IP Performance Metrics Working Group A.Morton Internet Draft L.Ciavattone Document: G.Ramachandran Category: Standards Track AT&T Labs S.Shalunov Internet2 J.Perser Spirent Packet Reordering Metric for IPPM Status of this Memo This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026 [1]. 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 made obsolete 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. 1. Abstract This memo defines a simple metric to determine if a network has maintained packet order. It provides motivations for the new metric, suggests a metric definition, and discusses the issues associated with measurement. The memo includes sample metrics to quantify the extent of reordering in several useful dimensions. Some examples of evaluation using the various sample metrics are included. 2. Conventions used in this document 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 [2]. Although RFC 2119 was written with protocols in mind, the key words are used in this document for similar reasons. They are used to ensure the results of measurements from two different implementations are comparable, and to note instances when an implementation could perturb the network. Morton, et al. Standards Track exp. Mar 2003 Page 1 Packet Reordering Metric for IPPM Nov 2002 3. Introduction Ordered delivery is a property of successful packet transfer attempts, where the packet sequence ascends for each arriving packet and there are no backward steps. An explicit sequence number, such as the sending time of each packet or an incrementing message number carried in each packet establishes the Source Sequence. The presence of reordering at the Destination is based on arrival order. This metric is consistent with RFC 2330 [3], and classifies arriving packets with sequence numbers smaller than their predecessors as out-of-order, or reordered. For example, if arriving packets are numbered 1,2,4,5,3, then packet 3 is reordered. This is equivalent to Paxon's reordering definition in [4], where "late" packets were declared reordered. The alternative is to emphasize "premature" packets instead (4 and 5 in the example). The metric's construction is very similar to the sequence space validation for received segments in RFC793 [5]. Earlier work to define ordered delivery includes [6], [7] and more ???. 3.1 Motivation A reordering metric is relevant for most applications, especially when assessing network support for Real-Time media streams. The extent of reordering may be sufficient to cause a received packet to be discarded by functions above the IP layer. Packet order is not expected to change during transfer, but several specific path characteristics can cause their order to change. Examples are: * When two paths, one with slightly longer transfer time, support a single packet stream or flow, then packets traversing the longer path may arrive out-of-order. Multiple paths may be used to achieve load balancing, or may arise from route instability. * To increase capacity, a network device designed with multiple processors serving a single port may reorder as a byproduct. * A layer 2 retransmission protocol that compensates for an error- prone link may cause packet reordering. * If for any reason, the packets in a buffer are not serviced in the order of their arrival, their order will change. * If packets in a flow are assigned to multiple buffers (following evaluation of traffic characteristics, for example), and the buffers have different occupations and/or service rates, then order will likely change. Morton, et al. Standards Track exp. Mar 2002 Page 2 Packet Reordering Metric for IPPM Nov 2002 The ability to restore order at the destination will likely have finite limits. Practical hosts have receiver buffers with finite size in terms of packets, bytes, or time (such as de-jitter buffers). Once the initial determination of reordering is made, it is useful to quantify the extent of reordering, or lateness, in all meaningful dimensions. 3.2 Goals and Objectives The definitions below intend to satisfy the goals of: 1. Determining whether or not packet order is maintained. 2. Quantifying the extent (achieving this second goal requires assumptions of upper layer functions and capabilities to restore order, and therefore several solutions). Reordering Metrics MUST: + be relevant to one or more known applications + be computable "on the fly" + work with Poisson and Periodic test streams + work even if the stream has duplicate or lost packets Reordering Metrics SHOULD: + have concatenating results for segments measured separately + have simplicity for easy consumption and understanding + have relevance to TCP performance + have relevance to Real-time application performance 4. An Ordered Arrival Singleton Metric The IPPM framework RFC 2330 [3] gives the definitions of singletons, samples, and statistics. The evaluation of packet order