Survey Report on Protocol Independent Multicast - Sparse Mode (PIM-SM) Implementations and Deployments

Protocol Independent Multicast - Sparse Mode (PIM-SM) was first published as in 1997. This version was then obsoleted by in 1998. The protocol was classified as Experimental in both documents. The protocol specification was then rewritten in whole and advanced to Proposed Standard as in 2006. Considering its multiple independent implementations developed and sufficient successful operational experience gained, the PIM WG decided to advance the PIM-SM protocol to Internet Standard. The conducted survey and this document are part of the work.

defines the stages in the standardization process, the requirements for moving a document between stages, and the types of documents used during this process. Section 4.1.2 of states that: The requirement for at least two independent and interoperable implementations applies to all of the options and features of the specification. In cases in which one or more options or features have not been demonstrated in at least two interoperable implementations, the specification may advance to the Draft Standard level only if those options or features are removed. updates the IETF Standards Process defined in . Primarily, it reduces the Standards Process from three Standards Track maturity levels to two. The second maturity level is a combination of Draft Standard and Standard as specified in . Section 2.2 of states that: (1) There are at least two independent interoperating implementations with widespread deployment and successful operational experience.(2)...(3) There are no unused features in the specification that greatly increase implementation complexity. Optional features that do not meet the aforesaid criteria have been identified by the PIM Working Group and will be removed. This document provides supporting documentation to advance the IETF stream's Protocol Independent Multicast - Sparse Mode (PIM-SM) protocol from Proposed Standard to Internet Standard.

A questionnaire was issued by the PIM WG co-chairs and announced widely to the vendors and operational community to obtain information on PIM-SM implementations and deployments. The survey concluded on 22 Oct 2012. The responses remain confidential and only combined results are published here, while responders chose whether to keep their affiliations confidential. The raw questionnaire is shown in , and a compilation of the responses is included in the following section.

Nine operators responded to the survey. They are SWITCH, National Research Council Canada, South Dakota School of Mines and Technology, Motorola Solutions, and five anonymous operators.

Since 1998, PIM-SM has been deployed for a wide variety of applications: Campus, Enterprise, Research and WAN networks, Broadband ISP, and Digital TV. There are five deployments based on implementations and two on implementations. PIM-SM for IPv6 has been deployed by three operators. Out of the nine operators, six have deployed PIM-SM implementations from multiple vendors. Operators reported minor interoperability issues and these were addressed by the vendors. There was no major interoperability concern reported by the operators.

Except for one deployment of PIM-SM with Multicast Extensions to OSPF (MOSPF), all other operators have deployed PIM-SM exclusively. No operators acknowledged deployments of either (*,*,RP) or PIM Multicast Border Route (PMBR) for interconnection between PIM-SM and other multicast domains.

The number of PIM-SM RPs deployed by operators ranges from a few (e.g., sixteen) to a massively scaled number (four hundred). Both static configuration and Bootstrap Router (BSR) have been deployed as RP discovery mechanisms. Anycast-RP has been deployed for RP redundancy. Two operators have deployed Anycast-RP using the Multicast Source Discovery Protocol (MSDP) . Three operators have deployed Anycast-RP using both MSDP and PIM for different scenarios. The best common practice seems to be to use static-RP configuration with Anycast-RP for redundancy.

Eight vendors reported PIM-SM implementations. They are XORP, Huawei Technologies, Cisco Systems, Motorola Solutions, Juniper Networks, and three other anonymous vendors.

Four vendors reported PIM-SM implementations based on and two reported PIM-SM implementations based on . Two other reported implementations are hybrids. Minor interoperability issues have been addressed by the vendors over the years and no concerns were reported by any vendor.

Most vendors have not implemented (*,*,RP) state as specified in either due to lack of deployment requirements or due to security concerns. Similarly, most vendors have also not implemented PMBR due to lack of deployment requirements or because it was considered too complex and non-scalable. Only one vendor, XORP, reported (*,*,RP) and PMBR implementation and they were implemented just because these were part of the specification.

Most vendors have implemented all of the following from the specification: Source-Specific Multicast (SSM) Join suppression Explicit tracking Register mechanism Shortest Path Tree (SPT) switchover at last-hop router Assert mechanism Hashing of group to RP mappings Some vendors do not implement explicit tracking and SSM.

PIM-SM has been widely implemented and deployed for different applications. The protocol is sufficiently well specified in resulting in interoperable implementation deployed by operators. There are no deployments and only one known implementation of (*,*,RP) and PMBR as specified in . Hence, it is necessary to remove these features from the specification as required by and .

The PIM WG is aware of at least three (and believes there are more) PIM-SM implementations that support the use of IPsec to protect PIM messages. For at least one of them, IPsec is not part of the PIM implementation itself -- one just configures IPsec with Security Policy Databases (SPDs) where interface, the ALL_PIM_ROUTERS multicast address, etc., can be used as selectors, according to .

The authors would like to thank Tim Chown and Bill Atwood, who helped to collect and anonymize the responses as the neutral third party. Special thanks are also given to Alexander Gall, William F. Maton Sotomayor, Steve Bauer, Sonum Mathur, Pavlin Radoslavov, Shuxue Fan, Sameer Gulrajani, and to the anonymous responders.

This section provides copies of the questionnaires exactly as distributed to operators and implementors.