RFC 7002 PROPOSED STANDARD RTP · RTCP · METRICS

RTCP XR Discard Count: RTP Control Protocol (RTCP) Extended Report (XR) Block for Discard Count Metric Reporting

A small RTCP report block that tells a sender how many packets arrived fine but were never played.

On rfc-editor.org ↗
PUBLISHED
September 2013
AUTHORS
A. Clark, G. Zorn, Q. Wu
STREAM
IETF
BLOCK TYPE
24
BLOCK LENGTH
2 (32-bit words)
BLOCK SIZE
3 words of 32 bits, 12 bytes
OVER-RANGE VALUE
0xFFFFFFFE
UNAVAILABLE MEASUREMENT
0xFFFFFFFF
SDP PARAMETER
pkt-discard-count
DOI
10.17487/RFC7002

RFC text

Contents
Original text · RFC Editor View on rfc-editor.org ↗
Internet Engineering Task Force (IETF)                          A. Clark
Request for Comments: 7002                                      Telchemy
Category: Standards Track                                        G. Zorn
ISSN: 2070-1721                                              Network Zen
                                                                   Q. Wu
                                                                  Huawei
                                                          September 2013


         RTP Control Protocol (RTCP) Extended Report (XR) Block
                   for Discard Count Metric Reporting

Abstract

   This document defines an RTP Control Protocol (RTCP) Extended Report
   (XR) block that allows the reporting of a simple discard count metric
   for use in a range of RTP applications.

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 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc7002.

Copyright Notice

   Copyright (c) 2013 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
   described in the Simplified BSD License.

Table of Contents

   1.  Introduction  . . . . . . . . . . . . . . . . . . . . . . . .   2
     1.1.  Discard Count Report Block  . . . . . . . . . . . . . . .   2
     1.2.  RTCP and RTCP Extended Reports  . . . . . . . . . . . . .   3
     1.3.  Performance Metrics Framework . . . . . . . . . . . . . .   3
     1.4.  Applicability . . . . . . . . . . . . . . . . . . . . . .   3
   2.  Terminology . . . . . . . . . . . . . . . . . . . . . . . . .   3
   3.  Discard Count Metrics Block . . . . . . . . . . . . . . . . .   4
     3.1.  Report Block Structure  . . . . . . . . . . . . . . . . .   5
     3.2.  Definition of Fields in the Discard Count Metrics Block .   5
   4.  SDP Signaling . . . . . . . . . . . . . . . . . . . . . . . .   7
     4.1.  SDP rtcp-xr Attribute Extension . . . . . . . . . . . . .   7
     4.2.  Offer/Answer Usage  . . . . . . . . . . . . . . . . . . .   7
   5.  IANA Considerations . . . . . . . . . . . . . . . . . . . . .   7
     5.1.  New RTCP XR Block Type Value  . . . . . . . . . . . . . .   8
     5.2.  New RTCP XR SDP Parameter . . . . . . . . . . . . . . . .   8
     5.3.  Contact Information for Registrations . . . . . . . . . .   8
   6.  Security Considerations . . . . . . . . . . . . . . . . . . .   8
   7.  Contributors  . . . . . . . . . . . . . . . . . . . . . . . .   8
   8.  Acknowledgments . . . . . . . . . . . . . . . . . . . . . . .   9
   9.  References  . . . . . . . . . . . . . . . . . . . . . . . . .   9
     9.1.  Normative References  . . . . . . . . . . . . . . . . . .   9
     9.2.  Informative References  . . . . . . . . . . . . . . . . .  10
   Appendix A.  Metrics Represented Using the Template from RFC 6390  11

1.  Introduction

1.1.  Discard Count Report Block

   This document defines a new block type to augment those defined in
   [RFC3611] for use in a range of RTP applications.  The new block type
   supports the reporting of the number of packets that are received
   correctly but are never played out, typically because they arrive too
   late (buffer underflow) or too early (buffer overflow) to be played
   out.  The metric is applicable both to systems that use packet loss
   repair techniques (such as forward error correction [RFC5109] or
   retransmission [RFC4588]) and to those that do not.

