AVT A. Begen Internet-Draft D. Wing Intended status: Standards Track Cisco Expires: February 13, 2011 T. VanCaenegem Alcatel-Lucent August 12, 2010 Token-Based Port Mapping Between Unicast and Multicast RTP Sessions draft-begen-avt-token-for-portmapping-01 Abstract This document presents an alternative port mapping solution that allows RTP receivers to choose their own ports for an auxiliary unicast session in RTP applications using both unicast and multicast services (almost) without the need for retrieving pre-authorization. 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). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at http://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 February 13, 2011. Copyright Notice Copyright (c) 2010 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 (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 Begen, et al. Expires February 13, 2011 [Page 1] Internet-Draft Token-Based Port Mapping August 2010 described in the Simplified BSD License. Table of Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. Requirements Notation . . . . . . . . . . . . . . . . . . . . 4 3. Token-Based Port Mapping . . . . . . . . . . . . . . . . . . . 5 3.1. Token Request and Retrieval . . . . . . . . . . . . . . . 5 3.2. Unicast Session Establishment . . . . . . . . . . . . . . 5 4. The portmapping-req Attribute . . . . . . . . . . . . . . . . 9 5. Message Formats . . . . . . . . . . . . . . . . . . . . . . . 10 6. Procedures for Token Construction . . . . . . . . . . . . . . 11 7. Validating Tokens . . . . . . . . . . . . . . . . . . . . . . 12 8. SDP Example . . . . . . . . . . . . . . . . . . . . . . . . . 13 9. Address Pooling NATs . . . . . . . . . . . . . . . . . . . . . 15 10. Security Considerations . . . . . . . . . . . . . . . . . . . 16 11. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 17 11.1. Registration of SDP Attributes . . . . . . . . . . . . . . 17 12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 18 13. References . . . . . . . . . . . . . . . . . . . . . . . . . . 19 13.1. Normative References . . . . . . . . . . . . . . . . . . . 19 13.2. Informative References . . . . . . . . . . . . . . . . . . 19 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 21 Begen, et al. Expires February 13, 2011 [Page 2] Internet-Draft Token-Based Port Mapping August 2010 1. Introduction [I-D.ietf-avt-ports-for-ucast-mcast-rtp] provides several scenarios for RTP applications that use one or more unicast and multicast RTP sessions together. These applications require a Port Mapping solution that allows receivers to choose their desired UDP ports for RTP and RTCP in unicast session(s). There is an inherent delay in learning the public port mapping and signaling it with the Offer/ Answer Model [RFC3264]. Thus, the receiver might wish to convey its port number(s) through a different mechanism. [I-D.ietf-avt-ports-for-ucast-mcast-rtp] offers a Cookie-based solution. This memo presents a more lightweight solution, which we call the Token solution. Following the same convention with [I-D.ietf-avt-ports-for-ucast-mcast-rtp], we will refer to the RTP endpoints that serve other RTP endpoints over a unicast session as the Servers, and the receiving RTP endpoints as Clients. Begen, et al. Expires February 13, 2011 [Page 3] Internet-Draft Token-Based Port Mapping August 2010 2. Requirements Notation 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 [RFC2119]. Begen, et al. Expires February 13, 2011 [Page 4] Internet-Draft Token-Based Port Mapping August 2010 3. Token-Based Port Mapping Token-based Port Mapping consists of two steps: Token request and retrieval, and unicast session establishment. These are described in the following sections. 