INTERNET-DRAFT Nick Duffield draft-ietf-psamp-framework-02.txt Albert Greenberg March 2003 Matthias Grossglauser Expires: September 2003 Jennifer Rexford AT&T Labs - Research Derek Chiou Avici Systems Benoit Claise Peram Marimuthu Ganesh Sadasivan Cisco Systems A Framework for Passive Packet Measurement Copyright (C) The Internet Society (2003). All Rights Reserved. This document is an Internet-Draft and is in full conformance with all provisions of Section 10 of RFC2026. 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. Abstract A wide range of traffic engineering and troubleshooting tasks rely on timely and detailed traffic measurements that can be consistently interpreted. We describe a framework for passive packet measurement that is (a) general enough to serve as the basis for a wide range of operational tasks, and (b) needs only a small set of packet selection operations that facilitate ubiquitous deployment in router interfaces or dedicated measurement devices, even at very high speeds. Comments on this document should be addressed to the PSAMP WG mailing list: psamp@ops.ietf.org To subscribe: psamp-request@ops.ietf.org, in body: subscribe Archive: https://ops.ietf.org/lists/psamp/ Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 1]
Internet-Draft Passive Packet Measurement March 2003 0 Contents 1 Motivation ................................................. 3 2 Elements, Terminology, and Architecture .................... 4 3 Requirements ............................................... 6 3.1 Selection Process Requirements ......................... 6 3.2 Reporting Process Requirements ......................... 7 3.3 Export Process Requirements ............................ 7 3.4 Configuration Requirements ............................. 7 4 Packet Selection ............................................ 8 4.1 Filtering .............................................. 8 4.2 Systematic Sampling .................................... 8 4.3 Random Sampling ........................................ 8 4.3.1 Uniform Random Sampling ............................ 8 4.3.2 Stratified Random Sampling ......................... 9 4.3.3 Non-uniform Independent Random Sampling ............ 9 4.4 Hash-based Selection ................................... 9 4.3.1 Consistent Flow Sampling ........................... 10 4.3.2 Trajectory Sampling ................................ 10 4.5 Generation of Pseudorandom Variates .................... 11 4.6 Criteria for Choice of Selection Operations ............ 11 4.6.1 Evaluating the Need for Distinct Selection Operations 11 4.6.2 Comparison of Uniform Sampling Methods ............. 12 4.7 Constraints on the Sampling Rate ....................... 12 4.8 Selection According to Packet Treatment ................ 12 4.9 Input Sequence Numbers for Primitive Selection Operations 12 4.10 Selection Operations and Application Requirements ..... 13 4.10.1 Mandatory Selection Operations .................... 13 4.10.2 Recommended Selection Operations .................. 13 4.10.3 Optional Selection Operations ..................... 14 5 Reporting .................................................. 14 5.1 Mandatory Reporting .................................... 14 5.2 Recommended Reporting .................................. 15 5.3 Report Interpretation ................................. 15 6 Export and Congestion Avoidance ............................ 16 6.1 Collector Destination .................................. 16 6.2 Local Export ........................................... 16 6.3 Reliable vs. Unreliable Transport ...................... 16 6.4 Limiting Delay in Exporting Measurement Packets ........ 17 6.5 Configurable Export Rate Limit ......................... 17 6.6 Congestion-aware Unreliable Transport .................. 17 6.7 Collector-based Rate Reconfiguration ................... 18 6.7.1 Changing the Export Rate and Other Rates ........... 18 6.7.2 Notions of Fairness ................................ 18 6.7.3 Behavior Under Overload and Failure ................ 19 7 Parallel Measurement Processes ............................. 19 8 Configuration and Management ............................... 19 9 Feasibility and Complexity ................................. 20 9.1 Feasibility ............ ............................... 20 9.1.1 Filtering .......................................... 20 9.1.2 Sampling ........................................... 20 9.1.3 Hashing ............................................ 20 Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 2]
Internet-Draft Passive Packet Measurement March 2003 9.1.4 Reporting .......................................... 20 9.1.5 Export ............................................. 21 9.2 Potential Hardware Complexity .......................... 21 10 Applications .............................................. 22 10.1 Baseline Measurement and Drill Down ................... 22 10.2 Passive Customer Performance Measurements ............. 23 10.3 Troubleshooting ....................................... 23 11 References ................................................ 24 12 Authors' Addresses ........................................ 25 13 Intellectual Property Statement ........................... 26 14 Full Copyright Statement .................................. 27 1 Motivation This document describes a framework in which to define a standard set of capabilities for network elements to sample subsets of packets by statistical and other methods. The framework will accommodate future work to (i) specify a set of selection operations by which packets are sampled (ii) specify the information that is to be made available for reporting on sampled packets; (iii) describe a protocol by which information on sampled packets is reported to applications; (iv) describe a protocol by which packet selection and reporting are configured. The motivation to standardize these capabilities comes from the need for measurement-based support for network management and control across multivendor domains. This requires domain wide consistency in the types of selection schemes available, the manner in which the resulting measurements are presented, and consequently, consistency of the interpretation that can be put on them. The capabilities are positioned as suppliers of packet samples to higher level consumers, including both remote collectors and applications, and on board measurement-based applications. Indeed, development of the standards within the framework described here should be open to influence by the requirements of standards in related IETF WGs, for example, IP Performance Metrics (IPPM) [PAMM98] and Internet Traffic Engineering (TEWG) [LCTV02]. Conversely, we expect that aspects of this framework not specifically concerned with the central issue of packet selection and report formation may be able to leverage work in other WGs. Potential examples are the format and export of measurement Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 3]
