<?xml version="1.0"encoding="US-ASCII"?>encoding="UTF-8"?> <!DOCTYPE rfc [ <!ENTITY nbsp " "> <!ENTITY zwsp "​"> <!ENTITY nbhy "‑"> <!ENTITY wj "⁠"> ]> <!-- edited with XMLSPY v5 rel. 3 U (http://www.xmlspy.com) by Daniel M Kohn (private) --><!DOCTYPE rfc SYSTEM "rfc2629.dtd"><rfccategory="std"xmlns:xi="http://www.w3.org/2001/XInclude" docName="draft-ietf-alto-performance-metrics-28"ipr="trust200902"> <?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?> <?rfc toc="yes" ?> <?rfc symrefs="yes" ?> <?rfc sortrefs="yes"?> <?rfc iprnotified="no" ?> <?rfc strict="yes" ?>number="9439" ipr="trust200902" obsoletes="" updates="" submissionType="IETF" category="std" consensus="true" xml:lang="en" tocInclude="true" symRefs="true" sortRefs="true" version="3"> <!-- xml2rfc v2v3 conversion 3.12.2 --> <front> <title abbrev="ALTO Performance CostMetrics">ALTOMetrics">Application-Layer Traffic Optimization (ALTO) Performance Cost Metrics</title> <seriesInfo name="RFC" value="9439"/> <author fullname="Qin Wu" initials="Q." surname="Wu"> <organization>Huawei</organization> <address> <postal> <extaddr>Yuhua District</extaddr> <street>101 SoftwareAvenue, Yuhua District</street>Avenue</street> <city>Nanjing</city> <region>Jiangsu</region> <code>210012</code> <country>China</country> </postal> <email>bill.wu@huawei.com</email> </address> </author> <author fullname="Y. Richard Yang" initials="Y." surname="Yang"> <organization>Yale University</organization> <address> <postal> <street>51 ProspectSt</street>St.</street> <city>New Haven</city> <region>CT</region> <code>06520</code> <country>United States of America</country> </postal> <email>yry@cs.yale.edu</email> </address> </author> <author fullname="Young Lee" initials="Y." surname="Lee"> <organization>Samsung</organization> <address><email>young.lee@gmail.com</email><email>younglee.tx@gmail.com</email> </address> </author> <author fullname="Dhruv Dhody" initials="D." surname="Dhody"> <organization>Huawei</organization> <address> <postal><street>Leela Palace</street> <city>Bangalore</city> <region>Karnataka</region> <code>560008</code><country>India</country> </postal> <email>dhruv.ietf@gmail.com</email> </address> </author> <author fullname="Sabine Randriamasy" initials="S." surname="Randriamasy"> <organization>NokiaBell Labs</organization>Networks France</organization> <address> <postal><street>Route de Villejust</street> <city>Nozay</city> <region></region> <code>91460</code><country>France</country> </postal> <email>sabine.randriamasy@nokia-bell-labs.com</email> </address> </author> <author fullname="Luis Miguel Contreras Murillo" initials="L." surname="Contreras"> <organization>Telefonica</organization> <address> <postal><street></street><street/> <city>Madrid</city><region></region> <code></code><region/> <code/> <country>Spain</country> </postal> <email>luismiguel.contrerasmurillo@telefonica.com</email> </address> </author> <dateyear="2022"year="2023" month="August" /><area>TSV Area</area> <workgroup>ALTO Working Group</workgroup> <keyword>RFC</keyword> <keyword>Request for Comments</keyword> <keyword>I-D</keyword> <keyword>Internet-Draft</keyword><area>tsv</area> <workgroup>alto</workgroup> <keyword>JavaScript ObjectNotation, Application-LayerNotation</keyword> <keyword>Application-Layer Traffic Optimization</keyword> <abstract> <t>The cost metric is a basic concept in Application-Layer Traffic Optimization (ALTO), and different applications may use different types of cost metrics. Since the ALTO base protocol (RFC 7285) defines only a single cost metric (namely, the generic "routingcost" metric), if an application wants to issue a cost map or an endpoint cost request in order to identify a resource provider that offers better performance metrics (e.g., lower delay or loss rate), the base protocol does not define the cost metric to be used.</t> <t>This document addresses this issue by extending the specification to provide a variety of network performance metrics, including network delay, delay variation(a.k.a, jitter),(a.k.a. jitter), packet loss rate, hop count, and bandwidth.