twork Working Group H. Kaplan Internet Draft Acme Packet Intended status: Standards Track October 14, 2011 Expires: April 21, 2012 A Media-based Traceroute Function for the Session Initiation Protocol (SIP) draft-kaplan-dispatch-sip-traceroute-00 Status of this Memo This Internet-Draft is submitted to IETF in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet- Drafts. 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." The list of current Internet-Drafts can be accessed at http://www.ietf.org/ietf/1id-abstracts.txt. The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html. This Internet-Draft will expire on April 14, 2011. Copyright and License 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 Kaplan, et al Expires April 24, 2012 [Page 1] Internet-Draft Media-Traceroute for SIP October 2011 Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the BSD License. Abstract SIP already provides the ability to perform hop-by-hop traceroute for SIP messages using the Max-Forwards header field, in order to determine the reachability path of requests to a target. A new mechanism for media-loopback calls is also being defined separately, which enables test calls to be generated which result in media being looped back to the originator. This document is a strawman proposal for a means of performing hop-by-hop traceroute-style test calls using the media-loopback mechanism, in order to test the media path when SIP sessions go through media-relaying B2BUAs. Table of Contents 1. Terminology...................................................2 2. Introduction..................................................3 2.1. Existing SIP Traceroute Mechanism........................4 3. Solution Overview.............................................4 4. Solution Details..............................................5 4.1. Generating a B2bua-Hops Header Field.....................5 4.2. Processing a Received B2bua-Hops Header Field............5 4.3. Answering the INVITE.....................................6 5. Open Issues...................................................6 6. Security Considerations.......................................6 7. IANA Considerations...........................................7 8. Acknowledgments...............................................7 9. References....................................................7 9.1. Normative References.....................................7 Authors' Addresses................................................7 1. Terminology 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 RFC 2119. The terminology in this document conforms to RFC 2828, "Internet Security Glossary". B2BUA: a SIP Back-to-Back User Agent, which is the logical combination of a User Agent Server (UAS) and User Agent Client (UAC). UAS: a SIP User Agent Server. Kaplan Expires - April 2011 [Page 2] Internet-Draft Media-Traceroute for SIP October 2011 UAC: a SIP User Agent Client. Traceroute: a mechanism to trace a path of hops from an originator to a destination. For IP, this is typically done using the TTL field of the IP header, starting at the value 1 and incrementing by 1 as each IP hop responds with an ICMP error. For SIP this can be done using Max-Forwards header field starting with the value 0, in a similar fashion to the TTL field. It is assumed the reader is already familiar with [draft-media- loopback]. 2. Introduction In many deployments, the media for SIP-created sessions does not flow directly from the originating user's UAC to the answering user's UAS. Often, SIP B2BUAs in the SIP signaling path participate in the media plane, either for injecting media such as rich- ringtones or music-on-hold, or for relaying media in order to provide functions such as transcoding, IPv4-IPv6 conversion, NAT traversal, SRTP termination, media steering, etc. As more and more SIP domains get deployed and interconnect, the odds of a SIP session crossing such media-plane B2BUAs increases, as well as the number of such B2BUAs any given SIP session may go through. In other words, any given SIP session may cross any number of B2BUAs both in the SIP signaling plane as well as media-plane. If failures or degradation occurs in the media plane, it is difficult to determine where in the media path they occur. In order to aid managing and troubleshooting SIP-based sessions and media crossing such B2BUAs, it would be useful to be able to test the media path to each B2BUA separately from the source. A mechanism to perform media-loopback test sessions is being defined in [draft- media-loopback], but it would be difficult to use the mechanism directly to test B2BUAs because typically the B2BUAs do not have an AoR to be targeted, nor is it known a priori which B2BUAs will be crossed for any given session. For example, suppose calls from Alice to Bob have media problems. Alice would like to test the media path to each B2BUA in the path to Bob separately, to determine which segment has the issues. Alice cannot target the B2BUAs directly for each test call, because she doesn't know what URIs to use to target them; nor would using such URIs guarantee the same media path be used as a call to Bob. A better solution would be to make a test call targeted to Bob, but with a SIP traceroute-type mechanism that makes the call terminate at the B2BUAs, such that she can perform test sessions to test the media path to each downstream B2BUA. Kaplan Expires - April 2011 [Page 3] Internet-Draft Media-Traceroute for SIP October 2011 This document defines how such a mechanism can be employed, using the [draft-media-loopback] mechanism with a new SIP header field such that a SIP User Agent can make multiple test calls, each reaching a B2BUA further downstream. Each B2BUA in the path that supports this mechanism would answer the media-loopback call, and thus the originating SIP UA can test the media path up to that B2BUA. 2.1. Existing SIP Traceroute Mechanism The Max-Forwards header field can already be used to perform a SIP- request traceroute mechanism by generating a SIP request initially using a Max-Forwards value of 0, receiving a 483 Too Many Hops response from the next-hop, and then incrementing the value for subsequent SIP requests, thereby reaching SIP devices further and further downstream and receiving 483 from each of them. Unfortunately, using the Max-Forwards header field for a media- loopback traceroute will not work for this document's purpose, for several reasons. The first problem is that Max-Forwards verification is done early in SIP message receive processing. Secondly, SIP Proxies that have no involvement in the media-plane handle it as a failure event, responding with 483. Third, a 483 response may not contain a Contact URI, so that the reached SIP device isn't even known. Lastly and most importantly, many SIP B2BUAs reset the Max-Forwards header field value to 70 or some new value - even those that have no involvement in the media-plane. 