Internet-Draft Passive Packet Measurement March 2003 reports, which may leverage the information model and export protocols of IP Flow Information Export (IPFIX) [QZCZCN02], and work in congestion aware unreliable transport in the Datagram Congestion Control Protocol (DCCP) [FHK02]. 2 Elements, Terminology, and Architecture This section defines the basic elements of the PSAMP framework. * PSAMP Device: a device hosting at least one of each of the following: an observation point, a measurement process, and an export process. * Observation Point: The observation point is a location in the network where packets can be observed. Examples are, a line to which a probe is attached, a shared medium, such as an Ethernet-based LAN, a single port of a router, or set of interfaces (physical or logical) of a router, an embedded measurement subsystem within an interface. * Measurement Process: the combination of a selection process followed by a reporting process. * Selection Process: A selection process selects packets for reporting at an observation point. The inputs to the selection process are the packets observed at the observation point (including packet encapsulation headers), information derived from the packets' treatment at the observation point, and selection state that may be maintained by the observation point. Selection is accomplished through operating on these inputs with one or more selection operations. * Selection Operation: A configurable packet selection operation. It takes as input the selection process input for a single packet. If the packet is selected, this same information may be considered as the output. Selection operations may change the selection state. * Selection State: the observation point may maintain state information for use by the reporting process, and/or by multiple selection operations, either on the same packet, or on different packets. Examples include sequence numbers of packets at the input of packet selectors, timestamps, iterators for pseudorandom number generators, calculated hash values, and indicators of whether a packet was selected by a given selection operation. * Composite Selection Operation: a selection operation that is expressed as an ordered composition of other selection operations. Thus a packet is selected by the composite operation if it is selected by all its constituent selection Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 4]
Internet-Draft Passive Packet Measurement March 2003 operations in order. * Primitive Selection Operation: a selection operation that is not a composite of other selection operations. * Reporting Process: the creation of a report stream of information on packets selected by a selection processes, in preparation for export. The input to a reporting process comprises that information available to a selection process, for the selected packets. The report stream contains two distinguished types of information: packet reports, and report interpretation. * Packet Reports: a configurable subset of the per packet input to the reporting process. * Report Interpretation: subsidiary information relating to one or more packets, that is used for interpretation of their packet reports. Examples include configuration parameters of the PSAMP device, and configuration parameters of the selection and reporting process. * Export Process: sends the output of one or more reporting process from the PSAMP device to one or more collectors. * Collector: a collector receives a report stream exported by one or more measurement processes. In some cases, the PSAMP device may serve as the collector. * Measurement packets: one or packet reports, and perhaps report interpretation, are bundled by the export process into a measurement packet for export to a collector. The various possibilities for the high level architecture of these elements is as follows. Note in the last case: the PSAMP device may also be a collector. OP = Observation Point, MP = Measurement Process, EP = Export Process +---------------------+ +------------------+ |PSAMP Device(1) | | Collector(1) | |Observation Point(s) | | | |MP(s)--->EP----------+---------------->| | |MP(s)--->EP----------+-------+-------->| | +---------------------+ | +------------------+ | +---------------------+ | +------------------+ |PSAMP Device(2) | +-------->| Collector(2) | |Observation Point(s) | | | |MP(s)--->EP----------+---------------->| | +---------------------+ +------------------+ +---------------------+ Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 5]
Internet-Draft Passive Packet Measurement March 2003 |PSAMP Device(3) | |Observation Point(s) | |MP(s)--->EP---+ | | | | |Collector(3)<-+ | +---------------------+ 3 Requirements 3.1 Selection Process Requirements. * Ubiquity: The selection operations must be simple enough to be implemented ubiquitously at maximal line rate. * Applicability: the set of selection operations must be rich enough to support a range of existing and emerging measurement based applications and protocols. This requires a workable trade-off between the range of traffic engineering applications and operational tasks it enables, and the complexity of the set of capabilities. * Extensibility: to allow for additional packet selection operations to support future applications. * Flexibility: to support selection of packets using different network protocols or encapsulation layers (e.g. IPv4, IPv6, MPLS, etc), and under packet encryption. * Visibility: robustness of packet selection w.r.t. attempts to evade measurement. * Parallel measurements: support multiple independent measurement processes at the same device. * Non-contingency: in order to satisfy the ubiquity requirement, the selection decision for each packet must not depend on future packets. Rather, the selection decision must be capable of being made on the basis of the selection process input up to and including the packet in question. This excludes selection functions that require caching of packet for selection contingent on subsequent packets. See also the timeliness requirement following. A range of candidate selection operations is given in Section 4. Some detailed requirements of all selection operations are given in Section 4.9. Those selection operations to be required by the PSAMP standard are described in Section 4.10. Parallel measurement processes are discussed in Section 8. A target set of applications for PSAMP to support are described in Section 10. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 6]
Internet-Draft Passive Packet Measurement March 2003 3.2 Reporting Process Requirements * Timeliness: reports on selected packets should be made available to the collector quickly enough to support near real time applications. * Transparency: allow transparent interpretation of measurements as communicated by PSAMP reporting, without need to obtain additional information from the measuring device. * Robustness: allow robust interpretation of measurements with respect to reports missing due to loss, e.g. in transport, or omission at the measurement device. Inclusion in reporting of information enabling accuracy of measurements to be determined. * Faithfulness: all reported quantities that relate to the packet treatment must reflect the router state and configuration encountered by the packet in the PSAMP device. * Privacy: selection of the content of packet reports will be cognizant of privacy and anonymity issues while being responsive to the needs of measurement applications, and in accordance with RFC 2804. Full packet capture of arbitrary packet streams is explicitly out of scope. A specific reporting processes meeting these requirements, and the requirement for ubiquity, is described in Section 5. 3.3 Export Process Requirements * Congestion Avoidance: export of a report stream across a network must be congestion avoiding in compliance with RFC 2914. * Secure Export: the ability to export securely, e.g. by encryption Candidate export processes meeting these requirements are described in Section 6. 3.4 Configuration Requirements * Ease of Configuration: of sampling and export parameters, e.g. for automated remote reconfiguration in response to measurements. * Secure Configuration: configuration via protocols that prevent unauthorized reconfiguration. Specific configuration capabilities that meet these requirements are discussed in Section 8. Feasibility and complexity of PSAMP operations is discussed in Section 9. Reuse of existing protocols will be encouraged provided the Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 7]