</t> <t>There are multiple sources (e.g.,estimationestimations based on measurements orservice-level agreement) to derivea Service Level Agreement) available for deriving a performance metric. This document introduces an additional "cost-context" field to the ALTO "cost-type" field to convey the source of a performance metric.</t> </abstract> </front> <middle> <section anchor="secintro"title="Introduction">numbered="true" toc="default"> <name>Introduction</name> <t>Application-Layer Traffic Optimization (ALTO) provides a means for network applications to obtain network information so that the applications can identify efficient application-layer traffic patterns using the networks. Cost metrics are used in both the ALTO cost map service and the ALTO endpoint cost service in the ALTO base protocol <xreftarget="RFC7285"></xref>.</t>target="RFC7285" format="default"/>.</t> <t>Since different applications may use different cost metrics, the ALTO base protocolintroduces an ALTOintroduced the "ALTO CostMetric Registry (Section 14.2 of <xref target="RFC7285"></xref>)Metrics" registry (<xref target="RFC7285" section="14.2" sectionFormat="of"/>) as a systematic mechanism to allow different metrics to be specified. For example, a delay-sensitive application may want to uselatency relatedlatency-related metrics, and a bandwidth-sensitive application may want to usebandwidth relatedbandwidth-related metrics. However, the ALTO base protocol has registered only a single cost metric, i.e., the generic "routingcost" metric(Section 14.2 of <xref target="RFC7285"></xref>);(<xref target="RFC7285" section="14.2" sectionFormat="of"/>); nolatencylatency- orbandwidth relatedbandwidth-related metrics are defined in the base protocol.</t> <t>This document registers a set of new cost metrics(Table 1)(<xref target="costmetric" sectionFormat="bare"/>) to allow applications to determine"where"where to connect based on network performancecriteriacriteria, includingdelaydelay- andbandwidth relatedbandwidth-related metrics.</t><figure> <artwork><![CDATA[ +--------------------+-------------+--------------------------------+ | Metric | Definition | Semantics Based On | | |<table anchor="costmetric"> <name>Cost Metrics Defined inthis doc | | +--------------------+-------------+--------------------------------+ | One-way Delay | Section 4.1 | Base: [RFC7471,8570,8571] | | | |This Document</name> <thead> <tr> <th>Metric</th> <th>Definition in This Document</th> <th>Semantics Based On</th> </tr> </thead> <tbody> <tr> <td>One-Way Delay</td> <td><xref target="oneway" sectionFormat="bare"/></td> <td>Base: <xref target="RFC7471"/> <xref target="RFC8570"/> <xref target="RFC8571"/> sum of Unidirectional Delay| | Round-trip Delay | Section 4.2 | Base:of links along the path</td> </tr> <tr> <td>Round-Trip Delay</td> <td><xref target="delayrt" sectionFormat="bare"/></td> <td>Base: Sum of two directions| | | | from above | | Delay Variation | Section 4.3 | Base: [RFC7471,8570,8571] | | | | sumof Unidirectional Delay</td> </tr> <tr> <td>Delay Variation</td> <td><xref target="delayvar" sectionFormat="bare"/></td> <td>Base: <xref target="RFC7471"/> <xref target="RFC8570"/> <xref target="RFC8571"/> Sum of Unidirectional Delay| | | |Variation| | Loss Rate | Section 4.4 | Base: [RFC7471,8570,8571] | | | |of links along the path</td> </tr> <tr> <td>Loss Rate</td> <td><xref target="lossrate" sectionFormat="bare"/></td> <td>Base: <xref target="RFC7471"/> <xref target="RFC8570"/> <xref target="RFC8571"/> aggr Unidirectional LinkLoss | | Residual Bandwidth | Section 5.2 | Base: [RFC7471,8570,8571] | | | |Loss</td> </tr> <tr> <td>Residual Bandwidth</td> <td><xref target="bwresidual" sectionFormat="bare"/></td> <td>Base: <xref target="RFC7471"/> <xref target="RFC8570"/> <xref target="RFC8571"/> min Unidirectional ResidualBW| | Available Bandwidth| Section 