   This metric is useful for identifying the existence, and
   characterizing the severity, of packet transport problems that may
   affect users' perceptions of a service delivered over RTP.

   This block may be used in conjunction with [RFC7003], which provides
   additional information on the pattern of discarded packets.  However,
   the metric in [RFC7003] may be used independently of the metrics in
   this block.

   When a Discard Count Metrics Block is sent together with a Burst/Gap
   Discard Metrics Block (defined in [RFC7003]) to the media sender or
   RTP-based network management system, the information carried in the
   Discard Count Metrics Block and the Burst/Gap Discard Metrics Block
   allows systems receiving the blocks to calculate burst/gap summary
   statistics (e.g., the gap discard rate).

   The metric belongs to the class of transport-related end-system
   metrics defined in [RFC6792].

1.2.  RTCP and RTCP Extended Reports

   The use of RTCP for reporting is defined in [RFC3550].  [RFC3611]
   defined an extensible structure for reporting using an RTCP Extended
   Report (XR).  This document defines a new Extended Report block for
   use with [RFC3550] and [RFC3611].

1.3.  Performance Metrics Framework

   "Guidelines for Considering New Performance Metric Development"
   [RFC6390] provides guidance on the definition and specification of
   performance metrics.  "Guidelines for Use of the RTP Monitoring
   Framework" [RFC6792] provides guidance for reporting block format
   using RTCP XR.  The metrics block described in this document is in
   accordance with the guidelines in [RFC6390] and [RFC6792].

1.4.  Applicability

   This metric is believed to be applicable to a large class of RTP
   applications that use a de-jitter buffer [RFC5481].

   Discards due to late or early arriving packets affect user
   experience.  The reporting of discards alerts senders and other
   receivers to the need to adjust their transmission or reception
   strategies.  The reports allow network managers to diagnose these
   user experience problems.

   The ability to detect duplicate packets can be used by managers to
   detect network layer or sender behavior, which may indicate network
   or device issues.  Based on the reports, these issues may be
   addressed prior to any impact on user experience.

2.  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 [RFC2119].

   In addition, the following terms are defined:

   Received, Lost, and Discarded

      A packet shall be regarded as lost if it fails to arrive within an
      implementation-specific time window.  A packet that arrives within
      this time window but is either too early or too late to be played
      out or is thrown away before playout due to packet duplication or
      redundancy shall be regarded as discarded.  A packet shall not be
      regarded as discarded if it arrives within this time window but is
      dropped during decoding by some higher layer decoder, e.g., due to
      a decoding error.  A packet shall be classified as one of the
      following: received (or OK), discarded, or lost.  The discard
      count metric counts only discarded packets.  The metric
      "cumulative number of packets lost" defined in [RFC3550] reports a
      count of packets lost from the media stream (single
      synchronization source (SSRC) within a single RTP session).
      Similarly, the metric "number of packets discarded" reports a
      count of packets discarded from the media stream (single SSRC
      within a single RTP session) arriving at the receiver.  Another
      metric defined in [RFC5725] is available to report on packets that
      are not recovered by any repair techniques that may be in use.

3.  Discard Count Metrics Block

   Metrics in this block report on the number of packets discarded in
   the stream arriving at the RTP end system.  The measurement of these
   metrics is made at the receiving end of the RTP stream.  Instances of
   this metrics block use the SSRC to refer to the separate auxiliary
   Measurement Information Block [RFC6776], which describes measurement
   periods in use (see [RFC6776], Section 4.2).  This metrics block
   relies on the measurement interval in the Measurement Information
   Block indicating the span of the report and MUST be sent in the same
   compound RTCP packet as the Measurement Information Block.  If the
   measurement interval is not received in the same compound RTCP packet
   as this metrics block, this metrics block MUST be discarded.

3.1.  Report Block Structure

   The structure of the Discard Count Metrics Block is as follows.

