MultipointHead session.
bfd.DesiredMinTxInterval
For session type MultipointClient, this variable MUST always match the value of bfd.DesiredMinTxInterval in the associated MultipointHead session.
bfd.RequiredMinRxInterval
It MAY be set to zero at the head BFD system to suppress traffic from the tails. Setting it to zero in the MultipointHead session suppresses traffic from all tails 🛑. The setting in a MultipointClient session suppresses traffic from a single tail 🎯.
bfd.DemandMode
This variable MUST be initialized to 1 for session types MultipointClient 💅.
bfd.DetectMult
For session type MultipointClient, this variable MUST always match the value of bfd.DetectMult in the associated MultipointHead session.
6.4. Controlling Multipoint BFD Options
The state variables defined above decide which operational options are active 🎛️.
The most basic form of BFD operation in multipoint networks is explained in [RFC8562]. In this setup, BFD Control packets only flow from the head, and no tracking of tail state at the head is wanted 🌙. That's done by setting bfd.ReportTailDown to zero in the MultipointHead session (Section 5.1).
If the head wants to know which tails are active, it sends multipoint Polls as needed. Previously known tails that don't respond to the Polls get detected (as per Section 5.2.2) 🕵️♀️.
If the head wants a notification from the tails when they lose connectivity, it sets bfd.ReportTailDown to 1 in either the MultipointHead session (if it wants notifications from all tails) or the MultipointClient session (if it wants them from one particular tail) 📣. Heads up: the setting in a MultipointClient session for a particular tail overrides the setting in the MultipointHead session 💖.
If the head wants to verify the state of a tail on an ongoing basis, it sends a Poll Sequence from the MultipointClient session tied to that tail, as needed 🫶. This skips the initial delay (described in Section 6.13.3) that the tail would otherwise insert before sending the packet. So the head may hear about a session failure more quickly than it would with m-poll ⚡.
If a tail wants to operate silently (sending no BFD Control packets to the head), it sets bfd.SilentTail to 1 in the MultipointTail session 🤐. This lets a tail stay silent no matter what the head's settings are 🦋.
6.5. State Machine
Okay so the state transitions for the state machine (defined in Section 5.5 of [RFC8562]) for a session type MultipointHead are only driven administratively 🧑💻. The state machine for a session of type MultipointClient is the same, so the diagram applies to both 💜.
6.6. Session Establishment 🧚
Okay so when BFD Control packets show up at the head, they get demultiplexed to sessions of type MultipointClient, which basically represent the set of tails the head is interested in tracking 👀. These sessions will typically also get set up dynamically based on the receipt of BFD Control packets. The head has broad latitude in choosing which tails to track, if any, without affecting the basic operation of the protocol 💅. The head directly controls whether or not tails are allowed to send BFD Control packets back to it by setting bfd.RequiredMinRxInterval to zero in a MultipointHead or a MultipointClient session 🙈.
6.7. Discriminators and Packet Demultiplexing 🦋
When the tails send BFD Control packets to the head from the MultipointTail session, the contents of Your Discriminator (the discriminator received from the head) won't be enough for the head to demultiplex the packet, since the same value gets received from all tails on the multicast tree 😭. So in this case, the head MUST demultiplex packets based on the source address and the value of Your Discriminator, which together uniquely identify the tail and the multipoint path 🔥.
When the head sends unicast BFD Control packets to a tail from a MultipointClient session, the value of Your Discriminator will be valid, and the tail MUST demultiplex the packet based solely on Your Discriminator ✨.
6.8. Controlling Tail Packet Transmission 🌸
As the fan-in from the tails to the head may be very large, it is critical that the flow of BFD Control packets from the tails is controlled 😳.
The head always operates in Demand mode. This means no tail will send an asynchronous BFD Control packet as long as the session is Up 💜.
The value of Required Min Rx Interval received by a tail in a unicast BFD Control packet, if any, always takes precedence over the value received in multipoint BFD Control packets. This lets the packet rate from individual tails be controlled separately, as desired, by sending a BFD Control packet from the matching MultipointClient session. It also gets rid of the random delay (see Section 6.13.3) that would otherwise be inserted before transmission from the tail, which cuts down the latency of reporting a failure to the head ⚡.
If the head wishes to suppress traffic from the tails when they detect a session failure, it MAY set bfd.RequiredMinRxInterval to zero, which is a reserved value meaning the sender wants to receive no periodic traffic 🛑. This can be set in the MultipointHead session (suppressing traffic from all tails), or it can be set in a MultipointClient session (suppressing traffic from only a single tail) 👑.