5.3 | Base: [RFC7471,8570,8571] | | | |BW</td> </tr> <tr> <td>Available Bandwidth</td> <td><xref target="bwavailable" sectionFormat="bare"/></td> <td>Base: <xref target="RFC7471"/> <xref target="RFC8570"/> <xref target="RFC8571"/> min UnidirectionalAvail. BW | | | | | | TCP Throughput | Section 5.1 | [I-D.ietf-tcpm-rfc8312bis] | | | | | | Hop Count | Section 4.5 | [RFC7285] | +--------------------+-------------+--------------------------------+ Table 1. Cost Metrics Defined in this Document. ]]></artwork> </figure>Available BW</td> </tr> <tr> <td>TCP Throughput</td> <td><xref target="tput" sectionFormat="bare"/></td> <td><xref target="RFC9438"/></td> </tr> <tr> <td>Hop Count</td> <td><xref target="hopcount" sectionFormat="bare"/></td> <td><xref target="RFC7285"/></td> </tr> </tbody> </table> <t>The first6six metrics listed inTable 1<xref target="costmetric" sectionFormat="bare"/> (i.e.,One-way Delay, Round-trip Delay, Delay Variation, Loss Rate, Residual Bandwidth,one-way delay, round-trip delay, delay variation, loss rate, residual bandwidth, andAvailable Bandwidth)available bandwidth) are derived from the set oftraffic engineeringTraffic Engineering (TE) performance metrics commonly defined in OSPF <xreftarget="RFC3630"></xref>,target="RFC3630" format="default"/> <xreftarget="RFC7471"></xref>;target="RFC7471" format="default"/>, IS-IS <xreftarget="RFC5305"></xref>,target="RFC5305" format="default"/> <xreftarget="RFC8570"></xref>;target="RFC8570" format="default"/>, andBGP-LSBGP - Link State (BGP-LS) <xreftarget="RFC8571"></xref>.target="RFC8571" format="default"/>. Deriving ALTO cost performance metrics from existing network-layertraffic engineeringTE performance metrics, and making it exposed toexpose to application-layer traffic optimization,ALTO, can be a typical mechanism used by network operators to deploy ALTO <xreftarget="RFC7971"></xref>,target="RFC7971" format="default"/> <xreftarget="FlowDirector"></xref>.target="FlowDirector" format="default"/>. This document defines the base semantics of these metrics by extending them from link metrics to end-to-end metrics for ALTO. The "Semantics Based On" column specifies at a high level how the end-to-endmetric ismetrics are computed from link metrics;thedetails will be specified in the following sections.</t> <t>Thecommon metricsMin/Max Unidirectional Link Delay metric as defined in[RFC8570][RFC8571]<xref target="RFC8570"/> andMax Link<xref target="RFC8571"/>, and Maximum (Link) Bandwidth as defined in[RFC3630,RFC5305]<xref target="RFC3630"/> and <xref target="RFC5305"/>, are not listed inTable 1<xref target="costmetric" sectionFormat="bare"/> because they can be handled by applying the statistical operators defined in this document. The metrics relatedwithto utilized bandwidth and reservable bandwidth (i.e.,MaxMaximum ReservableBW(Link) Bandwidth and UnreservedBWBandwidth as defined in[RFC3630,RFC5305])<xref target="RFC3630"/> and <xref target="RFC5305"/>) are outside the scope of this document.</t> <t>The7thseventh metric in <xref target="costmetric" sectionFormat="bare"/> (the estimated TCP-flow throughput metric) provides an estimation of the bandwidth of a TCP flow, using TCP throughput modeling, to support use cases of adaptive applications <xreftarget="Prophet"></xref>,target="Prophet" format="default"/> <xreftarget="G2"></xref>.target="G2" format="default"/>. Note that other transport-specific metrics can be defined in the future. For example, QUIC-related metrics <xreftarget="RFC9000"></xref>target="RFC9000" format="default"/> can be considered when the methodologyto measurefor measuring such metrics is more mature (e.g., see <xreftarget="I-D.corre-quic-throughput-testing"></xref>).</t>target="I-D.corre-quic-throughput-testing" format="default"/>).</t> <t>The8theighth metric in <xref target="costmetric" sectionFormat="bare"/> (the hop count metric)in Table 1ismentionedmentioned, but not defined, in the ALTO base protocol[RFC7285], but not defined, and<xref target="RFC7285"/>; this documentdefines it to be complete.