3.1. Token Request and Retrieval The first step is required to be completed only once. Once a Token is retrieved from a particular server, it may be used for all the unicast sessions the client will be running with this particular server. By default, Tokens are server specific. However, the client can use the same Token to communicate with different servers if these servers are provided with the same key used to generate the Token. The Token may become invalid if client's public IP address changes or when the server expires the token. In those cases, the client has to request a new Token. The Token is essentially an opaque encapsulation that conveys client's IP address information (as seen by the server) using a reversible transform only known to the server. When a request is received, the server creates a Token for this particular client, and sends it back to the client. Later, when the client wants to establish a unicast session, the Token will be validated by the server, making sure that the IP address information matches. This is effective against DoS attacks, i.e., an attacker cannot simply spoof another client's IP address and start possibly a high-bitrate unicast transmission [I-D.ietf-avt-rapid-acquisition-for-rtp] towards random clients. 3.2. Unicast Session Establishment We illustrate the second step on the same example presented in [I-D.ietf-avt-ports-for-ucast-mcast-rtp]. Consider an SSM distribution network where a distribution source multicasts RTP packets to a large number of clients, and one or more retransmission servers function as feedback targets to collect unicast RTCP feedback from these clients [RFC5760]. The retransmission servers also join the primary multicast session to receive the multicast packets and cache them for a certain time period. When a client detects missing packets in the primary multicast session, it requests a retransmission from one of the retransmission servers by using an RTCP NACK message [RFC4585]. The retransmission server pulls the requested packet(s) out of the cache and retransmits them to the requesting client [RFC4588]. The pertaining RTP and RTCP flows are sketched in Figure 1. Between Begen, et al. Expires February 13, 2011 [Page 5] Internet-Draft Token-Based Port Mapping August 2010 the client and server, there may be one or more NAT devices [RFC4787]. -------------- --- ---------- | |-------------------------------| |-->|P1 | | |-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-| |.->|P2 | | | | | | | | Distribution | ---------------- | | | | | Source | | | | | | | | |---->|P1 | | | | | | |.-.->|P2 | | | | | | | | | | | | | -------------- | P3|<.=.=.=.| |=.=|*c0 | | P3|<~~~~~~~| |~~~|*c1 | PRIMARY MULTICAST | | | | | | RTP SESSION with | | | | | | UNICAST FEEDBACK | | | N | | | | Retransmission | | A | | Client | - - - - - - - - - - -| - - - - - - - -| - - - -| - |- -| - - - - -|- | Server | | T | | | AUXILIARY UNICAST | | | | | | RTP SESSION | | | | | | | P3|........| |..>|*c1 | | P3|=.=.=.=.| |=.>|*c1 | | P4|<.=.=.=.| |=.=|*c2 | | | | | | | ---------------- --- ---------- -------> Multicast RTP Flow .-.-.-.> Multicast RTCP Flow .=.=.=.> Unicast RTCP Reports ~~~~~~~> Unicast RTCP Feedback Messages .......> Unicast RTP Flow Figure 1: Example scenario showing an SSM distribution with support for retransmissions from a server In this figure, we have the following multicast and unicast ports: o Ports P1 and P2 denote the destination RTP and RTCP ports in the primary multicast session, respectively. The clients listen to these ports to receive the multicast RTP and RTCP packets. Ports P1 and P2 are defined declaratively. o Port P3 denotes the RTCP port on the feedback target running on the retransmission server to collect the RTCP feedback messages, Begen, et al. Expires February 13, 2011 [Page 6] Internet-Draft Token-Based Port Mapping August 2010 and RTCP receiver and extended reports from the clients in the primary multicast session. This is also the port that the retransmission server uses to send the RTP packets and RTCP sender reports in the unicast session. Port P3 is defined declaratively. o Port P4 denotes the RTCP port on the retransmission server used to collect the RTCP receiver and extended reports for the unicast session. Port P4 is defined declaratively and MUST be different from port P3. o Ports *c0, *c1 and *c2 are chosen by the client. *c0 denotes the port on the client used to send the RTCP reports for the primary multicast session. *c1 denotes the port on the client used to send the unicast RTCP feedback in the primary multicast session and to receive the RTP packets and RTCP sender reports in the unicast session. *c2 denotes the port on the client used to send the RTCP receiver and extended reports in the unicast session. Ports c0, c1 and c2 MAY be the same port or different ports. However, there are two advantages of using the same port for both c0 and c1: 1. Some NATs only keep bindings active when a packet goes from the inside to the outside of the NAT (See REQ-6 of Section 4.3 of [RFC4787]). Long RTP bursts may exceed that timeout. If c0=c1, the occasional RTCP receiver reports sent from port c0 will ensure the NAT does not time out the public port associated with the incoming RTP burst to port c1. 