3. Solution Overview This document proposes a new SIP header field named "B2bua-Hops", with behavior similar to Max-Forwards. Like Max-Forwards, B2bua- Hops encodes an integer value from 0-255 which gets decremented by each SIP hop and if it reaches 0 stops subsequent forwarding. Unlike Max-Forwards, only B2BUAs involved in the media-plane would decrement the B2bua-Hops value. [Note: it is TBD if this should be done using a SIP header, or an SDP attribute instead, or both] To perform a SIP media-plane traceroute, the originating UAC generates a SIP INVITE to a target AoR, with SDP based on [draft- media-loopback], and a B2bua-Hops header value of 0. When the request reaches the first B2BUA that supports this mechanism, that B2BUA will answer the INVITE to establish the dialog and create a media-loopback session. The originating UAC can then generate another INVITE to the same target AoR with a B2bua-Hops header value Kaplan Expires - April 2011 [Page 4] Internet-Draft Media-Traceroute for SIP October 2011 of 1, which will reach the second B2BUA that supports this mechanism, and so on. Using this mechanism a SIP UAC can test the path from itself to each successive B2BUA on the path to a target. Such a mechanism could also be useful for establishing a permanent test call between an Enterprise and a Service Provider across a SIP Trunk, for example, or for automated measurement systems to test the media path between domains, etc. 4. Solution Details The focus of this document is enabling UACs to perform media-path loopback test calls using a traceroute-style mechanism. Currently this involves using a new SIP header, and the procedures for the header comprise the majority of the normative text in this section. Arguably the new SIP header is useful beyond the purposes of media- loopback testing, for example to detect looped calls. This document makes no judgment in that regard, and further discussion is needed in the IETF beforehand. 4.1. Generating a B2bua-Hops Header Field A SIP User Agent Client MAY generate a B2bua-Hops header field in any dialog-creating INVITE request, and SHOULD use the default value of 70 unless it has some reason to use another value. Using another value may make sense for cases such as traceroute testing, for example. [Note: it is TBD if the recommendation should be a "SHOULD" or "MUST" to generate the header in all out-of-dialog requests, instead of "MAY"] B2BUAs logically generate all SIP headers on their UAC side as a User Agent, but for the purposes of this document the value of the B2bua-Hops header field generated on the UAC side MUST be copied from the value received on the UAS side of the B2BUA. If the B2BUA is involved in the media plane, it MUST decrement the copied value by 1 before encoding. All dialog-creating INVITE requests generated by a B2BUA based on receiving a dialog-creating INVITE request MUST use the same (decremented) value. SIP devices MUST NOT reset or increase the B2bua-Hops value of received SIP messages, ever. 4.2. Processing a Received B2bua-Hops Header Field When a SIP B2BUA or UAS receives an out-of-dialog SIP INVITE request, it checks for the existence of the B2bua-Hops header field. Kaplan Expires - April 2011 [Page 5] Internet-Draft Media-Traceroute for SIP October 2011 If one exists and its value is 0, the device MAY act as the final recipient and process the request as defined in section 5; otherwise the request MUST be rejected with a 483 (Too many hops) response, and a local Contact URI MUST be included in the 483 response. If the B2bua-Hops header field value is greater than 0, the SIP device MUST decrement it by 1 as described in the previous section and continue SIP processing. [Note: it is TBD if responding with 483 is appropriate, vs. a new or other existing response code] 4.3. Answering the INVITE If a SIP B2BUA or UAS receives a dialog-creating INVITE request with a B2bua-Hops header value of 0, with SDP based on [draft-media- loopback], and the policies of the B2BUA/UAS allow it to answer such a request, then it is answered as if the target of the request were the local SIP B2BUA/UAS. The normal procedures of SIP apply, as well as [draft-media-loopback], as if the request had been targeted at the local device all along. [Open Issue: how does the UAC know the request reached a B2BUA vs. the final UAS? (e.g., how does it know when to stop testing?)] 5. Open Issues There are many ways to achieve the goals of this document, each with pros/cons. If the IETF decides to take this work on in a Working Group, the Working Group can debate which approach is the best. This section highlights some of the known open issues or options. - Should the mechanism be done using separate INVITEs as described in this document, or rather as a single INVITE and forked at each B2BUA, creating separate dialogs implicitly? - Should we use a SIP header or an SDP attribute, or both? - Should we attempt to "fix" Max-Forwards instead of creating a new header? (is it even fixable?) - Should we use B2bua-Hops for things beyond media-loopback? (e.g., in other method types, for all requests all the time) - How does the UAC know when the request finally reached the ultimate UAS? (e.g., use a param somewhere) 6. Security Considerations There are security implications for the mechanism defined in this document. Answering media-loopback calls in a B2BUA consumes resources on the B2BUA, and network bandwidth in between; therefor, B2BUAs should have some means of controlling who can make such test Kaplan Expires - April 2011 [Page 6] Internet-Draft Media-Traceroute for SIP October 2011 calls, how many concurrent calls can be established and maintained, and for how long. Such policies are typically vendor-specific and do not need to be defined in this document. 7. IANA Considerations This document makes no request of IANA yet - if the new header approach is maintained, then a new header will be registered. 8. Acknowledgments The general concept of performing media-loopback on a hop-by-hop basis using a decrementing header traceroute style approach came out of discussions several years ago, between the author, Kaynam Hedayat, Nagarjuna Venna, Patrick MeLampy, and others. Other people that have contributed to the topic over the years since then: Zaid Ally, Dianna Stiller, Jon Boone, and several others whom I have lost the names of since. Funding for the RFC Editor function is provided by the IETF Administrative Support Activity (IASA). 9. References 9.1. Normative References [draft-media-loopback] Hedayat, K., et al, "An Extension to the Session Description Protocol (SDP) for Media Loopback", draft-ietf- mmusic-media-loopback-16, September 2011. Author's Address Hadriel Kaplan Acme Packet Email: hkaplan@acmepacket.com Kaplan Expires - April 2011 [Page 7]