</t>provides a definition for it.</t> <t>These8eight performance metrics can be classified into two categories: those derived from the performance of individual packets (i.e.,One-way Delay, Round-trip Delay, Delay Variation, Loss Rate,one-way delay, round-trip delay, delay variation, loss rate, andHop Count),hop count) and those related to bandwidth/throughput(Residual(residual bandwidth, available bandwidth,and Available Bandwidth,and TCP throughput). These two categories are defined inSections <xrefSections <xref format="counter"target="secpktmetrics"></xref>target="secpktmetrics"/> and <xref format="counter"target="secbwmetrics"></xref>target="secbwmetrics"/>, respectively. Note that all metrics exceptRound-trip Delayround-trip delay are unidirectional. An ALTO client will need to query both directions if needed.</t> <t>The purpose of this document is to ensure proper usage of these8eight performance metrics in the context of ALTO. This document follows theguidelineguidelines defined inSection 14.2 of the ALTO base protocol<xreftarget="RFC7285"></xref>target="RFC7285" sectionFormat="of" section="14.2"/> on registering ALTO cost metrics. Hence, it specifies the identifier, the intended semantics, and the security considerations of each one of the metrics specified inTable 1.</t><xref target="costmetric" sectionFormat="bare"/>.</t> <t>The definitions of the intended semantics of the metrics tend to becoarse-grained,coarse grained and are for guidance only, and they may work well for ALTO. On the other hand, a performance measurement framework, such as the IP PerformanceMeasurementMetrics (IPPM) framework, may provide more detailsinfor defining a performance metric. This document introduces a mechanism called "cost-context" to provide additional details, when they are available; see <xreftarget="sec2"></xref>.</t> <!-- <t>Additionally, future versions of this document may define network metric values that stem from both measurements and provider policies such as many metrics related to end-to-end path bandwidth.</t> -->target="sec3" format="default"/>.</t> <t>Following the ALTO base protocol, this document uses JSON to specify the value type of each defined metric. See <xreftarget="RFC8259"></xref>target="RFC8259" format="default"/> for JSON data typespecification.specifications. In particular, <xreftarget="RFC7285"></xref>target="RFC7285" format="default"/> specifies that cost values should be assumed by defaultas JSONNumber.to be 'JSONNumber'. When defining the value representation of each metric inTable 1,<xref target="costmetric" sectionFormat="bare"/>, this document conforms to[RFC7285],<xref target="RFC7285"/> but specifies additional, generic constraints on valid JSONNumbers for each metric. For example, each new metric inTable 1<xref target="costmetric" sectionFormat="bare"/> will be specified as non-negative (>= 0); Hop Count is specified to be an integer.</t> <t>An ALTO server may provide only a subset of the metrics described in this document. For example, those that are subject to privacy concerns should not be provided to unauthorized ALTO clients. Hence, all cost metrics defined in this document are optional; not all of them need to be exposed to a given application. When an ALTO server supports a cost metric defined in this document, it announces the metric in its information resource directory (IRD) as defined inSection 9.2 of<xreftarget="RFC7285"></xref>.</t>target="RFC7285" section="9.2" sectionFormat="of"/>.</t> <t>An ALTO server introducing these metrics should consider related security issues. As a generic security considerationon theregarding reliability and trust in the exposed metric values, applicationsSHOULD rapidly give up<bcp14>SHOULD</bcp14> promptly stop using ALTO-based guidance if they detect that the exposed information does not preserve their performance level or even degrades it. <xreftarget="secsecconsider"></xref>target="secsecconsider" format="default"/> discusses security considerations in more detail.