requires several supporting concepts. The first is a sequence number applied to packets at the source to uniquely identify the order of packet transmission. The sequence number may be established by a simple message number, a byte stream number, or it may be the actual time when each packet departs from the Src. The second supporting concept is a stored value which is the "next expected" packet number. Under normal conditions, the value of Next Expected (NextExp) is the sequence number of the previous packet (plus 1 for message numbering). In byte stream numbering, NextExp is a value 1 byte greater than the last in-order packet sequence number + payload. If Src time is used as the sequence number, NextExp is the Src time from the last in-order packet + 1 clock tick. Morton, et al. Standards Track exp. Mar 2002 Page 3 Packet Reordering Metric for IPPM Nov 2002 Each packet within a packet stream can be evaluated for its order singleton metric. 4.1 Metric Name: Type-P-Non-Reversing-Order 4.2 Metric Parameters: + Src, the IP address of a host + Dst, the IP address of a host + SrcTime, the time of packet emission from the Src (or wire time) + SrcNum, the packet sequence number applied at the Src, in units of messages or bytes. + NextExp, the Next Expected Sequence number at the Dst, in units of messages, time, or bytes. + PayloadSize, the number of bytes contained in the information field and referred to when the SrcNum sequence is based on byte transfer. 4.3 Definition: In-order packets have sequence numbers (or Src times) greater than or equal to the value of Next Expected. Each new in-order packet will increase the Next Expected (typically by 1 for message numbering, or the payload size plus 1 for byte numbering). The Next Expected value cannot decrease, thereby specifying non-reversing order as the basis to identify reordered packets. A reordered packet outcome occurs when a single IP packet at the Dst Measurement Point results in the following: The packet has a Src sequence number lower than the Next Expected (NextExp), and therefore the packet is reordered. The Next Expected value does not change on the arrival of this packet. This definition can also be specified in pseudo-code. On successful arrival of a packet with sequence number n: if n >= NextExp, /* n is in-order */ then NextExp = n + PayloadSize + 1; else /* when n < NextExp */ designate packet n as reordered; When using message-based sequence numbering or Src time, PayloadSize=0. 4.4 Discussion Morton, et al. Standards Track exp. Mar 2002 Page 4 Packet Reordering Metric for IPPM Nov 2002 Any arriving packet bearing a sequence number from the sequence that establishes the Next Expected value can be evaluated to determine if it is in-order, or reordered, based on a previous packet's arrival. In the case where Next Expected is Undefined (because the arriving packet is the first successful transfer), the packet is designated in-order. 5. Sample Metrics It is highly desirable to assert the degree to which a packet is out-of-order, or reordered with respect to a sample of packets. This section defines several metrics that quantify the extent of reordering in various units of measure. Each metric highlights a relevant application. 5.1 n-Reordering [Note: This is the 10/2002 definition of n-Reordering. This definition focuses on TCP sender and receiver behavior, and in particular, New Reno TCP behavior when n=3.] Metric Name: Type-P-packet-n-reordering-Poisson/Periodic-Stream Parameter Notation: Let n be a positive integer (a parameter). Let k be a positive integer (sample size, the number of packets sent). Let l be a non-negative integer representing the number of packets that were received out of the k packets sent. (Note that there is no relationship between k and l: on one hand, losses can make l less than k; on the other hand, duplicates can make l greater than k.) Assign each sent packet a sequence number, 1 to k. Let s[1], ..., s[l] be the original sequence numbers of the received packets, in the order of arrival (duplicates are possible). Definition 1: Received packet number i (n < i <= l) is called n- reordered if and only if for all j such that i-n <= j < i we have s[j] > s[i]. Note: This definition is illustrated by C code in Appendix A. It computes n-reordering for a particular value of n (when actually writing applications that would report the metric, one would probably report it for several values of n, such as 1, 2, 3, 4 -- and maybe a few more consecutive values). Claim: If a packet is n-reordered and 0 < n' < n, then the packet is also n'-reordered. Let m be the number of n-reordered packets in the sample. Definition 2: The degree of n-reordering of the sample is m/(l-n). Morton, et al. Standards Track exp. Mar 2002 Page 5 Packet Reordering Metric for IPPM Nov 2002 Definition 3: The degree of reordering of the sample is its degree of 1-reordering. <<< s[i]. A sample's position offset may be expressed as a histogram, to easily summarize the extent and frequency of various offsets. 