| Mboned | M. Abrahamsson |
| Internet-Draft | T-Systems |
| Intended status: Best Current Practice | T. Chown |
| Expires: January 3, 2019 | Jisc |
| L. Giuliano | |
| Juniper Networks, Inc. | |
| T. Eckert | |
| Huawei | |
| July 2, 2018 |
Deprecating ASM for Interdomain Multicast
draft-acg-mboned-deprecate-interdomain-asm-02
This document recommends deprecation of the use of Any-Source Multicast (ASM) for interdomain multicast. It recommends the use of Source-Specific Multicast (SSM) for interdomain multicast applications, and that hosts and routers that are expected to handle such applications fully support SSM. The recommendations in this document do not preclude the continued use of ASM within a single organisation or domain, and are especially easy to adopt when already using the preferred ASM protocol options there (PIM-SM).
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 "Key words for use in RFCs to Indicate Requirement Levels" [RFC2119].
This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."
This Internet-Draft will expire on January 3, 2019.
Copyright (c) 2018 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.
IP Multicast has been deployed in various forms, within private networks, the wider Internet, and federated networks such as national or regional research networks. While a number of service models have been published, and in many cases revised over time, there has been no strong recommendation made by the IETF on the appropriateness of those models to certain scenarios, even though vendors and federations have often made such recommendations.
This document addresses this gap by making a BCP-level recommendation to deprecate the use of ASM for interdomain multicast, leaving SSM as the remaining interdomain mode of multicast. This recommendation thus also implicitly states that all hosts and routers that are expected to support interdomain multicast applications fully support SSM.
This document does not make any statement on the use of ASM within a single domain or organisation, and therefore does not preclude its use. Indeed, there are application contexts for which ASM is currently still widely considered well-suited within a single domain.
The main issue in most cases with moving to SSM is application support. Many applications will first get used intradomain but only later interdomain. Therefore, this document recommends making applications support SSM, even when they are initially meant to be just used intradomain. As explained below, SSM applications are readily compatible with existing intradomain ASM deployments that follow the current IETF standard protocol recommendations.
The general IP multicast service model [RFC1112] is that sender(s) send to a multicast group address, receivers express an interest in traffic sent to a given multicast group address, and that routers use multicast routing protocols to determine how to deliver traffic from the sender(s) to the receivers.
Two high-level flavours of this service model have evolved over time. In Any-Source Multicast (ASM), any number of sources may transmit multicast packets, and those sources may come and go over the course of a multicast session without being known a priori. In ASM, receivers express interest only in a given multicast group address, and the multicast routing protocol facilitates source discovery at the network layer. ASM is simply the name given to the 1989 RFC1112 IP Multicast model when in around 2000 the idea for the alternative SSM model was taking shape: In Source-Specific Multicast (SSM) the specific source(s) that may send traffic to the group are known in advance by the receivers, or may be determined during a session, typically through an out-of-band protocol sitting above the network layer. Thus in SSM, receivers express interest in both a multicast group address and specific associated source address(es).
IANA has reserved specific ranges of IPv4 and IPv6 address space for multicast addressing. Guidelines for IPv4 multicast address assignments can be found in [RFC5771], while guidelines for IPv6 multicast address assignments can be found in [RFC2375] and [RFC3307]. The IPv6 multicast address format is described in [RFC4291].
The most commonly deployed ASM routing protocol is Protocol Independent Multicast - Sparse Mode, or PIM-SM, as detailed in [RFC7761]. PIM-SM, as the name suggests, was designed to be used in scenarios where the subnets with receivers are sparsely distributed throughout the network. Because it does not know sender addresses in advance, PIM-SM uses the concept of a Rendezvous Point (RP) to 'marry up' senders and receivers, and all routers in a PIM-SM domain are configured to use specific RP(s), either explicitly or through dynamic RP discovery protocols.
To enable PIM-SM to work between multiple domains, i.e., to allow an RP in one domain to learn the existence of a source in another domain, an inter-RP signalling protocol known as Multicast Source Discovery Protocol (MSDP) [RFC3618] is used. Deployment scenarios for MSDP are given in [RFC4611]. MSDP has remained an Experimental protocol since its publication in 2003. One core reason for this is the need to flood information about all active sources for all active applications to the RPs in all the domains in an MSDP peering mesh - even if there is no receiver for a given application in a domain. This is the key scalability and security issue with MSDP and also the reason why it was not extended to support IPv6.