    0                   1                   2                   3
    0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |     BT=24     | I |DT |  resv |      Block Length = 2         |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |                        SSRC of Source                         |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   |                        Discard Count                          |
   +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

                     Figure 1: Report Block Structure

3.2.  Definition of Fields in the Discard Count Metrics Block

   Block Type (BT): 8 bits

      A Discard Count Metrics Block is identified by the constant 24.

   Interval Metric flag (I): 2 bits

      This field indicates whether the reported metric is an Interval,
      Cumulative, or Sampled metric [RFC6792]:

         I=10: Interval Duration - the reported value applies to the
         most recent measurement interval duration between successive
         metrics reports.

         I=11: Cumulative Duration - the reported value applies to the
         accumulation period characteristic of cumulative measurements.

         I=01: Sampled Value - the reported value is a sampled
         instantaneous value.

      In this document, the discard count metric can only be measured
      over definite intervals and cannot be sampled.  Accordingly, the
      value I=01, indicating a sampled value, MUST NOT be sent, and MUST
      be discarded when received.  In addition, the value I=00 is
      reserved and also MUST NOT be sent, and MUST be discarded when
      received.

   Discard Type (DT): 2 bits

      This field is used to identify the discard type used in this
      report block.  The discard type is defined as follows:

         00: Report packet discarded or being thrown away before playout
         due to packet duplication.

         01: Report packet discarded due to too early to be played out.

         10: Report packet discarded due to too late to be played out.

      The value DT=11 is reserved for future definition and MUST NOT be
      sent, and MUST be discarded when received.

      An endpoint MAY report any combination of discard types in each
      reporting interval by including several Discard Count Metrics
      Blocks in a single RTCP XR packet.

      Some systems send duplicate RTP packets for robustness or error
      resilience.  This is NOT RECOMMENDED since it breaks RTCP packet
      statistics.  If duplication is desired for error resilience, the
      mechanism described in [RTPDUP] can be used, since this will not
      cause breakage of RTP streams or RTCP statistics.

   Reserved (resv): 4 bits

      These bits are reserved.  They MUST be set to zero by senders and
      ignored by receivers (see [RFC6709], Section 4.2).

   Block Length: 16 bits

      The length of this report block in 32-bit words, minus one, in
      accordance with the definition in [RFC3611].  This field MUST be
      set to 2 to match the fixed length of the report block.  The block
      MUST be discarded if the block length is set to a different value.

   SSRC of Source: 32 bits

      As defined in Section 4.1 of [RFC3611].

   Discard Count

      Number of packets discarded over the period (Interval or
      Cumulative) covered by this report.

      The measured value is an unsigned value.  If the measured value
      exceeds 0xFFFFFFFD, the value 0xFFFFFFFE MUST be reported to
      indicate an over-range measurement.  If the measurement is
      unavailable, the value 0xFFFFFFFF MUST be reported.

      Note that the number of packets expected in the period associated
      with this metric (whether Interval or Cumulative) is available
      from the difference between a pair of extended sequence numbers in
      the Measurement Information Block [RFC6776], so it need not be
      repeated in this block.

4.  SDP Signaling

   [RFC3611] defines the use of the Session Description Protocol (SDP)
   [RFC4566] for signaling the use of XR blocks.  However, XR blocks MAY
   be used without prior signaling (see Section 5 of RFC 3611).

4.1.  SDP rtcp-xr Attribute Extension

   This section augments the SDP [RFC4566] attribute "rtcp-xr" defined
   in [RFC3611] by providing an additional value of "xr-format" to
   signal the use of the report block defined in this document.  The
   ABNF [RFC5234] syntax is as follows.

   xr-format =/ xr-pdc-block
   xr-pdc-block = "pkt-discard-count"


4.2.  Offer/Answer Usage

   When SDP is used in Offer/Answer context, the SDP Offer/Answer usage
   defined in [RFC3611] for unilateral "rtcp-xr" attribute parameters
   applies.  For detailed usage of Offer/Answer for unilateral
   parameters, refer to Section 5.2 of [RFC3611].