2. Having c0=c1 conserves NAT port bindings. Thus, it is strongly RECOMMENDED that c0=c1. Once the server receives the RTCP NACK, the server sends the RTP burst to the IP address and UDP port the RTCP NACK came from. In addition to the ports, we use the following notation: o DS: IP address of the distribution source o G: Destination multicast address o S: IP address of the retransmission server o C: IP address of the client o C': Public IP address of the client (as seen by the server) We assume that the information declaratively defined is available as part of the session description information and is provided to the Begen, et al. Expires February 13, 2011 [Page 7] Internet-Draft Token-Based Port Mapping August 2010 clients. The Session Description Protocol (SDP) [RFC4566] and other session description methods can be used for this purpose. The following steps summarize the Token-based solution: 1. The client ascertains server address (S) and port numbers (P3 and P4) from the session description. 2. The client determines its port numbers (*c0, *c1 and *c2). 3. If the client does not have a valid Token for this particular server: A. The client first sends a message to the server via a new RTCP message, called PortMappingRequest. This message can be sent from any port on the client side. The server learns client's public IP address (C') from the received message. NOTE: The client can send this message anytime it wants (e.g., during initialization), and does not normally ever need to re-send this message (See Section 7). B. The server generates an opaque encapsulation (called Token) that conveys client's IP address information using a reversible transform only known to the server. See Section 6. C. The server sends the Token back to the client using a new RTCP message, called PortMappingResponse. This message MUST be sent from the port at which the server received the request. 4. The client includes the Token when necessary in the subsequent messages sent to the server. Note that the unicast session is only established after the server has received a feedback message (along with a valid Token) from the client for which it needs to react by sending unicast data. Until a unicast session is established, neither the server nor the client needs to send RTCP reports for the unicast session. 5. Normal flows ensue as shown in Figure 1. If the client uses the same port for both c0 and c1, the RTCP receiver and extended reports sent for the primary multicast session keep the P3->c1 binding alive. If the client uses different ports for c0 and c1, an explicit keep-alive message [I-D.ietf-avt-app-rtp-keepalive] may be needed to keep the P3->c1 binding alive during the lifetime of the unicast session, if that unicast session's lifetime exceeds the NAT's mapping refresh time. Begen, et al. Expires February 13, 2011 [Page 8] Internet-Draft Token-Based Port Mapping August 2010 4. The portmapping-req Attribute This new SDP attribute is used declaratively to indicate the port for obtaining a Token. Its presence also indicates that a Token MUST be included in the feedback messages sent to the server. The formal description of the 'portmapping-req' attribute is defined by the following ABNF [RFC5234] syntax: portmapping-req-attribute = "a=portmapping-req:" port CRLF Here, the 'port' token is defined as specified in Section 9 of [RFC4566]. The 'portmapping-req' attribute MAY be used as a media-level attribute; it MUST NOT be used as a session-level attribute. Begen, et al. Expires February 13, 2011 [Page 9] Internet-Draft Token-Based Port Mapping August 2010 5. Message Formats Editor's note: This section will define the message formats for requesting a Token (PortMappingRequest) and sending a Token (PortMappingResponse). TBC. Begen, et al. Expires February 13, 2011 [Page 10] Internet-Draft Token-Based Port Mapping August 2010 6. Procedures for Token Construction Editor's notes: The Token may contain o Client's IP address o A timestamp to protect against replay attacks o HMAC [RFC2104] of the above information (where only the server knows the HMAC secret) The server conveys the expiration date in the clear to the client via the PortMappingResponse message. Thus, the client can request a new Token before the current one expires. Details are TBC. Begen, et al. Expires February 13, 2011 [Page 11] Internet-Draft Token-Based Port Mapping August 2010 7. Validating Tokens Upon receipt of an RTCP feedback message containing a Token, the server validates the Token. The server considers a Token valid if the source IP address of the RTCP feedback message matches the IP address in the Token, and if the Token has not expired. The IP address is encoded into the Token by the server, using an algorithm known only to the server. This, combined with the expiration, provides protection against DoS attacks so that a client using a certain IP address cannot cause one or more RTP packets to be sent to another client with a different IP address. Begen, et al. Expires February 13, 2011 [Page 12] Internet-Draft Token-Based Port Mapping August 2010 8. SDP Example The SDP describing the scenario given in Figure 1 can be written as: v=0 o=ali 1122334455 1122334466 IN IP4 nack.example.com s=Local Retransmissions t=0 0 a=group:FID 1 2 a=rtcp-unicast:rsi m=video 41000 RTP/AVPF 98 i=Primary Multicast Stream c=IN IP4 233.252.0.2/255 a=source-filter:incl IN IP4 233.252.0.2 198.51.100.1 ; Note 1 a=rtpmap:98 MP2T/90000s a=multicast-rtcp:41500 ; Note 1 a=rtcp:42000 IN IP4 192.0.2.1 ; Note 2 a=rtcp-fb:98 nack ; Note 2 a=mid:1 m=video 42000 RTP/AVPF 99 ; Note 3 i=Unicast Retransmission Stream c=IN IP4 192.0.2.1 a=rtpmap:99 rtx/90000 a=rtcp:42500 ; Note 4 a=fmtp:99 apt=98; rtx-time=5000 a=portmapping-req:30000 ; Note 5 a=mid:2 Figure 2: SDP describing an SSM distribution with support for retransmissions from a local server In this SDP, we highlight the following notes: Note 1: The source stream is multicast from a distribution source with a source IP address of 198.51.100.1 (DS) to the multicast destination address of 233.252.0.2 (G) and port 41000 (P1). The associated RTCP packets are multicast in the same group to port 41500 (P2). Note 2: A retransmission server including feedback target functionality with an IP address of 192.0.2.1 (S) and port of 42000 (P3) is specified with the 'rtcp' attribute. The feedback functionality is enabled for the RTP stream with payload type 98 through the 'rtcp-fb' attribute [RFC4585]. Note 3: The port specified in the second "m" line (for the unicast stream) does not mean anything in this scenario as the client does not send any RTP traffic back to the server. To make this clear, we Begen, et al. Expires February 13, 2011 [Page 13] Internet-Draft Token-Based Port Mapping August 2010 might mandate to use the discard port in this line. Note 4: The server uses port 42500 (P4) for the unicast sessions. Note 5: The "a=portmapping-req" line indicates that a Token MUST be retrieved first before a unicast session can be established and that the Token request MUST be sent to port 30000. Begen, et al. Expires February 13, 2011 [Page 14] Internet-Draft Token-Based Port Mapping August 2010 9. Address Pooling NATs Large-scale NAT (LSN) devices have a pool of public IPv4 addresses and map internal hosts to one of those public IPv4 addresses. As long as an internal host maintains an active mapping in the NAT, the same IPv4 address is assigned to new connections. However, once all of the host's mappings have been deleted (e.g., because of timeout), it is possible that a new connection from that same host will be assigned a different IPv4 address from the pool. When that occurs, the Token will be considered invalid by the server, causing an additional round trip for the client to acquire a fresh token. Any traffic from the host which traverses the NAT will prevent this problem. As the host is sending RTCP receiver reports at least every 5 seconds (Section 6.2 of [RFC3550]) for the multicast session it is receiving, those