To this day, there is no IETF Proposed Standard level interdomain solution for IPv4 ASM multicast because MSDP was the "best" component for the interdomain discovery problem, and it stayed Experimental. Other protocol options where investigated at the same time but did achieve at most achieve IETF informational status and are now historic (e.g: [RFC3913]).
Due to the availability of more bits in an IPv6 address than in IPv4, an IPv6-specific mechanism was able to be designed in support of interdomain ASM with PIM-SM. Embedded-RP [RFC3956] allows routers supporting the protocol to determine the RP for the group without any prior configuration or discovery protocols, simply by observing the unicast RP address that is embedded (included) in the IPv6 multicast group address. Embedded-RP allows PIM-SM operation across any IPv6 network (intradomain but especially interdomain) in which there is an end-to-end path of routers supporting the mechanism.
SSM is detailed in [RFC4607]. It is in effect a subset of PIM-SM where no RP is used. Note that there is no separate document for PIM-SSM, it just leverages a subset of [RFC7761].
PIM-SSM expects that sender source address(es) are known in advance by receivers; i.e., a given source's IP address is known (by some out of band mechanism), and thus the receiver's router can send a PIM JOIN directly towards the sender, without needing to use an RP.
IPv4 addresses in the 232/8 (232.0.0.0 to 232.255.255.255) range are designated as source-specific multicast (SSM) destination addresses and are reserved for use by source-specific applications and protocols. See [RFC4607]. For IPv6, the address prefix FF3x::/32 is reserved for source-specific multicast use.
In enterprise and campus scenarios, ASM in the form of PIM-SM is likely the most commonly deployed multicast protocol and has generally replaced PIM-DM [RFC3973], which is an IETF Experimental category RFC, while PIM-SM is full Internet Standard. The configuration and management of an RP (even with RP redundancy) within a single domain is well understood operational practice. However, if interworking with external PIM domains is needed in IPv4 multicast deployments, interdomain MSDP is required to exchange information about sources between domain RPs. The problems with this use of MSDP are as explained above. They are the problems that make MSDP an Experimental protocol, and that make it (in these deployments) a complex and fragile protocol to administer and troubleshoot (flooding RPF rules, state attack protection, undesired source filtering, ...).
PIM-SM is a general purpose protocol that can handle all use cases. In particular, it was designed for cases such as videoconferencing where multiple sources may come and go during a multicast session. But for cases where a single, persistent source for a group is used, and receivers can be configured to know of that source, PIM-SM has unnecessary complexity. In these applications it is typically only necessary to extend the configuration or signaling for the IP multicast group to be used with the additional information on the IP multicast source to be used. There are also various techniques to use a single logical "anycast" source address supported by two or more redundant senders to give additional reliability for SSM, and to offer simpler deployment by using that address as a "static"/"well-known" address.
As explained above, MSDP was not taken forward to IPv6. Instead, the proposed interdomain ASM solution for PIM-SM with IPv6 is Embedded-RP, which allows the RP address for a multicast group to be embedded in the group address, making RP discovery automatic, if all routers on the path between a receiver and a sender support the protocol. Embedded-RP can support lightweight ad-hoc deployments. However, it relies on a single RP for an entire group that could only be made resilient within one domain. While this approach solves the MSDP issues, it does not solve the problem of unauthorised sources sending traffic to ASM multicast groups; this security issues is one of biggest problem of interdomain multicast. Embedded-RP was run successfully between European and US academic networks during the 6NET project in 2004/05. Its usage generally remains constrained to academic networks.
As stated in RFC 4607, SSM is particularly well-suited to dissemination-style applications with one or more senders whose identities are known (by some mechanism) before the application starts running - or applications that have some existing signaling indicating multicast groups, where the additional source address can easily be added - for example electronic programming guide signalling in IPTV applications. PIM-SSM is therefore very well-suited to applications such as classic linear broadcast TV over IP.