Internet-Draft Passive Packet Measurement March 2003 protocol capabilities are compatible with the requirements laid out in this section. 4 Packet Selection The function of packet selection is to select a subset from the stream of all packets visible at an observation point. Selection can be used to select packets of based on their content, and/or to reduce the rate of packets reports regardless of content. This section details some candidate primitive selection operations for standardization that satisfy the requirements of Section 3.1. Not all operations listed here are intended for standardization. Those that are are listed in Section 4.10. Packet selection techniques are discussed in more detail in [ZMR03]. 4.1 Filtering Filtering is the selection of packets based only the packet content, the treatment of the packet at the observation point, and deterministic functions of these occurring in the selection state. The packet is selection if these quantities fall into a specified range. Hash-based packet selection (see Section 4.3) can also be regarded as a filter) An example is a match/mask filter applied to a combination of bit positions. The packet is selected if the bits and the match are equal after taking the logical AND of both with the mask. Higher level interfaces may be used to specify mask and matches for particular fields, for example, for IP addresses. Filtering on information derived from packet treatment, e.g., AS numbers derived from routing state, is another possibility; see Section 4.8. Filtering based on calculated hashes is described separately in Section 4.4. 4.2 Systematic Sampling In systematic sampling, the triggers for sampling are periodic, either in time or in packet count. All packets occurring in a selection interval (either in time or packet count) beyond the trigger are selected. The case that the selection interval covers only the first available packet for count-based sampling is often called 1 in N sampling: packets are selected with count period N. More generally, some number M<N of consecutive packets are selected. 4.3 Random Sampling 4.3.1 Uniform Random Sampling Packets are selected according to a probabilistic law. In the first instance, we consider Independent Uniform Random Sampling: packets are selected independently with some uniform probability 1/N. This count-driven sampling is sometimes referred to a geometric random Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 8]
Internet-Draft Passive Packet Measurement March 2003 sampling, since the difference in count between successive selected packets are independent random variables with a geometric distribution of mean N. A time-driven analog, exponential random sampling, has the time between triggers exponentially distributed. Both geometric and exponential random sampling are examples of what is known as additive random sampling, defined as sampling where the intervals or counts between successive samples are independent identically distributed random variable. 4.3.2 Stratified Random Sampling Generally, in stratified random sampling, packets are assigned to strata according to an attribute, then a number of elements are drawn randomly from each stratum. Stratification reduces variance of single packet statistics if the variance between strata is greater than the variance within strata. In uniform stratified sampling, the number of elements in each stratum is the same, as is the number selected from each stratum. Thus each packet has the same selection probability, but some combination selections are disallowed. With the non-contingency requirement on sampling, the only allowed uniform stratification is that based on packet count. Each group of N successive packets forms a stratum, then some number M<N of each are drawn at random. This is non-contingent because the random positions of the selected packets can be generated in advance for each stratum. 4.3.3 Non-Uniform Independent Random Sampling Also known as non-uniform probability sampling, or content dependent sampling, this is a variant of independent random sampling in which the sampling probabilities can depend on the selection process input. This can be used to weight sampling probabilities in order e.g. to boost the chance of sampling packets that are rare but are deemed important. Unbiased estimators for quantitative statistics are recovered by renormalization of sample values; see [HT52]. 4.4 Hash-based Selection Hash-based selection offers both a way to emulate random selection by generating from the packets' content pseudorandom variates on which to make packet selection decisions, while consistently select subsets of packets that share a common property. A hash function h operates on the selection function input for a packet, and maps in onto a range R. The packet is selected if the resulting hash falls in a specified range S. Thus hashing is a filter. But for good hash functions (see below) the inverse image inv(h(S)) will be extremely complex, and hence h would not be expressible as, say, a match/mask filter or a simple combination of these. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 9]
Internet-Draft Passive Packet Measurement March 2003 A hash function should have good mixing properties, in that changing one bit of the input should change many bits of the output. Then the distribution of hashes will be be fairly uniform, independent of the distribution of the input. The sampling rate is then #S/#R. If the input comprises distinct packet fields, c1 ... cm, selection decisions will appear uncorrelated with the contents of any individual field, if the complementary fields are sufficiently variable and uncorrelated with cj. Even with a publicly known hash-function, accidental or deliberate evasion (or overwhelming) of hash-based sampling is militated against by keeping the selection range private, and/or employing a parameterizable hash function and keeping the parameter private. 4.4.1 Consistent Flow Sampling Hash-based sampling can be used to select all packets from a pseudorandom set of flows. The flow key of each packet is hash sampled, and selected packets are reported on. All packets with a given key are either selected or not selected together. 4.4.2 Trajectory Sampling In trajectory sampling, PSAMP devices in a network hash-sample packets using identical hash function and selection range. The domain of the hash is restricted to those fields that are invariant from hop to hop. Fields such as Time-to-Live, which is decremented per hop, and header CRC, which is recalculated per hop, are thus excluded from the hash domain. Thus a given packet is selected at all either all points on its path through the network, or at none. The domain of the hash function needs to be wider than just a flow key, if packets are to be selected pseudorandomly within flows; see [DuGr01]. A report on each selected packet is exported to a collector. The collector can reconstruct trajectories of the selected packets provided it can match different reports on the same packet, and distinguish these from reports on different packets. For this purpose, reports may also contain a second distinct hash, the identification hash, of the selected packets and/or timing information. The identification hash can be considered as part of the selection state. Applications of trajectory sampling include (i) estimation of the network path matrix, i.e., the traffic intensities accordng to network path, broken down by flow; (ii) detection of routing loops, as indicated by self-intersecting trajectories; (iii) passive performance measurement: prematurely terminating trajectories indicate packet loss, and packet latencies can be determined if reports include (synchronized) timestamps; (iv) network attack tracing, of the actual paths taken by attack packets with spoofed source addresses. Applications are discussed in Section 10. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 10]