RTCP messages will be sufficient to prevent this problem. Begen, et al. Expires February 13, 2011 [Page 15] Internet-Draft Token-Based Port Mapping August 2010 10. Security Considerations The Token, which is generated based on a client's IP address and expiration date, provides protection against DoS attacks. An attacker using a certain IP address cannot cause one or more RTP packets to be sent to a victim client who has a different IP address. Begen, et al. Expires February 13, 2011 [Page 16] Internet-Draft Token-Based Port Mapping August 2010 11. IANA Considerations The following contact information shall be used for all registrations in this document: Ali Begen abegen@cisco.com 11.1. Registration of SDP Attributes This document registers a new attribute name in SDP. SDP Attribute ("att-field"): Attribute name: portmapping-req Long form: Port for requesting Token Type of name: att-field Type of attribute: Media level Subject to charset: No Purpose: See this document Reference: This document Values: See this document Begen, et al. Expires February 13, 2011 [Page 17] Internet-Draft Token-Based Port Mapping August 2010 12. Acknowledgments The approach presented in this document came out after discussions with various individuals in the AVT and MMUSIC WGs, and the breakout session held in the Anaheim meeting. We thank each of these individuals. Begen, et al. Expires February 13, 2011 [Page 18] Internet-Draft Token-Based Port Mapping August 2010 13. References 13.1. Normative References [RFC3550] Schulzrinne, H., Casner, S., Frederick, R., and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", STD 64, RFC 3550, July 2003. [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [RFC4566] Handley, M., Jacobson, V., and C. Perkins, "SDP: Session Description Protocol", RFC 4566, July 2006. [RFC4585] Ott, J., Wenger, S., Sato, N., Burmeister, C., and J. Rey, "Extended RTP Profile for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/AVPF)", RFC 4585, July 2006. [RFC5760] Ott, J., Chesterfield, J., and E. Schooler, "RTP Control Protocol (RTCP) Extensions for Single-Source Multicast Sessions with Unicast Feedback", RFC 5760, February 2010. [RFC5234] Crocker, D. and P. Overell, "Augmented BNF for Syntax Specifications: ABNF", STD 68, RFC 5234, January 2008. 13.2. Informative References [I-D.ietf-avt-ports-for-ucast-mcast-rtp] Begen, A. and B. Steeg, "Port Mapping Between Unicast and Multicast RTP Sessions", draft-ietf-avt-ports-for-ucast-mcast-rtp-02 (work in progress), May 2010. [RFC3264] Rosenberg, J. and H. Schulzrinne, "An Offer/Answer Model with Session Description Protocol (SDP)", RFC 3264, June 2002. [I-D.ietf-avt-rapid-acquisition-for-rtp] Steeg, B., Begen, A., Caenegem, T., and Z. Vax, "Unicast- Based Rapid Acquisition of Multicast RTP Sessions", draft-ietf-avt-rapid-acquisition-for-rtp-11 (work in progress), July 2010. [RFC4787] Audet, F. and C. Jennings, "Network Address Translation (NAT) Behavioral Requirements for Unicast UDP", BCP 127, RFC 4787, January 2007. Begen, et al. Expires February 13, 2011 [Page 19] Internet-Draft Token-Based Port Mapping August 2010 [RFC4588] Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R. Hakenberg, "RTP Retransmission Payload Format", RFC 4588, July 2006. [I-D.ietf-avt-app-rtp-keepalive] Marjou, X. and A. Sollaud, "Application Mechanism for keeping alive the Network Address Translator (NAT) mappings associated to RTP flows.", draft-ietf-avt-app-rtp-keepalive-08 (work in progress), June 2010. [RFC5506] Johansson, I. and M. Westerlund, "Support for Reduced-Size Real-Time Transport Control Protocol (RTCP): Opportunities and Consequences", RFC 5506, April 2009. [RFC2104] Krawczyk, H., Bellare, M., and R. Canetti, "HMAC: Keyed- Hashing for Message Authentication", RFC 2104, February 1997. Begen, et al. Expires February 13, 2011 [Page 20] Internet-Draft Token-Based Port Mapping August 2010 Authors' Addresses Ali Begen Cisco 181 Bay Street Toronto, ON M5J 2T3 Canada Email: abegen@cisco.com Dan Wing Cisco Systems, Inc. 170 West Tasman Dr. San Jose, CA 95134 USA Email: dwing@cisco.com Tom VanCaenegem Alcatel-Lucent Copernicuslaan 50 Antwerpen, 2018 Belgium Email: Tom.Van_Caenegem@alcatel-lucent.be Begen, et al. Expires February 13, 2011 [Page 21]