5.  IANA Considerations

   New block types for RTCP XR are subject to IANA registration.  For
   general guidelines on IANA considerations for RTCP XR, refer to
   [RFC3611].

5.1.  New RTCP XR Block Type Value

   This document assigns the block type value 24 in the IANA "RTP
   Control Protocol Extended Reports (RTCP XR) Block Type Registry" to
   the "Discard Count Metrics Block".

5.2.  New RTCP XR SDP Parameter

   This document also registers a new parameter "pkt-discard-count" in
   the "RTP Control Protocol Extended Reports (RTCP XR) Session
   Description Protocol (SDP) Parameters Registry".

5.3.  Contact Information for Registrations

   The following contact information is provided for all registrations
   in this document:

   Qin Wu (sunseawq@huawei.com)
   101 Software Avenue, Yuhua District
   Nanjing, Jiangsu  210012
   China

6.  Security Considerations

   In some situations, returning very detailed error information (e.g.,
   over-range measurement or measurement unavailable) using this report
   block can provide an attacker with insight into the security
   processing.  Where this is a concern, the implementation should apply
   encryption and authentication to this report block.  For example,
   this can be achieved by using the Audio-Visual Profile with Feedback
   (AVPF) profile together with the Secure RTP profile, as defined in
   [RFC3711]; an appropriate combination of those two profiles ("SAVPF")
   is specified in [RFC5124].

   Besides this, it is believed that this RTCP XR block introduces no
   new security considerations beyond those described in [RFC3611].
   This block does not provide per-packet statistics, so the risk to
   confidentiality documented in Section 7, paragraph 3 of [RFC3611]
   does not apply.

7.  Contributors

   Geoff Hunt wrote the initial draft of this document.

8.  Acknowledgments

   The authors gratefully acknowledge reviews and feedback provided by
   Bruce Adams, Philip Arden, Amit Arora, Claire Bi, Bob Biskner,
   Gonzalo Camarillo, Benoit Claise, Kevin Connor, Claus Dahm, Spencer
   Dawkins, Randy Ethier, Roni Even, Stephen Farrel, Jim Frauenthal,
   Kevin Gross, Albert Higashi, Tom Hock, Shane Holthaus, Paul Jones,
   Rajesh Kumar, Keith Lantz, Jonathan Lennox, Mohamed Mostafa, Amy
   Pendleton, Colin Perkins, Mike Ramalho, Ravi Raviraj, Dan Romascanu,
   Albrecht Schwarz, Varun Singh, Tom Taylor, Dan Wing, and Hideaki
   Yamada.

9.  References

9.1.  Normative References

   [RFC2119]  Bradner, S., "Key words for use in RFCs to Indicate
              Requirement Levels", BCP 14, RFC 2119, March 1997.

   [RFC3550]  Schulzrinne, H., Casner, S., Frederick, R., and V.
              Jacobson, "RTP: A Transport Protocol for Real-Time
              Applications", STD 64, RFC 3550, July 2003.

   [RFC3611]  Friedman, T., Caceres, R., and A. Clark, "RTP Control
              Protocol Extended Reports (RTCP XR)", RFC 3611, November
              2003.

   [RFC3711]  Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K.
              Norrman, "The Secure Real-time Transport Protocol (SRTP)",
              RFC 3711, March 2004.

   [RFC4566]  Handley, M., Jacobson, V., and C. Perkins, "SDP: Session
              Description Protocol", RFC 4566, July 2006.

   [RFC5124]  Ott, J. and E. Carrara, "Extended Secure RTP Profile for
              Real-time Transport Control Protocol (RTCP)-Based Feedback
              (RTP/SAVPF)", RFC 5124, February 2008.

   [RFC5234]  Crocker, D. and P. Overell, "Augmented BNF for Syntax
              Specifications: ABNF", STD 68, RFC 5234, January 2008.