SSM requires applications, host operating systems and their subnet routers using it to support the new(er) IGMPv3 [RFC3376] and MLDv2 [RFC3810] protocols. While delayed delivery of support in some OSes has meant that adoption of SSM has been slower than might have been expected, or hoped, and was a historical reason to use ASM rather than SSM, support for IGMPv3 and MLDv2 has become widespread in common OSes for several years (Windows, MacOS, Linux/Android).
A significant benefit of SSM is its reduced complexity through eliminating the network-based source discovery required in ASM. This means there are no RPs, shared trees, Shortest Path Tree (SPT) switchovers, PIM registers, MSDP, or data-driven state creation elements to support, or any requirement to run dynamic RP discovery and redundancy signaling protocols such as BSR. SSM is really just a small subset of PIM-SM, alongside the integration with IGMPv3 / MLDv2 where the source address signaled from the receiver is immediately used to create (S,G) state. Eliminating network-based source discovery for interdomain multicast means the vast majority of the complexity issues go away.
This reduced complexity makes SSM radically simpler to manage, troubleshoot and operate, particularly for network backbone operators, and this is the main operator motivation for the recommendation to deprecate the use of ASM in interdomain scenarios. Note that SSM operation is all standardised in PIM-SM (RFC7761). There is no separate specification for PIM-SSM.
RFC 4607 details many benefits of SSM, including:
Further discussion can also be found in [RFC3569].
SSM is considered more secure in that it supports access control, i.e. you only get packets from the sources you explicitly ask for, as opposed to ASM where anyone can decide to send traffic to a PIM-SM group address. This topic is expanded upon in [RFC4609].
This document recommends that the use of ASM is deprecated for interdomain multicast, and thus implicitly that hosts and routers that are expected to support such interdomain applications fully support SSM and its associated protocols. Best current practices for deploying interdomain multicast using SSM are documented in [RFC8313].
The recommendation applies to the use of ASM between domains where either MSDP (IPv4) or Embedded-RP (IPv6) is used for sharing knowledge of remote sources (MSDP) or RPs (Embedded-RP).
This document also recommends against the interdomain use of PIM-SM with a (potentially redundant) RP, where multicast tunnels are used between domains.
An interdomain use of ASM multicast in the context of this document is primarily one where PIM-SM for ASM, e.g., with RPs/MSDP/Embedded-RP, is run on routers operated by two or more separate operational entities (domains, organisations).
The more inclusive interpretation of this recommendation is that it also extends to the case where PIM may only be operated in a single operator domain, but where user hosts or non-PIM network edge devices are under different operator control. A typical example of this case is an SP providing IPTV (single operator domain for PIM) to subscribers operating an IGMP proxy home gateway and IGMPv3/MLDv2 hosts (computer, tablets, set-top boxes).
While MSDP is an Experimental category IETF standard, this document does not propose making MSDP Historic, given its use may be desirable for intradomain multicast use cases (e.g., RP redundancy intradomain). This may change in future documents should a successor to MSDP for intradomain RP redundancy ([RFC4610]) be defined to add better support for some currently missing operational requirements.
This document recommends that all host and router platforms supporting multicast, and any security appliances that may handle multicast traffic, support IGMPv3 [RFC3376] and MLDv2 [RFC3810] (based on the version IP they intend to support). The updated IPv6 Node Requirements RFC [I-D.ietf-6man-rfc6434-bis] states that MLDv2 support is a MUST in all implementations. Such support is already widespread in common host and router platforms.
Further guidance on IGMPv3 and MLDv2 is given in [RFC4604].
It is sometimes desirable to limit the propagation of multicast messages in a layer 2 network, typically through a layer 2 switch device. In such cases multicast snooping can be used, by which the switch device observes the IGMP/MLD traffic passing through it, and then attempts to make intelligent decisions about on which physical ports it should forward multicast. Typically, ports that have not expressed an interest in receiving multicast for a given group would not have traffic for that group forwarded through them. Such snooping capability should therefore support IGMPv3 and MLDv2. There is further discussion in [RFC4541].
There is a wide range of applications today that only support ASM (mostly for historic reasons), whether as software packages, or code embedded in devices such as set-top boxes.