Internet-Draft Passive Packet Measurement March 2003 4.5 Generation of Pseudorandom Variates Although pseudorandom number generators with well understood properties have been developed, they may not be the method of choice in setting where computational resources are scarce. A convenient alternative is to use packet content as a source of randomness. Hash-based sampling is an example: the hash (suitably renormalized) is a pseudorandom variate in the interval [0,1]. Other schemes may use packet fields in iterators for pseudorandom numbers. The point here, is that the statistical properties of the idealized packet selection law (such as independence of sampling decisions for different packets, or independence on packet content) may not be exactly shared by an implementation, but only approximately so. Although the selection decisions for non-uniform independent random sampling (see Section 4.3.3 above) also depend on the packet content, this form of sampling is distinguished from the use of packet content to generate variates. In the former case, the content only determines the selection probabilities: selection could then proceed e.g by use of a variates obtained by an independent pseudorandom number generator. In the latter case, the content determines the pseudorandom variates rather than the probabilities. 4.6 Criteria for Choice of Selection Operations 4.6.1 Evaluating the Need for Distinct Selection Operations In current practice, sampling has been performed using particular algorithms, including: - pseudorandom independent sampling with probability 1/N; - systematic sampling of every Nth packet. The question arises as to whether both of these should be standardized as distinct selection operations, or whether they can be regarded as different implementations of a single selection operation. To determine the answer to this question, we need to consider (a) measured or assumed statistical properties of the packet stream, e.g., one or more of the following: - contents of different packets are statistically independent - correlations between contents of different packets decay at a specified rate - contents of certain fields within the same packet are significantly variable and exhibit small cross correlation (b) the desired reference sampling model, e.g., one of: - sample packets with long term probability 1/N - sample packets independent with probability 1/N (c) the set of possible alternatives and implementations, e.g., one of: Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 11]
Internet-Draft Passive Packet Measurement March 2003 - pseudorandom independent sampling with probability 1/N - systematic sampling with period N - hash-based sampling with target probability 1/N (d) the tolerance for error in the applications that use the measurements. We can say that a given alternative from (c) reproduces a reference model (b) for the applications if the results obtained using them are sufficiently accurate in (d) for traffic satisfying an assumed statistical properties in (a). Clearly, application to evaluate methods in (c) requires developing agreement on the relevant properties in (a), (b) and (d). Example: systematic sampling with period N will not count the occurrence of closely space packets (less than N counts apart) from the same flow. Thus for applications that are concerned with the joint statistics of multiple packets within flows, systematic sampling may not reproduce the results obtained with random sampling sufficiently accurately. 4.6.2 Comparison of Uniform Sampling Methods A comparison of sampling methods, and their accuracy for estimating single packet statistics (e.g. mean and distribution of packet length) has been performed in [CPB93]. It was found that estimation using count-based methods was uniformly more accurate than that using time-based methods. Time based methods were found to be particularly inaccurate for assessing interarrival times, since they may miss traffic bursts. There was comparatively little difference between the systematic, stratified and random sampling, at least for the single packet statistics examined. For this reasons, we believe PSAMP should focus on count-based methods. Amongst these, accuracy of single packet statistics is not a great deciding factor. Systematic and random sampling are easier to implement than stratified sampling. 4.7. Constraints on the Sampling Rate Sampling at full line rate, i.e. with probability 1, is not excluded in principle, although resource constraints may not support it in practice. 4.8 Selection According to Packet Treatment Router architectural considerations may preclude some information concerning the packet treatment, e.g routing state, being available at line rate for selection of packets. However, if selection not based on routing state has reduced down from line rate, subselection based on routing state may be feasible. 4.9 Input sequence numbers for primitive selection operations. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 12]
Internet-Draft Passive Packet Measurement March 2003 Each instance of a primitive selection operation MUST maintain a count of packets presented at its input. The counter value is to be included as a sequence number for selected packets. This enables applications to determine the attained rate at which packets are selected, and hence correctly normalize network usage estimates regardless of loss of information, whether this occurs because of discard of packet reports in the PSAMP device, or loss of measurement packets in transmission or collection; see [PPM01]. The sequence numbers are considered as part of the packet's selection state. 4.10 Standardized Selection Operations and Application Requirements In this section we list selection operations in three categories: mandatory (MUST), recommended (SHOULD) and optional (MAY). In each case, we list the applications (described in Section 10) that are supported. 4.10.1 Mandatory Selection Operations PSAMP devices MUST support the following: (i) either count-based 1 in N systematic sampling or count-based simple random sampling. Either operation supports widespread baseline measurement. They reflect current practice: both exist in various routers that are currently available. Furthermore, initial implementation of PSAMP may be constrained to be implemented in software, for devices whose hardware was not designed with PSAMP in mind. It is reasonable that such devices be PSAMP compliant by offering (i) above. 4.10.2 Recommended Selection Operations PSAMP devices SHOULD support the following: (ii) both options in 4.10.1(i) above. (iii) general count-based systematic sampling. (iv) The PSAMP device identifies packet fields relating to different protocols, including IPv4, IPv6, MPLS and AToM. (v) filtering by match/mask, or by single numerical range, on all/any of the fields identified in (iv), together with packet treatment, if any, performed by the PSAMP device. (vi) hash-based selection on a configurable input comprising any/all of the fields identified in (iv), using one or more hash functions to be standardized. (vii) composite sampling operations comprising filtering and any Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 13]