5.2.2 Metric Name: Type-P-packet-Late-Time-Poisson/Periodic-Stream Metric Parameters: In addition to the parameters defined for Type-P- Non-Reversing-Order, we specify: + DstTime, the time that each packet in the stream arrives at Dst Definition: Lateness in time is calculated using Dst times. Late Time = DstTime(reordered packet)-DstTime(packet at discontinuity) Using similar notation to that of Section 5.1, an equivalent definition is: The Late Time of Reordered Packet i is t = DstTime[i]-DstTime[j], for min{j|1<=js[i], or SrcTime[j]>SrcTime[i]. 5.2.3 Metric Name: Type-P-packet-Byte-Offset-Poisson/Periodic-Stream Metric Parameters: We use the same parameters defined above. Definition: Byte stream offset is the sum of the payload sizes of all intervening packets between the reordered packet and the discontinuity (including the packet at the discontinuity). Morton, et al. Standards Track exp. Mar 2002 Page 7 Packet Reordering Metric for IPPM Nov 2002 When reordered packet has DstOrder=m Byte Offset = Sum[PayloadSize(packet, DstOrder=m-1), PayloadSize(packet, DstOrder=m-2), ... PayloadSize(packet at discontinuity)] 5.2.4 Discussion The Offset metrics can predict whether reordered packets will be useful in a general, but limited receiver buffer system. The limit may be the number of bytes or packets the buffer can store, or the time of storage prior to a cyclic play-out instant (as with de- jitter buffers). Note that the One-way IPDV [8] gives the delay variation for a packet w.r.t. the preceding packet in the source sequence. Lateness and IPDV give an indication of whether a buffer at Dst has sufficient storage to accommodate the network's behavior and restore order. When an earlier packet in the Src sequence is lost, IPDV will necessarily be undefined for adjacent packets, and Late Time may provide the only way to evaluate the usefulness of a packet. In the case of de-jitter buffers, there are circumstances where the receiver employs loss concealment at the intended play-out time of a late packet. However, if this packet arrives out of order, the Late Time determines whether the packet is still useful. IPDV no longer applies, because the receiver establishes a new play-out schedule with additional buffer delay to accommodate similar events in the future - this requires very minimal processing. When packets in the stream have variable sizes, it may be most useful to characterize Offset in terms of the payload size(s) of stored packets (using byte stream numbering). For a sample of packets in a stream, results may be reported as a ratio of reordered packets to total packets sent by the source during the test. If separate reordering events can be distinguished, then an event count may also be reported (along with the event description, such as the number of reordered packets and their offsets). The distribution of various Offset metrics may also be reported and summarized as average, range, etc. 6. Measurement Issues The results of tests will be dependent on the time interval between measurement packets (both at the Src, and during transport where spacing may change). Clearly, packets launched infrequently (e.g., 1 per 10 seconds) are unlikely to be reordered. Test streams may prefer to use a periodic sending interval so that a known temporal bias is maintained, also bringing simplified results analysis [Ref to npmps]. In this case, the periodic sending interval Morton, et al. Standards Track exp. Mar 2002 Page 8 Packet Reordering Metric for IPPM Nov 2002 should be chosen to reproduce the closest Src packet spacing expected. <<< the received packets are represented as: Morton, et al. Standards Track exp. Mar 2002 Page 10 Packet Reordering Metric for IPPM Nov 2002 \/ s = 1, 2, 3, 5, 6, 7, 8, 4, 9, 10 i = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 /\ when n=1, 7<=J<8, and 8 > 4, so the packet at i=8 is 1-reordered. when n=2, 6<=J<8, and 7 > 4, so the packet at i=8 is 2-reordered. when n=3, 5<=J<8, and 6 > 4, so the packet at i=8 is 3-reordered. when n=4, 4<=J<8, and 5 > 4, so the packet at i=8 is 4-reordered. when n=5, 3<=J<8, but 3 < 4, no more reordering. We note that the Position Offset is equal to the Max(n) with n- reordering. Table 2 Example with Packets 5 and 6 Reordered, Sending order(SrcNum@Src): 1,2,3,4,5,6,7,8,9,10 SrcNum Src Dst Dst Posit. Late @Dst NextExp Time Time Delay IPDV Order Offset Time 1 1 0 68 68 1 2 2 20 88 68 0 2 3 3 40 108 68 0 3 4 4 60 128 68 0 4 7 5 120 188 68 -22 5 5 8 80 189 109 41 6 1 1 6 8 100 190 90 -19 7 2 2 8 8 140 208 68 0 8 9 9 160 228 68 0 9 10 10 180 248 68 0 10 Table 2 shows a case where packets 5 and 6 arrive just behind packet 7, so both 5 and 6 are declared out-of-order. Their positional offsets (6-5=1 and 7-5=2, using DstOrder again) and Late times (189- 