The recommendation to use SSM for interdomain multicast means that applications should use SSM, and operate correctly in an SSM environment, triggering IGMPv3/MLDv2 messages to signal use of SSM.
It is often thought that ASM is required for multicast applications where there are multiple sources. However, RFC 4607 also describes how SSM can be used instead of PIM-SM for multi-party applications:
Given all common OSes support SSM, it is then down to the programming language and APIs used as to whether the necessary SSM APIs are available. SSM support became first ubiquitous for C/C++/Python, and key exceptions today include websockets used in web-browser based applications (see e.g.: https://github.com/nodejs/node/pull/15735/files introducing SSM support there in 2017).
Some useful considerations for multicast applications can still be found in the relatively old [RFC3170].
If feasible, it is recommended to make applications use SSM, even if they are initially only meant to be used in intradomain environments supporting ASM. Because PIM-SSM is a subset of PIM-SM, it should be possible to readily make existing intradomain PIM-SM networks compatible with SSM application receivers, therefore allowing continued use of an existing ASM PIM-SM deployment in a network with no or very little changes. SSM's benefits of simplified address management and significantly reduced operational complexity apply equally to intradomain use.
However, for some applications it may be prohibitively difficult to add support for signaling of source IP addresses into the application.
Currently, there is no good document summarising best current practices to convert ASM applications into SSM applications, or how to most easily support SSM in greenfield application designs. This would be useful guidance for the IETF to work on.
In the case of existing ASM applications that cannot readily be ported to SSM, it may be possible to use some form of protocol mapping, i.e., to have a mechanism to translate a (*,G) join or leave to a (S,G) join or leave, for a specific source, S. The general challenge in performing such mapping is determining where the configured source address, S, comes from.
There are existing vendor-specific mechanisms deployed that achieve this function, but none are documented in IETF documents. This appears to be a useful area for the IETF to work on, but it should be noted that any such effort would only be an interim transition mechanism, and such mappings do not remove the requirement for applications to be allocated ASM group addresses for the communications.
The reason why these mechanisms should not be considered a long-term solution is because they introduce network operator management work, and need some form of address management, both of which are not required in SSM.
A key benefit of SSM is that a multicast application does not need to be allocated a specific multicast group by the network, rather as SSM is inherently source-specific, it can use any group address, G, in the reserved range of IPv4 or IPv6 SSM addresses for its own source address, S.
In principle, if interdomain ASM is deprecated, backbone operators could begin filtering the ranges of group addresses used by ASM. In practice, this is not recommended given there will be a transition period from ASM to SSM, where some form of ASM-SSM mappings may be used, and filtering may preclude such operations.
The use of ASM within a single multicast domain, such as a campus or enterprise, with an RP for the site, is still relatively common today. There are even global enterprise networks that have successfully been using PIM-SM for many years. The operators of such networks most often use Anycast-RP [RFC4610] or MSDP for RP resilience, at the expense of the extra complexity in managing that configuration. These existing practices are unaffected by this document.
This document does not preclude continued use of ASM in the intradomain scenario. If an organisation, or AS, wishes to use multiple multicast domains within its own network border, that is a choice for that organisation to make, and it may then use MSDP or Embedded-RP internally within its own network.
Traffic over non-controlled networks, which most interdomain paths are, must support congestion control. This is achievable with rate adaptation, layered codecs, circuit breakers and/or other appropriate mechanisms. See [RFC8085].
This document adds no new security considerations. It instead removes security issues incurred by interdomain ASM with PIM-SM/MSDP: infrastructure control plane attacks and application and bandwidth/congestion attacks from unauthorised sources sending to ASM multicast groups. RFC 4609 describes the additional security benefits of using SSM instead of ASM.
This document makes no request of IANA.
Note to RFC Editor: this section may be removed upon publication as an RFC.
The authors would like to thank members of the IETF mboned WG for discussions on the content of this document, with specific thanks to the following people for their contributions to the document: Hitoshi Asaeda, Dale Carder, Jake Holland, Albert Manfredi, Mike McBride, Per Nihlen, Greg Shepherd, James Stevens, Stig Venaas, Nils Warnke, and Sandy Zhang.