Internet-Draft Passive Packet Measurement March 2003 other selection operation, configurably composed in either order. (viii) at least two parallel independent selection processes, one or both of which can be composite. Filtering, and its combination with sampling, support drill down analysis. Having at least two parallel independent selection processes (and their associated reporting processes) allows drill down to be simultaneously performed with baseline measurements. Hash-based sampling supports trajectory sampling. The operations (ii)-(viii) are listed as recommended, rather than mandatory, in recognition of the fact that some PSAMP devices e.g. simple switches or hubs, may not have the native capabilities to provide detailed protocol information, or perform computations for any selection other than sampling. Conversely, when a PSAMP device is, in its usual capacity, capable of recognizing use of a given protocol in the packet, then the protocol fields MUST be identifiable in (iv) above, and the contents made available for input to filters and hash-based sampling. Thus, for example, a device which routes at the IP level, must make IP field available for filtering. Similarly, any information relating to packet treatment performed by the device MUST be made available for filtering. Thus if the device is a router, routing state MUST be made available for filtering. (See also Section 4.8). 4.10.3 Optional Selection Operations PSAMP devices MAY support the following: (i) non-uniform independent random sampling, based on fields identified in 4.10.2(iv). 5 Reporting Information eligible for inclusion in packet reports includes (i) the packet content itself (including encapsulating headers), but not the requiring the full packet payload; (ii) information relating to the packet treatment: incoming and outgoing interfaces, subinterfaces and channel identifiers, routing state applied to or derived from the packet e.g. next hop IP address, routing prefixes, source and destination AS numbers; (iii) selection state associated with the packet, e.g. timestamps, sequence numbers, hash values. 5.1 Mandatory Reporting All PSAMP devices MUST report the following for each selected packet: Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 14]
Internet-Draft Passive Packet Measurement March 2003 (i) identifiers for the input and output interfaces of the PSAMP device that were traversed by the packet. (ii) the input sequence number(s) of any elementary selection operation(s) that acted on the packet. (iii) some number of contiguous bytes from the start of the packet. The motivation is that some devices may not have the resource capacity or functionality to identify fields within a packet. The burden of interpretation is placed on the collector or applications that it supplies. 5.2 Recommended Reporting PSAMP devices SHOULD report as follows for each selected packet: (iv) identification of packet fields relating to different protocols, including IPv4, IPv6, MPLS and AToM. (v) configurable inclusion of any/all fields from (iv) in the packet report, together with information from packet treatment if present. (vi) selection state associated with the packet, including timestamps and hashes calculated. Similar considerations apply as those of Section 4.10.2: if a device has the native capability to recognize protocol fields and/or treat packets, the the field contents and packet treatment MUST be made available for reporting. 5.3 Report Interpretation Information for use in report interpretation includes (i) configuration parameters of the selectors of the packets reported on; (ii) format of the packet reports (iii) configuration parameters and state information of the network element; (iv) indication of the inherent accuracy of the reported quantities, e.g., of timestamps. The requirements for robustness and transparency are motivations for including report interpretation in the report stream. Inclusion makes the report stream self-defining. The PSAMP framework excludes reliance on an alternative model in which interpretation is recovered out of band. This latter approach is not robust with respect to undocumented changes in selection configuration, and leaves an architectural hostage for network management systems to coherently manage both configuration and data collection. It is not envisaged that all report interpretation must be included Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 15]
Internet-Draft Passive Packet Measurement March 2003 in every packet report. Many of the quantities listed above are expected to be relatively static; they could be communicated periodically, and upon change. To conserve network bandwidth and resources at the collector, the PSAMP device may compress the measurement packets before export. Compression should be quite effective since the sampled packets may share many fields in common, e.g. if a filter focuses on packets with certain values in particular header fields. Using compression, however, could impact the timeliness of reports. Any consequent delay should not violate the timeliness requirement for availability of packet reports at the collector. 6 Export and Congestion Avoidance 6.1 Collector Destination When exporting to a remote collector, the collector is identified by IP address and port number. 6.2 Local Export The report stream may be directly exported to on-board measurement based applications, for example those that for composite statistics from more than one packet. Local export may be presented through an interface direct to the higher level applications, i.e., through an API, rather than employing the transport used for off-board export. A possible example of the local export could be that the selected packets from the PSAMP measurement process serve as the input for the IPFIX protocol, i.e. delivering flow records out of the packets selected out via PSAMP. Note that IPFIX being still developed, this is just listed as a possible example. 6.3 Reliable vs. Unreliable Transport The export of the report stream does not require reliable export. On the contrary, retransmission of lost measurement packets consumes additional network resources and require maintenance of state by the export process. The PSAMP device would have to be able to receive and process acknowledgments, and to store unacknowledged data. Furthermore, the PSAMP device may not possess its own network address (for example an embedded measurement subsystem in an interface) at which to receive acknowledgments. These requirements would be a significant impediment to having ubiquitous support PSAMP. Instead, it is proposed that PSAMP devices support an unreliable export mechanism. Sequence numbers on the measurement packets would indicate when loss has occurred, and the analysis of the collected measurement data can account for this loss. In some sense, packet loss becomes another form of sampling (albeit a less desirable, and Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 16]