   [RFC6709]  Carpenter, B., Aboba, B., and S. Cheshire, "Design
              Considerations for Protocol Extensions", RFC 6709,
              September 2012.

   [RFC6776]  Clark, A. and Q. Wu, "Measurement Identity and Information
              Reporting Using a Source Description (SDES) Item and an
              RTCP Extended Report (XR) Block", RFC 6776, October 2012.

9.2.  Informative References

   [RFC4588]  Rey, J., Leon, D., Miyazaki, A., Varsa, V., and R.
              Hakenberg, "RTP Retransmission Payload Format", RFC 4588,
              July 2006.

   [RFC5109]  Li, A., "RTP Payload Format for Generic Forward Error
              Correction", RFC 5109, December 2007.

   [RFC5481]  Morton, A. and B. Claise, "Packet Delay Variation
              Applicability Statement", RFC 5481, March 2009.

   [RFC5725]  Begen, A., Hsu, D., and M. Lague, "Post-Repair Loss RLE
              Report Block Type for RTP Control Protocol (RTCP) Extended
              Reports (XRs)", RFC 5725, February 2010.

   [RFC6390]  Clark, A. and B. Claise, "Guidelines for Considering New
              Performance Metric Development", BCP 170, RFC 6390,
              October 2011.

   [RFC6792]  Wu, Q., Hunt, G., and P. Arden, "Guidelines for Use of the
              RTP Monitoring Framework", RFC 6792, November 2012.

   [RFC7003]  Clark, A., Huang, R., and Q. Wu, Ed., "RTP Control
              Protocol(RTCP) Extended Report (XR) Block for Burst/Gap
              Discard Metric Reporting", RFC 7003, September 2013.

   [RTPDUP]   Begen, A. and C. Perkins, "Duplicating RTP Streams", Work
              in Progress, March 2013.

Appendix A.  Metrics Represented Using the Template from RFC 6390

   a.  Number of Packets Discarded Metric

       *  Metric Name: Number of RTP packets discarded.

       *  Metric Description: Number of RTP packets discarded over the
          period covered by this report.

       *  Method of Measurement or Calculation: See Section 3.2, Discard
          Count definition.

       *  Units of Measurement: See Section 3.2, Discard Count
          definition.

       *  Measurement Point(s) with Potential Measurement Domain: See
          Section 3, 1st paragraph.

       *  Measurement Timing: See Section 3, 1st paragraph for
          measurement timing and Section 3.2 for Interval Metric flag.

       *  Use and Applications: See Section 1.4.

       *  Reporting Model: See RFC 3611.

Authors' Addresses

   Alan Clark
   Telchemy Incorporated
   2905 Premiere Parkway, Suite 280
   Duluth, GA  30097
   USA

   EMail: alan.d.clark@telchemy.com


   Glen Zorn
   Network Zen
   227/358 Thanon Sanphawut
   Bang Na, Bangkok  10260
   Thailand

   Phone: +66 (0) 8-1000-4155
   EMail: glenzorn@gmail.com


   Qin Wu
   Huawei
   101 Software Avenue, Yuhua District
   Nanjing, Jiangsu  210012
   China

   EMail: sunseawq@huawei.com

ISO Plain Language

On this page
PROBLEMA receiver cannot always tell whether a packet that is missing from a media stream was lost in the network or arrived and was thrown away. The second case matters, because a packet that arrives too late to be played out is a transport problem the sender may be able to fix.
MECHANISMThe RFC defines an RTCP Extended Report block, block type 24, that carries the number of packets discarded at the receiver, split by reason: a duplicate, a packet that arrived too early, and a packet that arrived too late. The block also says whether the count covers one reporting interval or the whole accumulation period.
WHAT IT SHOWSSenders and network managers can see how much of the stream was thrown away and why, and can adjust their sending or the network before users notice. Used with the burst and gap discard block of RFC 7003, the two allow summary rates such as the gap discard rate to be calculated.