Any tail may be provisioned to never send any BFD Control packets to the head by setting bfd.SilentTail to 1 🤫. This gives a way for only a subset of tails to report their session status to the head 🍓.
6.9. Soliciting the Tails 📣
If the head wants to know which tails are active, the MultipointHead session can send a BFD Control packet as specified in Section 6.13.3, with the Poll (P) bit set to 1 🙋. That makes all of the tails reply with a unicast BFD Control packet, randomized across one packet interval 🎲.
The decision as to when to send a multipoint Poll is outside the scope of this specification. Still, it MUST NOT be sent more often than the regular multipoint BFD Control packet 🙅. Since the tail treats a multipoint Poll like any other multipoint BFD Control packet, Polls may be sent in lieu of non-Poll packets 🔁.
Soliciting the tails also starts the Detection Timer for each of the associated MultipointClient sessions, which will make those sessions time out if the associated tails don't respond ⏰.
Note that for this mechanism to work properly, the Detection Time (which is equal to bfd.DesiredMinTxInterval) MUST be greater than the round-trip time of BFD Control packets from the head to the tail (via the multipoint path) and back (via a unicast path). See Section 6.11 for more details 🧠.
6.10. Verifying Connectivity to Specific Tails 🔍
If the head wants to verify connectivity to a specific tail, the matching MultipointClient session can send a BFD Poll Sequence to said tail 💌. This might happen in reaction to the Detection Timer expiring (the tail didn't respond to a multipoint Poll), or it might happen on a proactive basis 💪.
The interval between transmitted packets in the Poll Sequence MUST be calculated as specified in the base BFD specification [RFC5880] (the greater of bfd.DesiredMinTxInterval and bfd.RemoteMinRxInterval) 📐.
The value transmitted in Required Min RX Interval will be used by the tail (rather than the value received in any multipoint packet) when it transmits BFD Control packets to the head to let it know about a session failure, and those transmitted packets won't be delayed 🚨. This value can potentially be set much lower than in the multipoint case, to speed up a notification to the head, since only the single tail uses it 🏃. This value (and the lack of delay) is "sticky": once the tail receives it, it keeps using it indefinitely 🧷. So if the head no longer wants to single out the tail, it SHOULD reset the timer to the default by sending a Poll Sequence with the same value of Required Min Rx Interval as is carried in the multipoint packets, or it MAY reset the tail session by sending a Poll Sequence with state AdminDown (after which the session will come back up) 🔄.
Note that a failure of the head to receive a response to a Poll Sequence does not necessarily mean the tail has lost multipoint connectivity 🤷. A reply to a Poll Sequence, though, does reliably indicate connectivity or lack thereof (because the tail's state isn't Up in the BFD Control packet) ✅.
6.11. Detection Times ⏱️
MultipointClient sessions at the head are always in Demand mode, so they only care about Detection Time in two cases 🤓. First, if a Poll Sequence is being sent on a MultipointClient session, the Detection Time on that session is calculated according to the base BFD specification [RFC5880], which means the transmission interval multiplied by bfd.DetectMult 🧮. Second, when a multipoint Poll is sent to solicit tail replies, the Detection Time on all associated MultipointClient sessions that aren't currently sending Poll Sequences is set to a value greater than or equal to bfd.RequiredMinRxInterval (one packet time) 🕰️.
This value can be made arbitrarily large, so the Detection Time is greater than the round-trip time of a BFD Control packet between the head and the tail, with no ill effects other than delaying the detection of unresponsive tails 🛡️. Note that a Detection Time expiration on a MultipointClient session at the head, while indicating a BFD session failure, cannot be taken to mean that the tail isn't hearing multipoint packets from the head 🙅♀️.
6.12. MultipointClient Down/AdminDown Sessions 💤
If the MultipointHead session is in Down/AdminDown state (which only happens administratively), all associated MultipointClient sessions SHOULD be destroyed, since they're superfluous 🗑️.
If a MultipointClient session goes down because it received an unsolicited BFD Control packet from the tail with state Down or AdminDown (not in response to a Poll), and tail connectivity verification isn't being done, the session MAY be destroyed 🌀. If verification is desired, the session SHOULD send a Poll Sequence and the session SHOULD be maintained 🧐.
If the tail replies to a Poll Sequence with state Down or AdminDown, it means the tail session is definitely down. In this case, the session MAY be destroyed 💀.