Internet-Draft Passive Packet Measurement March 2003 less controlled, form of sampling). 6.4 Limiting Delay in Exporting Measurement Packets The device may queue the report stream in order to export multiple reports in a single measurement packet. Any consequent delay should still allow for timely availability of packet reports at the collector. 6.5 Configurable Export Rate Limit The export process must be able to limit its export rate; otherwise it could overload the network and/or the collector. Note this problem would be exacerbated if using reliable transport mode, since the PSAMP device would retransmit any lost packets, thereby imposing an additional load on the network. At times, the device may generate new packet reports faster than the allowed export rate. In this situation, the device should discard the excess reports rather than transmitting them to the collector. Sequence numbers reported for selector input enable correction for lost reports. An additional sequence number for dispatched measurement packets enables the collector to determine the degree of loss in transmission. There are two options for a configurable rate limit. First, if the transport protocol has a configurable rate limit, that can be used. The second option is to limit the rate at which measurement packets are supplied to the transport protocol. A candidate for implementation of rate limiting is the leaky bucket, with tokens corresponding e.g. to bytes or packets. The export rate limit must be configurable per export process. Note that since congestion loss can occur at any link on the export path, it is not sufficient to limit rate simply as a function of the bandwidth of the interface out of which export takes place. 6.6 Congestion-aware Unreliable Transport Exported measurement traffic competes for resources with other Internet transfers. Congestion-aware export is important to ensure that the measurement packets do not overwhelm the capacity of the network or unduly degrade the performance of other applications, while making good use of available bandwidth resources. The PSAMP WG will evaluate (at least) the following alternatives for congestion aware unreliable transport: (i) protocols under development, including the Datagram Congestion Control Protocol (DCCP); see [FHK02] (ii) protocols adopted in the future by the IPFIX WG, (iii) collector-based rate reconfiguration, as now described. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 17]
Internet-Draft Passive Packet Measurement March 2003 6.7 Collector-based Rate Reconfiguration Since collector-based rate reconfiguration is a new proposal, this draft will discuss it in some detail. The collector can detect congestion loss along the path from the PSAMP device through lost packets, manifest as gaps in the sequence numbers, or the absence of packets for a period of time. The server can run an appropriate congestion-control algorithm to compute a new export rate limit, then reconfigure the PSAMP device with the new rate. This is an attractive alternative to requiring the PSAMP device to receive acknowledgment packets. Implementing the congestion control algorithm in the collector has the added advantages of flexibility in adapting the sending rate and the ability to incorporate new congestion-control algorithms as they become available. 6.7.1 Changing the Export Rate and Other Rates Forcing the PSAMP device to discard excess reports is an effective control under short term congestion. Alternatively, the device could be reconfigured to select fewer packets, and/or send smaller reports on each selected packet. This may be a more appropriate reaction to long-term congestion. In some cases, a collector may receive measurement reports from more than one device, and could decide to reduce the export or other rates at one device rather than another, in order to prioritize the measurement data. This type of flexibility is valuable for network operators that collect measurement data from multiple locations to drive multiple applications. 6.7.2 Notions of Fairness In some cases, it may be reasonable to allow the collector to have flexibility in deciding how aggressively to respond to congestion. For example, the PSAMP device and the collector may have a very small round-trip time relative to other traffic. Conventional TCP-friendly congestion control would allocate a very large share of the bandwidth to this traffic. Instead, the collector could apply an algorithm that reacts more aggressively to congestion to give a larger share of the bandwidth to other traffic (with larger RTTs). In other cases, the measurement packets may require a larger share of the bandwidth than other flows. For example, consider a link that carries tens of thousands of flows, including some non TCP-friendly DoS attack traffic. Restricting the PSAMP traffic to a fair share allocation may be too restrictive, and might limit the collection of the data necessary to diagnose the DoS attack which overloads links over which measurement packets are carried. In order to maintain report collection during periods of congestion, Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 18]
Internet-Draft Passive Packet Measurement March 2003 PSAMP report streams may claim more than a fair share of link bandwidth, provided the number of report streams in competition with fair sharing traffic is limited. The collector could also employ policies that allocate bandwidth in certain proportions amongst different measurement processes. 6.7.3 Behavior Under Overload and Failure The congestion control algorithm has to be robust to severe overload or complete loss of connectivity between the PSAMP device and the collector, and also to the failure of the device or the collector. For example, in a scenario where the collector is unable to reconfigure the export rate because of loss of reverse (collector to PSAMP device) connectivity, it is desirable that the device reduce the export rate automatically. Similarly, if no measurement reports reach the collector because of loss of forward connectivity, the collector should not react to this by increasing the export rate. This problem may be solved through periodic heartbeat packets in both directions (i.e., measurement reports in the forward direction, configuration refresh messages in the reverse direction). This allows each side to detect a loss in connectivity or outright failure and to react appropriately. 7 Parallel Measurement Processes Because of the increasing number of distinct measurement applications, with varying requirements, it is desirable to set up parallel measurement processes on a stream of packets. Each process should consist of independently-configurable selection, reporting and export processes. Each of the parallel measurement processes should be, as far as possible, independent. However, resource constraints may prevent complete reporting on a packet selected by multiple selection processes. In this case, reporting for the packet must be complete for at least one measurement process; other measurement processes need only report that they selected the packet. The priority amongst measurement processes to report packets must be configurable. It is not proposed to standardize the number of parallel measurement processes available beyond the recommendation of Section 4.10.2. 8 Configuration and Management A key requirement for PSAMP is the easy reconfiguration of the parameters of the measurement process: those for selection, packet reports and export. Examples are (i) support of measurement based applications that want to drill-down on traffic detail in real-time; (ii) collector-based rate reconfiguration. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 19]