§1What the Block Counts

The RFC defines a new block type to augment those defined in RFC 3611, for use in a range of RTP applications. The new block type supports the reporting of the number of packets that are received correctly but are never played out, typically because they arrive too late, which the RFC calls buffer underflow, or too early, which it calls buffer overflow. The metric applies both to systems that use packet loss repair techniques such as forward error correction (RFC 5109) or retransmission (RFC 4588), and to those that do not.

The metric is useful for identifying the existence of packet transport problems, and for characterising their severity, where they may affect users' perceptions of a service delivered over RTP.

The block may be used together with the burst and gap discard block of RFC 7003, which reports the pattern of discarded packets. When both are sent to the media sender or to an RTP-based network management system, the information in the two allows the receiving system to calculate burst and gap summary statistics such as the gap discard rate. The metric in RFC 7003 can also be used on its own. The metric belongs to the class of transport-related end-system metrics defined in RFC 6792.

RTCP reporting is defined in RFC 3550, and RFC 3611 defined the extensible structure for reporting using an RTCP Extended Report (XR), which is the structure this new block is added to. The block follows the guidance on defining performance metrics in RFC 6390 and on reporting block formats in RFC 6792.

This block counts the packets that arrived safely but were then thrown away instead of being played. That happens when a packet arrives so late that its moment has passed, or so early that the receiver has no room for it.

The count is worth reporting because it separates a timing problem from a loss problem. A packet that never arrived is one thing; a packet that arrived and was dropped is another, and the sender can often do something about the second. A companion block, defined in RFC 7003, adds the pattern of the drops, and the two together let the receiver work out summary rates such as the gap discard rate.

§2Received, Lost and Discarded

The RFC uses three terms for the fate of a packet. A packet is regarded as lost if it fails to arrive within an implementation-specific time window. A packet that arrives within that window but is either too early or too late to be played out, or is thrown away before playout because of packet duplication or redundancy, is regarded as discarded. A packet is not regarded as discarded if it arrives within the window but is dropped during decoding by a higher layer decoder, for example because of a decoding error.

Every packet falls into exactly one of three classes: received, which the RFC also calls OK, discarded, or lost. The discard count metric counts only discarded packets. The metric for the cumulative number of packets lost, defined in RFC 3550, reports packets lost from the media stream, a single synchronization source within a single RTP session. The metric for the number of packets discarded reports the packets discarded from the media stream arriving at the receiver, again for a single synchronization source within a single RTP session. A further metric defined in RFC 5725 is available to report on packets that no repair technique recovers.

Three words matter here. A packet is lost if it does not arrive within a time window the implementation chooses. A packet is discarded if it arrives in time but is still thrown away, because it arrived too early, too late, or was a duplicate. A packet that arrives and is then dropped by a decoder is neither, because nothing was lost and nothing was thrown away by the transport.

Every packet is exactly one of received, discarded or lost. This block counts only the discarded ones.

THE THREE FATES OF A PACKET
OutcomeWhen it appliesCounted by this block
ReceivedThe packet arrived in time and was played, or was dropped later by a decoderNo
DiscardedThe packet arrived in time but was thrown away: a duplicate, too early, or too lateYes
LostThe packet did not arrive within the implementation's time windowNo

§3.1The Block Layout

The metrics in the block report on the number of packets discarded in the stream arriving at the RTP end system, and the measurement is made at the receiving end. Instances of the block use the SSRC to refer to the separate auxiliary Measurement Information Block of RFC 6776, which describes the measurement periods in use. The block relies on the measurement interval in that block to indicate the span of the report, and it MUST be sent in the same compound RTCP packet as it. If the measurement interval is not received in the same compound RTCP packet, the metrics block MUST be discarded.

The block itself has a fixed layout of three 32-bit words. The first word holds the block type, the interval metric flag, the discard type and the reserved bits, followed by the block length. The second word holds the SSRC of the source, as defined in Section 4.1 of RFC 3611. The third word holds the discard count itself.