188=1, 190-188=2) are small. In the notation of n-reordering, the received packets are represented as: \/ \/ s = 1, 2, 3, 4, 7, 5, 6, 8, 9, 10 i = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 /\ /\ Considering packet 5[6] first: when n=1, 5<=J<6, and 7 > 5, so the packet at i=6 is 1-reordered. when n=2, 4<=J<6, but 4 < 5, same for all earlier packets. Considering packet 6[7] next: when n=1, 6<=J<7, and 5 < 6, so the packet at I=7 is not n-reordered for any n, even though: when N=2, 5<=J<7, and 7 > 6, because n-reordering requires s[j]>s[i] Morton, et al. Standards Track exp. Mar 2002 Page 11 Packet Reordering Metric for IPPM Nov 2002 for all j such that i-n <= j < i (see Definition 1 in section 5.1). A hypothetical sender/receiver pair may retransmit packet 5[8] unnecessarily, since it is 1-reordered (in agreement with the singleton metric). However, the receiver cannot advance packet 7[5] to the higher layers until after packet 6[7] arrives. Therefore, the singleton metric correctly determined that 6[7] is reordered, and the n-reordering metric indicates that the hypothetical receiver can deal with its arrival efficiently (no unnecessary retransmission). Table 3 Example with Packets 4, 5, and 6 reordered Sending order(SrcNum@Src): 1,2,3,4,5,6,7,8,9,10,11 SrcNum Src Dst Dst Posit. Late @Dst NextExp Time Time Delay IPDV Order Offset Time 1 1 0 68 68 1 2 2 20 88 68 0 2 3 3 40 108 68 0 3 7 4 120 188 68 -68 4 8 8 140 208 68 0 5 9 9 160 228 68 0 6 10 10 180 248 68 0 7 4 11 60 250 190 122 8 4 62 5 11 80 252 172 -18 9 5 64 6 11 100 256 156 -16 10 6 68 11 11 200 268 68 0 11 The case in Table 3 is where three packets in sequence have long transit times (packets with SrcNum 4,5,and 6). Delay, Late time, and Position Offset capture this very well, and indicate variation in reordering extent, while IPDV indicates that the spacing between packets 4,5,and 6 has changed. The histogram of Position Offsets would be: Bin 1 2 3 4 5 6 7 Frequency 0 0 0 1 1 1 0 In the notation of n-reordering, the received packets are represented as: s = 1, 2, 3, 7, 8, 9,10, 4, 5, 6, 11 i = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,11 Considering packet 4[8] first: when n=1, 7<=J<8, and 10> 4, so the packet at i=8 is 1-reordered. when n=2, 6<=J<8, and 9 > 4, so the packet at i=8 is 2-reordered. when n=3, 5<=J<8, and 8 > 4, so the packet at i=8 is 3-reordered. when n=4, 4<=J<8, and 7 > 4, so the packet at i=8 is 4-reordered. when n=5, 3<=J<8, but 3 < 4, same for all earlier packets. Considering packet 5[9] next: Morton, et al. Standards Track exp. Mar 2002 Page 12 Packet Reordering Metric for IPPM Nov 2002 when n=1, 8<=J<9, and 4 < 5, so the packet at I=9 is not n-reordered This example shows again that the n-reordering definition identifies a single packet (SrcNum=4) with a sufficient degree of reordering to result in one unnecessary packet retransmission by the New Reno TCP sender. Also, the delayed arrival of SrcNum=5 and SrcNum=6 will allow the receiver process to pass Src packets 7 through 10 up the protocol stack (the singleton metric indicates 5 and 6 are reordered). 8. Security Considerations [mostly borrowed from npmps] 8.1 Denial of Service Attacks This metric requires a stream of packets sent from one host (Src) to another host (Dst) through intervening networks. This method could be abused for denial of service attacks directed at Dst and/or the intervening network(s). Administrators of Src, Dst, and the intervening network(s) should establish bilateral or multi-lateral agreements regarding the timing, size, and frequency of collection of sample metrics. Use of this method in excess of the terms agreed between the participants may be cause for immediate rejection or discard of packets or other escalation procedures defined between the affected parties. 8.2 User data confidentiality Active use of this method generates packets for a sample, rather than taking samples based on user data, and does not threaten user data confidentiality. Passive measurement must restrict attention to the headers of interest. Since user payloads may be temporarily stored for length analysis, suitable precautions MUST be taken to keep this information safe and confidential. 8.3 Interference with the metric It may be possible to identify that a certain packet or stream of packets is part of a sample. With that knowledge at Dst and/or the intervening networks, it is possible to change the processing of the packets (e.g. increasing or decreasing delay) that may distort the measured performance. It may also be possible to generate additional packets that appear to be part of the sample metric. These additional packets are likely to perturb the results of the sample measurement. To discourage the kind of interference mentioned above, packet interference checks, such as cryptographic hash, may be used. 9. IANA Considerations Morton, et al. Standards Track exp. Mar 2002 Page 13 Packet Reordering Metric for IPPM Nov 2002 Since this metric does not define a protocol or well-known values, there are no IANA considerations in this memo. 10. References 1 Bradner, S., "The Internet Standards Process -- Revision 3", BCP 9, RFC 2026, October 1996. 