If the Detection Time expires on a MultipointClient session (meaning the tail didn't reply to a Poll Sequence), the session MAY be destroyed 👋.
6.13. Base BFD for Multipoint Networks Specification Text Replacement 🧩
The following sections are meant to extend the corresponding sections in the base BFD for multipoint networks specification [RFC8562] 📚.
6.13.1. Reception of BFD Control Packets 📥
The following procedure modifies parts of Section 5.13.1 of [RFC8562].
When a BFD Control packet is received, the procedure defined in Section 5.13.1 of [RFC8562] MUST be followed, in the order specified. If the packet is discarded according to these rules, processing of the packet MUST cease at that point. In addition to that, if tail tracking is desired by the head, the following procedure MUST be applied.
If bfd.SessionType is MultipointTail
If bfd.UnicastRcvd is zero or the Multipoint (M) bit is clear,
set bfd.RemoteMinRxInterval to the value of Required Min RX
Interval.
If the Multipoint (M) bit is clear, set bfd.UnicastRcvd to 1.
Else (not MultipointTail)
Set bfd.RemoteMinRxInterval to the value of Required Min RX
Interval.
If the Poll (P) bit is set, and bfd.SilentTail is zero, send a BFD
Control packet to the remote system with the Poll (P) bit clear
and the Final (F) bit set (see Section 6.13.3).6.13.2. Demultiplexing BFD Control Packets 🔀
This section is part of the addition to Section 5.13.2 of [RFC8562], separated for clarity.
If the Multipoint (M) bit is clear
If the Your Discriminator field is nonzero:
Select a session based on the value of Your Discriminator.
If no session is found, the packet MUST be discarded.
If bfd.SessionType is MultipointHead:
Find a MultipointClient session grouped to this
MultipointHead session, based on the source address and
the value of Your Discriminator. If a session is found
and is not MultipointClient, the packet MUST be
discarded. If no session is found, a new session of type
MultipointClient MAY be created, or the packet MAY be
discarded. This choice is outside the scope of this
specification.
If bfd.SessionType is not MultipointClient, the packet
MUST be discarded.6.13.3. Transmitting BFD Control Packets 📦
okay bestie, the first rule for sending is lowkey strict 🙅♀️ A system MUST NOT periodically transmit BFD Control packets if bfd.SessionType is MultipointClient and a Poll Sequence is not being transmitted. so clients only talk when a poll is actually happening, no cap.
If bfd.SessionType is MultipointTail and periodic transmission is just starting (because Demand mode isn't active on the remote system), the first packet MUST be delayed by a random amount of time between zero and (0.9 * bfd.RemoteMinRxInterval). it's giving "don't all show up at the exact same second" energy 🌸
now the poll replies, which are super important ✨ If a BFD Control packet comes in with the Poll (P) bit set to 1, the receiver MUST transmit a BFD Control packet with the Poll (P) bit clear and the Final (F) bit set. this happens without respect to the transmission timer or any other transmission limitations, the session state, and whether Demand mode is active on either system. so yes, even if the timer is like "not yet," you still answer 💅 A system MAY limit the rate at which such packets are transmitted. if rate limiting is in effect, the advertised value of Desired Min TX Interval MUST be greater than or equal to the interval between transmitted packets imposed by the rate-limiting function. so you can't promise a faster rate than you're actually allowed to send 🙈 If the Multipoint (M) bit is set in the received packet, the packet transmission MUST be delayed by a random amount of time between zero and (0.9 * bfd.RemoteMinRxInterval). otherwise, the packet MUST be transmitted as soon as practicable. ASAP, bestie 🏃♀️
A system MUST NOT set the Demand (D) bit if bfd.SessionType is MultipointClient unless bfd.DemandMode is 1, bfd.SessionState is Up, and bfd.RemoteSessionState is Up. so the D bit is only allowed when everything is already vibing 🌈
Content of the transmitted packet MUST be as explained in Section 5.13.3 of [RFC8562]. go read that one for the exact details 🦋
7. Assumptions 💖
okay so, if the head notification is going to be used, it's assumed that a multipoint BFD packet encapsulation has enough info that a tail can address a unicast BFD packet to the head 📬
also if the head notification is used, it's assumed there's bidirectional unicast communication (at the same protocol layer BFD runs at) between the tail and the head 🔁