The count is measured at the receiving end of the stream. The report has to be sent together with the separate block that says which time period the report covers, and in the same packet. If that second block is missing, the receiver is told to throw the count away, because a count without a period has no meaning.

The block itself is three words long: a header word, a word naming the stream the count belongs to, and the count.

THE DISCARD COUNT METRICS BLOCK
Discard Count Metrics Block
FieldNameSizeValueDescription
BTBT1 byte24Block type, the constant 24.
INTERVAL-FLAGI2 bitsInterval metric flag: interval, cumulative or sampled.
DISCARD-TYPEDT2 bitsDiscard type: duplicate, too early or too late.
RESERVEDresv4 bitsReserved bits; senders set them to zero and receivers ignore them.
BLOCK-LENGTHBlock Length2 bytes2Length of the block in 32-bit words minus one.
SSRCSSRC of Source4 bytesThe synchronization source the count belongs to.
DISCARD-COUNTDiscard Count4 bytesNumber of packets discarded over the period covered by the report.

§3.2The Fields and Their Values

The block type is 8 bits and identifies the block with the constant 24. The interval metric flag is 2 bits and says whether the reported value covers the most recent measurement interval, I=10, the accumulation period of a cumulative measurement, I=11, or is a sampled instantaneous value, I=01. The discard count can only be measured over definite intervals and cannot be sampled, so the sampled value MUST NOT be sent and MUST be discarded when received, and the reserved value I=00 MUST NOT be sent and MUST be discarded when received.

The discard type is 2 bits and identifies why the packets were discarded: 00 for a packet discarded or being thrown away before playout because of packet duplication; 01 for a packet discarded because it was too early to be played out; and 10 for a packet discarded because it was too late to be played out. The value 11 is reserved for future definition, MUST NOT be sent and MUST be discarded when received. An endpoint MAY report any combination of discard types in each reporting interval by including several Discard Count Metrics Blocks in a single RTCP XR packet. The RFC notes that some systems send duplicate RTP packets for robustness, which is NOT RECOMMENDED because it breaks RTCP packet statistics, and points to the duplication mechanism of [RTPDUP] as an alternative that does not break the statistics.

The reserved field is 4 bits. Senders MUST set them to zero and receivers ignore them. The block length is 16 bits and gives the length of the report block in 32-bit words minus one, in accordance with RFC 3611. It MUST be set to 2, and the block MUST be discarded if the length is set to any other value.

The discard count is the number of packets discarded over the period covered by the report, whether an interval or a cumulative period. The measured value is unsigned. If it exceeds 0xFFFFFFFD, the value 0xFFFFFFFE MUST be reported to indicate an over-range measurement, and if the measurement is unavailable, the value 0xFFFFFFFF MUST be reported. The number of packets expected over the period is not repeated in this block, because it can be worked out from the difference between a pair of extended sequence numbers in the Measurement Information Block.

The header of the block says three things: how the count is measured, why the packets were dropped, and how long the block is. Only three reasons are defined: the packet was a duplicate, it was too early, or it was too late. A receiver that sees the fourth value, or a length other than the one defined, is told to ignore the block.

The count itself is a plain number. Two special values report trouble with the measurement: one means the number was too large to fit, the other that no measurement was available at all.

THE DISCARD TYPES THE BLOCK CAN REPORT
  • 00Discarded as a duplicate — The packet was discarded or being thrown away before playout because of packet duplication.
  • 01Discarded, too early — The packet was discarded because it was too early to be played out.
  • 10Discarded, too late — The packet was discarded because it was too late to be played out.
  • 11Reserved — Reserved for future definition. It must not be sent, and a block carrying it must be discarded.

§4Signalling the Block with SDP

RFC 3611 defines the use of the Session Description Protocol (RFC 4566) for signalling the use of XR blocks, although XR blocks MAY be used without any prior signalling, as described in Section 5 of that RFC.