2 Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", RFC 2119, March 1997. 3 Paxson, V., Almes, G., Mahdavi, J., and Mathis, M., "Framework for IP Performance Metrics", RFC 2330, May 1998. 4 V.Paxson, "Measurements and Analysis of End-to-End Internet Dynamics," Ph.D. dissertation, U.C. Berkeley, 1997, ftp://ftp.ee.lbl.gov/papers/vp-thesis/dis.ps.gz. 5 Postel, J., "Transmission Control Protocol", STD 7, RFC 793, September 1981. Obtain via: http://www.rfc-editor.org/rfc/rfc793.txt 6 L.Ciavattone and A.Morton, "Out-of-Sequence Packet Parameter Definition (for Y.1540)", Contribution number T1A1.3/2000-047, October 30, 2000. ftp://ftp.t1.org/pub/t1a1/2000-A13/0a130470.doc 7 J.C.R.Bennett, C.Partridge, and N.Shectman, "Packet Reordering is Not Pathological Network Behavior," IEEE/ACM Transactions on Neteworking, vol.7, no.6, pp.789-798, December 1999. 8 Demichelis, C., and Chimento, P., "IP Packet Delay Variation Metric for IPPM", work in progress. 11. Acknowledgments The authors would like to acknowledge the helpful discussions with Matt Mathis and Jon Bennett. We gratefully acknowledge the foundation laid by the authors of the IP performance Framework [3]. 12. Appendix A (informative) Two example c-code implementations of reordering definitions follow: Example 1 n-reordering ============================================ #include #define MAX_N 100 #define min(a, b) ((a) < (b)? (a): (b)) #define loop(x) ((x) >= 0? x: x + MAX_N) Morton, et al. Standards Track exp. Mar 2002 Page 14 Packet Reordering Metric for IPPM Nov 2002 /* * Read new sequence number and return it. Return a sentinel value of EOF * (at least once) when there are no more sequence numbers. In this example, * the sequence numbers come from stdin; in an actual test, they would come * from the network. */ int read_sequence_number() { int res, rc; rc = scanf("%d\n", &res); if (rc == 1) return res; else return EOF; } int main() { int m[MAX_N]; /* We have m[j-1] == number of * j-reordered packets. */ int ring[MAX_N]; /* Last sequence numbers seen. */ int r = 0; /* Ring pointer for next write. */ int l = 0; /* Number of sequence numbers read. */ int s; /* Last sequence number read. */ int j; for (j = 0; j < MAX_N; j++) m[j] = 0; for (; (s = read_sequence_number()) != EOF; l++, r = (r+1) % MAX_N) { for (j=0; j #define MAX_N 100 #define min(a, b) ((a) < (b)? (a): (b)) #define loop(x) ((x) >= 0? x: x + MAX_N) /* Global counters */ int receive_packets=0; /* number of recieved */ int reorder_packets=0; /* number of reordered packets */ /* function to test if current packet has been reordered * returns 0 = not reordered * 1 = reordered */ int testorder1(int seqnum) // Al { static int NextExp = 1; int iReturn = 0; if (seqnum >= NextExp) { NextExp = seqnum+1; } else { iReturn = 1; } return iReturn; } int testorder2(int seqnum) // Stanislav { static int ring[MAX_N]; /* Last sequence numbers seen. */ static int r = 0; /* Ring pointer for next write. */ int l = 0; /* Number of sequence numbers read. */ int j; int iReturn = 0; l++; r = (r+1) % MAX_N; for (j=0; j Len Ciavattone AT&T Labs Room C4 - 2B29 200 Laurel Ave. South Middletown, NJ 07748 USA Phone +1 732 420 1239 Gomathi Ramachandran AT&T Labs Room C4 - 3D22 200 Laurel Ave. South Middletown, NJ 07748 USA Phone +1 732 420 2353 Stanislav Shalunov Internet2 200 Business Park Drive, Suite 307 Armonk, NY 10504 Phone: + 1 914 765 1182 EMail: Jerry Perser Spirent Communications 26750 Agoura Road Calabasas, CA 91302 USA Phone: + 1 818 676 2300 EMail: Full Copyright Statement Morton, et al. Standards Track exp. Mar 2002 Page 17 Packet Reordering Metric for IPPM Nov 2002 "Copyright (C) The Internet Society (date). All Rights Reserved. This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implmentation may be prepared, copied, published and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this paragraph are included on all such copies and derivative works. However, this document itself may not be modified in any way, such as by removing the copyright notice or references to the Internet Society or other Internet organizations, except as needed for the purpose of developing Internet standards in which case the procedures for copyrights defined in the Internet Standards process must be followed, or as required to translate it into languages other than English. The limited permissions granted above are perpetual and will not be revoked by the Internet Society or its successors or assigns. This document and the information contained herein is provided on an "AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Morton, et al. Standards Track exp. Mar 2002 Page 18