Internet-Draft Passive Packet Measurement March 2003 To facilitate reconfiguration and retrieval of parameters, they are to reside in a Management Information Base (MIB). CLI and SNMP access to these parameters must be available. 9 Feasibility and Complexity In order for PSAMP to be supported across the entire spectrum of networking equipment, it must be simple and inexpensive to implement. One can envision easy-to-implement instances of the mechanisms described within this draft. Thus, for that subset of instances, it should be straightforward for virtually all system vendors to include them within their products. Indeed, sampling and filtering operations are already realized in available equipment. Here we give some specific arguments to demonstrate feasibility and comment on the complexity of hardware implementations. We stress here that the point of these arguments is not to favor or recommend any particular implementation, or to suggest a path for standardization, but rather to demonstrate that the set of possible implementations is not empty. 9.1 Feasibility 9.1.1 Filtering Filtering consists of a small number of mask (bit-wise logical), comparison and range (greater than) operations. Implementation of at least a small number of such operations is straightforward. For example, filters for security access control lists (ACLs) are widely implemented. This could be as simple as an exact match on certain fields, or involve more complex comparisons and ranges. 9.1.2 Sampling Sampling based on either counters (counter set, decrement, test for equal to zero) or range matching on the hash of a packet (greater than) is possible given a small number of selectors, although there may be some differences in ease of implementation for hardware vs. software platforms. 9.1.3 Hashing Hashing functions vary greatly in complexity. Execution of a small number of sufficient simple hash functions is implementable at line rate. Concerning the input to the hash function, hop-invariant IP header fields (IP address, identification) and TCP/UDP header fields (port numbers, TCP sequence number) drawn from the first 40 bytes of the packet have been found to possess a considerable variability; see [DuGr01]. 9.1.4 Reporting Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 20]
Internet-Draft Passive Packet Measurement March 2003 The simplest packet report would duplicate the first n bytes of the packet. However, such an uncompressed format may tax the bandwidth capabilities of the PSAMP device for high sampling rates; reporting selected fields would save on bandwidth within the PSAMP device. Thus there is a trade-off between simplicity and bandwidth limitations within the PSAMP device. 9.1.5 Export Ease of exporting measurement packets depends on the system architecture. Most systems should be able to support export by insertion of measurement packets, even through the software path. 9.2 Potential Hardware Complexity We now comment on the complexity of possible hardware implementations. Achieving low constants for performance while minimizing hardware resources is, of course, a challenge, especially at very high clock frequencies. Most of these operations, however, are very basic and their implementations very well understood; in fact, the average ASIC designer simply uses canned library instances of these operations rather than design them from scratch. In addition, networking equipment generally does not need to run at the fastest clock rates, further reducing the effort required to get reasonably efficient implementations. Simple bit-wise logical operations are easy to implement in hardware. Such operations (NAND/NOR/XNOR/NOT) directly translate to four-transistor gates. Each bit of a multiple-bit logical operation is completely independent and thus can be performed in parallel incurring no additional performance cost above a single bit operation. Comparisons (EQ/NEQ) take O(lg(M)) stages of logic, where M is the number of bits involved in the comparison. The lg(M) is required to accumulate the result into a single bit. Greater than operations, as used to determine whether a hash falls in a selection range, are a determination of the most significant not-equivalent bit in the two operands. The operand with that most-significant-not-equal bit set to be one is greater than the other. Thus, a greater than operation is also an O(lg(M)) stages of logic operation. Optimized implementations of arithmetic operations are also O(lg(M)) due to propagation of the carry bit. Setting a counter is simply loading a register with a state. Such an operation is simple and fast O(1). Incrementing or decrementing a counter is a read, followed by an arithmetic operation followed by a store. Making the register dual-ported does take additional space, but it is a well-understood technique. Thus, the Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 21]
Internet-Draft Passive Packet Measurement March 2003 increment/decrement is also an O(lg(M)) operation. Hashing functions come in a variety of forms. The computation involved in a standard Cyclic Redundancy Code (CRC) for example are essentially a set of XOR operations, where the intermediate result is stored and XORed with the next chunk of data. There are only O(1) operations and no log complexity operations. Thus, a simple hash function, such as CRC or generalizations thereof, can be implemented in hardware very efficiently. At the other end of the range of complexity, the MD5 function uses a large number of bit-wise conditional operations and arithmetic operations. The former are O(1) operations and the latter are O(lg(M)). MD5 specifies 256 32b ADD operations per 16B of input processed. Consider processing 10Gb/sec at 100MHz (this processing rate appears to be currently available). This requires processing 12.5B/cycle, and hence at least 200 adders, a sizeable number. Because of data dependencies within the MD5 algorithm, the adders cannot be simply run in parallel, thus requiring either faster clock rates and/or more advanced architectures. Thus selection hashing functions as complex as MD5 may be precluded from ubiquitous use at full line rate. This motivates exploring the use of selection hash functions with complexity somewhere between that of MD5 and CRC. However, identification hashing with MD5 on only selected packets is feasible at a sufficiently low sampling rate. 10 Applications We first describe several representative operational applications that require traffic measurements at various levels of temporal and spatial granularity enabled by a PSAMP device. Some of the goals here appear similar to those of IPFIX, at least in the broad classes of applications supported. However, there are two major differences: - PSAMP aims for ubiquitous deployment, thus offering broader reach for existing applications. - PSAMP can support new applications. 10.1 Baseline Measurement and Drill Down Packet sampling is ideally suited to determine the composition of the traffic across a network. The approach is to enable measurement on a cut-set of the network links such that each packet entering the network is seen at least once, for example, on all ingress links. Unfiltered sampling with a relatively low rate establishes baseline measurements of the network traffic. Reports include packet attributes of common interest: source and destination address and port numbers, prefix, protocol number, type of service, etc. Traffic matrices are indicated by reporting source and destination AS matrices. Absolute traffic volumes are estimated by renormalizing the sampled traffic volumes through division by Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 22]