This RFC extends the rtcp-xr SDP attribute of RFC 3611 with an additional value of xr-format to signal the use of the report block defined here. The ABNF is xr-format =/ xr-pdc-block and xr-pdc-block = "pkt-discard-count". Where SDP is used in an offer and answer context, the unilateral rtcp-xr attribute parameter usage defined in Section 5.2 of RFC 3611 applies.

IANA has assigned the block type value 24 in the RTCP XR Block Type Registry to the Discard Count Metrics Block, and the new parameter pkt-discard-count in the RTCP XR SDP Parameters Registry.

The block can be used without any negotiation at all. If the two ends do want to agree on it first, it is announced in the session description, in the same attribute that announces the other extended report blocks, with the value pkt-discard-count.

§6Security Considerations

In some situations, returning very detailed error information, such as an over-range measurement or an unavailable measurement, using this report block can give an attacker insight into the security processing. Where this is a concern, the implementation should apply encryption and authentication to the report block. The RFC gives the example of the Audio-Visual Profile with Feedback profile together with the Secure RTP profile, as defined in RFC 3711, a combination specified as SAVPF in RFC 5124.

Beyond that, the RFC believes the block introduces no new security considerations beyond those described in RFC 3611. The block does not provide per-packet statistics, so the risk to confidentiality described in Section 7 of RFC 3611 does not apply.

The only risk the document points at is that the two special values in the count, which report a failed measurement, could tell an attacker something about how the system is working. Where that matters, the report should be encrypted and authenticated, for example with the secure profile for RTP feedback.

Other than that, the block adds no new risk, because it reports a summary count rather than anything about individual packets.

§7References

[RFC2119]RFC 2119Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997.March 1997[RFC3550]RFC 3550Schulzrinne, H., Casner, S., Frederick, R., and V. Jacobson, "RTP: A Transport Protocol for Real-Time Applications", STD 64, RFC 3550, July 2003.July 2003[RFC3611]RFC 3611Friedman, T., Caceres, R., and A. Clark, "RTP Control Protocol Extended Reports (RTCP XR)", RFC 3611, November 2003.November 2003[RFC3711]RFC 3711Baugher, M., McGrew, D., Naslund, M., Carrara, E., and K. Norrman, "The Secure Real-time Transport Protocol (SRTP)", RFC 3711, March 2004.March 2004[RFC4566]RFC 4566Handley, M., Jacobson, V., and C. Perkins, "SDP: Session Description Protocol", RFC 4566, July 2006.July 2006[RFC5109]RFC 5109Li, A., "RTP Payload Format for Generic Forward Error Correction", RFC 5109, December 2007.December 2007[RFC5124]RFC 5124Ott, J. and E. Carrara, "Extended Secure RTP Profile for Real-time Transport Control Protocol (RTCP)-Based Feedback (RTP/SAVPF)", RFC 5124, February 2008.February 2008[RFC5725]RFC 5725Begen, A., Hsu, D., and M. Lague, "Post-Repair Loss RLE Report Block Type for RTP Control Protocol (RTCP) Extended Reports (XRs)", RFC 5725, February 2010.February 2010[RFC6390]RFC 6390Clark, A. and B. Claise, "Guidelines for Considering New Performance Metric Development", BCP 170, RFC 6390, October 2011.October 2011[RFC6709]RFC 6709Carpenter, B., Aboba, B., and S. Cheshire, "Design Considerations for Protocol Extensions", RFC 6709, September 2012.September 2012[RFC6776]RFC 6776Clark, A. and Q. Wu, "Measurement Identity and Information Reporting Using a Source Description (SDES) Item and an RTCP Extended Report (XR) Block", RFC 6776, October 2012.October 2012[RFC6792]RFC 6792Wu, Q., Hunt, G., and P. Arden, "Guidelines for Use of the RTP Monitoring Framework", RFC 6792, November 2012.November 2012[RFC7003]RFC 7003Clark, A., Huang, R., and Q. Wu, Ed., "RTP Control Protocol (RTCP) Extended Report (XR) Block for Burst/Gap Discard Metric Reporting", RFC 7003, September 2013.September 2013