</t><!-- <t>The definitions of a set of cost metrics can allow us to extend the ALTO base protocol (e.g., allowing output and constraints use different cost metrics), but such extensions are not in the scope of this document.</t> --></section> <sectiontitle="Requirements Language">numbered="true" toc="default"> <name>Requirements Language</name> <t>The key words"MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY","<bcp14>MUST</bcp14>", "<bcp14>MUST NOT</bcp14>", "<bcp14>REQUIRED</bcp14>", "<bcp14>SHALL</bcp14>", "<bcp14>SHALL NOT</bcp14>", "<bcp14>SHOULD</bcp14>", "<bcp14>SHOULD NOT</bcp14>", "<bcp14>RECOMMENDED</bcp14>", "<bcp14>NOT RECOMMENDED</bcp14>", "<bcp14>MAY</bcp14>", and"OPTIONAL""<bcp14>OPTIONAL</bcp14>" in this document are to be interpreted as described inBCP 14BCP 14 <xref target="RFC2119"/> <xreftarget="RFC2119"></xref><xref target="RFC8174"></xref>target="RFC8174"/> when, and only when, they appear in all capitals, as shown here.</t> </section> <sectionanchor="sec2" title="Performanceanchor="sec3" numbered="true" toc="default"> <name>Performance MetricAttributes">Attributes</name> <t>The definitions of the metrics in this document arecoarse-grained,coarse grained, based on network-layertraffic engineeringTE performance metrics, and for guidance only. A fine-grained framework as specified in <xreftarget="RFC6390"></xref>target="RFC6390" format="default"/> requires that the fine-grained specification of a network performance metric include6six components:(i)(1) Metric Name,(ii)(2) Metric Description,(iii)(3) Method of Measurement or Calculation,(iv)(4) Units of Measurement,(v)(5) Measurement Points, and(vi)(6) Measurement Timing. Requiring that an ALTO serverprovidesprovide precise, fine-grained values for all6six components for each metric that it exposes may not be feasible or necessary for all ALTO use cases. For example, an ALTO server computing its metrics from network-layertraffic-engineeringTE performance metrics may not have information about the method of measurement or calculation (e.g., measured traffic patterns).</t> <t>To address the issue and realize ALTO usecases,cases for the metrics listed inTable 1,<xref target="costmetric" sectionFormat="bare"/>, this document defines performance metric identifierswhichthat can be used in the ALTOprotocolProtocol with the following well-defined(i)items: (1) Metric Name,(ii)(2) Metric Description,(iv)(3) Units of Measurement, and(v)(4) Measurement Points, which are always specified by the specific ALTO services; for example, the endpoint cost service is between the two endpoints. Hence, the ALTO performance metric identifiers provide basic metric attributes.</t> <t>To allow the flexibility of allowing an ALTO server to provide fine-grained information such as Method of Measurement orCalculation,Calculation according to its policy and use cases, this document introduces context information so that the server can provide these additional details.</t> <section anchor="meta"title="Performancenumbered="true" toc="default"> <name>Performance Metric Context:"cost-context"">"cost-context"</name> <t>The core additional details of a performance metric specify"how"how the metric is obtained. This is referred to as the source of the metric. Specifically, this document defines three types of coarse-grained metric information sources: "nominal",and "sla" (service level agreement),"sla", and "estimation".</t> <t>For a given type of source, precise interpretation of a performance metric value can depend on specific measurement and computation parameters.