Internet-Draft Passive Packet Measurement March 2003 either the target sampling rate, or the attained sampling rate (as derived by interface packet counters included in the report stream) Suppose an operator or a measurement based application detects an interesting subset of traffic identified by a particular packet attribute. Real-time drill-down to that subset is achieved by instantiating a new measurement process at the PSAMP device from which the subset was reported. The selection process of the new measurement process filters according to the attribute of interest, and composes with sampling if necessary to manage the rate of packet selection. 10.2 Customer Performance Hash-based sampling enables the tracking of the performance experience by customer traffic, customers identified by a list of source or destination prefixes, or by ingress or egress interfaces. Operational uses include the verification of SLAs, and troubleshooting following a customer complaint. In this application, Trajectory Sampling is enabled at all ingress and egress interfaces. The label hash is used to match up ingress and egress samples. Rates of loss in transit between ingress and egress are estimated from the proportion of trajectories for which no egress report is received. Note loss of customer packets is distinguishable from loss of packet reports through use of report sequence numbers. Assuming synchronization of clock between PSAMP devices, delay of customer traffic across the network may also be measured. Extending hash-sampling to all interfaces in the network would enable attribution of poor performance to individual network links. 10.3 Troubleshooting PSAMP can also be used to diagnose problems whose occurrence is evident from aggregate statistics, per interface utilization and packet loss statistics. These statistics are typically moving averages over relatively long time windows, e.g., 5 minutes, and serve as a coarse-grain indication of operational health of the network. The most common method of obtaining such measurements are through the appropriate SNMP MIBs (MIB-II and vendor-specific MIBs.) Suppose an operator detects a link that is persistently overloaded and experiences significant packet drop rates. There is a wide range of potential causes: routing parameters (e.g., OSPF link weights) that are poorly adapted to the traffic matrix, e.g., because of a shift in that matrix; a denial of service attack or a flash crowd; a routing problem (link flapping). In most cases, aggregate link statistics are not sufficient to distinguish between Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 23]
Internet-Draft Passive Packet Measurement March 2003 such causes, and to decide on an appropriate corrective action. For example, if routing over two links is unstable, and the links flap between being overloaded and inactive, this might be averaged out in a 5 minute window, indicating moderate loads on both links. Baseline PSAMP measurement the congested link, as described in Section 10.1, enables measurements that are fine grained in both space and time. The operator has to be able to determine how many bytes/packets are generated for each source/destination address, port number, and prefix, or other attributes, such as protocol number, MPLS forwarding equivalence class (FEC), type of service, etc. This allows to pinpoint precisely the nature of the offending traffic. For example, in the case of a DDoS attack, the operator would see a significant fraction of traffic with an identical destination address. In certain circumstances, precise information about the spatial flow of traffic through the network domain is required to detect and diagnose problems and verify correct network behavior. In the case of the overloaded link, it would be very helpful to know the precise set of paths that packets traversing this link follow. This would readily reveal a routing problem such as a loop, or a link with a misconfigured weight. More generally, complex diagnosis scenarios can benefit from measurement of traffic intensities (and other attributes) over a set of paths that is constrained in some way. For example, if a multihomed customer complains about performance problems on one of the access links from a particular source address prefix, the operator should be able to examine in detail the traffic from that source prefix which also traverses the specified access link towards the customer. While it is in principle possible to obtain the spatial flow of traffic through auxiliary network state information, e.g., by downloading routing and forwarding tables from routers, this information is often unreliable, outdated, voluminous, and contingent on a network model. For operational purposes, a direct observation of traffic flow is more reliable, as it does not depend on any such auxiliary information. For example, if there was a bug in a router's software, direct observation would allow to diagnose the effect of this bug, while an indirect method would not. Trajectory sampling by enabling common hash-based sampling on all routers in a domain supports such diagnoses. In particular, routing loops are revealed as cycles in trajectories. 11 References [B88] R.T. Braden, A pseudo-machine for packet monitoring and statistics, in Proc ACM SIGCOMM 1988 [ClPB93] K.C. Claffy, G.C. Polyzos, H.-W. Braun, Application of Sampling Methodologies to Network Traffic Characterization, Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 24]
Internet-Draft Passive Packet Measurement March 2003 Proceedings of ACM SIGCOMM'93, San Francisco, CA, USA, September 13-17, 1993 [DuGr01] N. G. Duffield and M. Grossglauser, Trajectory Sampling for Direct Traffic Observation, IEEE/ACM Trans. on Networking, 9(3), pp. 280-292, June 2001. [FHK02] S. Floyd, M. Handley. E. Kohler, Problem Statement for DCCP, Internet Draft draft-ietf-dccp-problem-00.txt, work in progress, October 2002. [HT52] D.G. Horvitz and D.J. Thompson, A Generalization of Sampling without replacement from a Finite Universe. J. Amer. Statist. Assoc. Vol. 47, pp. 663-685, 1952. [LCTV02] W.S. Lai, B.Christian, R.W. Tibbs, S. Van den Berghe, A Framework for Internet Traffic Engineering Measurement, Internet Draft draft-ietf-tewg-measure-04.txt, work in progress, September 2002. [PPM01] P. Phaal, S. Panchen, N. McKee, InMon Corporation's sFlow: A Method for Monitoring Traffic in Switched and Routed Networks, RFC 3176, September 2001 [PAMM98] V. Paxson, G. Almes, J. Mahdavi, M. Mathis, Framework for IP Performance Metrics, RFC 2330, May 1998 [QZCZCN02] J. Quittek, T. Zseby, B. Claise, S. Zander, G. Carle, K.C. Norseth, Requirements for IP Flow Information Export, Internet Draft draft-ietf-ipfix-reqs-08.txt, work in progress, January 2003. [SPSJTKS01] A. C. Snoeren, C. Partridge, L. A. Sanchez, C. E. Jones, F. Tchakountio, S. T. Kent, W. T. Strayer, Hash-Based IP Traceback, Proc. ACM SIGCOMM 2001, San Diego, CA, September 2001. [ZMR02] T. Zseby, M. Molina, F. Raspall, Sampling and Filtering Techniques for IP Packet Selection, Internet Draft draft-ietf-psamp-sample-tech-01.txt, work in progress, March 2003. 12 Authors' Addresses Nick Duffield AT&T Labs - Research Room B-139 180 Park Ave Florham Park NJ 07932, USA Phone: +1 973-360-8726 Email: duffield@research.att.com Albert Greenberg AT&T Labs - Research Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 25]
Internet-Draft Passive Packet Measurement March 2003 Room A-161 180 Park Ave Florham Park NJ 07932, USA Phone: +1 973-360-8730 Email: albert@research.att.com Matthias Grossglauser AT&T Labs - Research Room A-167 180 Park Ave Florham Park NJ 07932, USA Phone: +1 973-360-7172 Email: mgross@research.att.com Jennifer Rexford AT&T Labs - Research Room A-169 180 Park Ave Florham Park NJ 07932, USA Phone: +1 973-360-8728 Email: jrex@research.att.com Derek Chiou Avici Systems 101 Billerica Ave North Billerica, MA 01862 Phone: +1 978-964-2017 Email: dchiou@avici.com Peram Marimuthu Cisco Systems 170, W. Tasman Drive San Jose, CA 95134 Phone: (408) 527-6314 Email: peram@cisco.com Ganesh Sadasivan Cisco Systems 170 W. Tasman Drive San Jose, CA 95134 Phone: (408) 527-0251 Email: gsadasiv@cisco.com 13 Intellectual Property Statement AT&T Corporation may own intellectual property applicable to this contribution. The IETF has been notified of AT&T's licensing intent for the specification contained in this document. See http://www.ietf.org/ietf/IPR/ATT-GENERAL.txt for AT&T's IPR statement. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 26]
Internet-Draft Passive Packet Measurement March 2003 14 Full Copyright Statement Copyright (C) The Internet Society (1999). 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 implementation 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. Duffield et. al. draft-ietf-psamp-framework-02.txt [Page 27]