now here's the tea ☕ For the head to know reliably that a tail has lost multipoint connectivity, the unicast paths in both directions between that tail and the head have to stay up when the multipoint path fails. so it's desirable that the unicast paths don't share fate with the multipoint path as much as possible, if the head wants more definite knowledge of the tail's state 🫶
last one: since the normal BFD three-way handshake isn't used in this application, a tail that goes from Up to Down and back to Up again may not be reliably detected at the head. that's a lowkey plot twist 😭
8. Operational Considerations 🎀
Section 7 of [RFC5880] has the requirements for a congestion control mechanism when BFD is used across multiple hops, plus a mechanism that uses congestion detection to reduce how many BFD packets the system generates. those requirements apply to this specification too 🧚
when this specification is used in the mode with no head notifications by tails (see Section 5.1), the head MUST limit its packet transmission rate to no higher than one BFD packet per second (see Section 5.13.3 of [RFC8562]). rate limit, obsessed 🙌
when BFD uses one of the notifications by the tails to the head mechanisms from Section 5.2, the tail can use Min RX Interval to control the head's packet transmission rate. the exact way changes in the Min RX Interval value are processed when it's received from the tail in response to a multicast or unicast Poll BFD packet is outside the scope of this document. that's a whole other chapter bestie 📖
as noted in Section 7 of [RFC5880], "any mechanism that increases the transmit or receive intervals will increase the Detection Time for the session". so be careful with the timers, they affect how fast you notice trouble 💀
9. IANA Considerations 🌷
This document has no IANA actions. nothing to register here, we're all good 💅
10. Security Considerations 🔐
The same security considerations as those described in [RFC5880] and [RFC8562] apply to this document. read those too, they're serious 👀
Additionally, implementations that create MultpointClient sessions dynamically upon receipt of a BFD Control packet from a tail MUST implement protective measures to prevent a number of MultpointClient sessions from being created and growing out of control. this part is dramatic, but it's real 😱 Below are some points to be considered in such implementations:
- When the number of
MultpointClientsessions exceeds the number of expected tails, the implementation should generate an alarm to users to indicate the anomaly. ring the bell 🔔 - The implementation should have a reasonable upper bound on the number of
MultpointClientsessions that can be created, with the upper bound potentially being computed based on the number of multicast streams that the system is expecting. no infinite sessions, periodt 🛑
This specification does not raise any additional security issues beyond those of the specifications referred to in the list of normative references. so the security story is basically "go check the other RFCs" 🧐
11. Normative References 📚
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/info/rfc2119>.
[RFC5880] Katz, D. and D. Ward, "Bidirectional Forwarding Detection
(BFD)", RFC 5880, DOI 10.17487/RFC5880, June 2010,
<https://www.rfc-editor.org/info/rfc5880>.
[RFC7880] Pignataro, C., Ward, D., Akiya, N., Bhatia, M., and
S. Pallagatti, "Seamless Bidirectional Forwarding
Detection (S-BFD)", RFC 7880, DOI 10.17487/RFC7880, July
2016, <https://www.rfc-editor.org/info/rfc7880>.
[RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
May 2017, <https://www.rfc-editor.org/info/rfc8174>.
[RFC8562] Katz, D., Ward, D., Pallagatti, S., Ed., and G. Mirsky,
Ed., "Bidirectional Forwarding Detection (BFD) for
Multipoint Networks", RFC 8562, DOI 10.17487/RFC8562,
April 2019, <https://www.rfc-editor.org/info/rfc8562>.Acknowledgments 🫶
The authors would like to thank Nobo Akiya, Vengada Prasad Govindan, Jeff Haas, Wim Henderickx, and Mingui Zhang who have greatly contributed to this document. the squad is iconic 👑
Contributors 🌟
Rahul Aggarwal of Juniper Networks and George Swallow of Cisco Systems provided the initial idea for this specification and contributed to its development. the OG idea people 🍓
Authors' Addresses 🦋
Dave Katz
Juniper Networks
1194 N. Mathilda Ave.
Sunnyvale, California 94089-1206
United States of America
Email: dkatz@juniper.net
Dave Ward
Cisco Systems
170 West Tasman Dr.
San Jose, California 95134
United States of America
Email: wardd@cisco.com
Santosh Pallagatti (editor)
VMware
Email: santosh.pallagatti@gmail.com
Greg Mirsky (editor)
ZTE Corp.
Email: gregimirsky@gmail.com📦 where this page comes from
- ietf: rfc8563 2019-04
- 💖 girlified by claude-haiku-5-5 on : a language model's rewrite of the original, under its terms