</t> <t>To make it possible to specify the source and the aforementioned parameters, this document introduces an optional "cost-context" field to the "cost-type" field defined by the ALTO base protocol(Section 10.7 of <xref target="RFC7285"></xref>)(<xref target="RFC7285" section="10.7" sectionFormat="of"/>) asthe following:</t> <figure> <artwork><![CDATA[follows:</t> <sourcecode type="json"><![CDATA[ object { CostMetric cost-metric; CostMode cost-mode; [CostContext cost-context;] [JSONString description;] } CostType; object { JSONString cost-source; [JSONValue parameters;] } CostContext;]]></artwork> </figure>]]></sourcecode> <t>"cost-context" will not be used as a key to distinguish among performance metrics. Hence, an ALTO information resourceMUST NOT<bcp14>MUST NOT</bcp14> announce multiple CostType entries with the same "cost-metric","cost-mode""cost-mode", and "cost-context". They must be placed into different information resources.</t> <t>The "cost-source" field of the "cost-context" field is defined as a string consisting of onlyUS-ASCIIASCII alphanumeric characters (U+0030-U+0039, U+0041-U+005A, and U+0061-U+007A). Thecost-source"cost-source" field is used in this document to indicate a string of this format.</t> <t>As mentioned above, this document defines three values for "cost-source": "nominal", "sla", and "estimation". The "cost-source" field of the "cost-context" fieldMUST<bcp14>MUST</bcp14> be one that is registered in the "ALTO CostSource"Source Types" registry(Section 8).</t>(<xref target="ianaconsider"/>).</t> <t>The "nominal" category indicates that the metric value is statically configured by the underlying devices. Not all metrics have reasonable "nominal" values. For example, throughput can have a nominal value, which indicates the configured transmission rate of the involved devices; latency typically does not have a nominal value.</t> <t>The "sla" category indicates that the metric value is derived from somecommitmentcommitment, which this document refers to asservice-level agreementa Service Level Agreement (SLA). Some operators also use terms such as "target" or "committed" values. For an "sla" metric, it isRECOMMENDED<bcp14>RECOMMENDED</bcp14> that the "parameters" field provide a link to the SLA definition.</t> <t>The "estimation" category indicates that the metric value is computed through an estimation process. An ALTO server may compute "estimation" values by retrieving and/or aggregating information from routing protocols (e.g., see <xreftarget="RFC7471"></xref>,target="RFC7471" format="default"/>, <xreftarget="RFC8570"></xref>,target="RFC8570" format="default"/>, and <xreftarget="RFC8571"></xref>),target="RFC8571" format="default"/>), traffic measurement management tools (e.g.,TWAMPthe Two-Way Active Measurement Protocol (TWAMP) <xreftarget="RFC5357"></xref>),target="RFC5357" format="default"/>), and measurement frameworks (e.g., IPPM), with corresponding operational issues. An illustration of potential information flows used for estimating these metrics is shown inFigure 1.<xreftarget="secopconsider"></xref>target="fig1"/>. <xref target="secopconsider" format="default"/> discusses in more detail the operational issues and how a network may address them.<figure> <artwork><![CDATA[</t> <figure anchor="fig1"> <name>A Framework to Compute Estimation of Performance Metrics</name> <artwork name="" type="" align="left" alt=""><![CDATA[ +--------+ +--------+ +--------+ | Client | | Client | | Client | +----^---+ +---^----+ +---^----+ | | | +-----------|-----------+North-Bound|ALTOprotocol Interface (NBI)|Protocol | | +--+-----+ retrieval +-----------+ | ALTO |<----------------| Routing | | Server | and aggregation| Protocols | | |<-------------+ |Protocols| +--------+ | +-----------+ | | +------------+ | |Performance | ---| Monitoring | | Tools | +------------+Figure 1. A framework to compute estimation to performance metrics]]></artwork></figure></t> <!-- <t> A particular type of "estimation" is direct "import", which indicates that the metric value is imported directly from a specific existing protocol or system. Specifying "import" as the source instead of the more generic "estimation" may allow better tracking of information flow. For an "import" metric, it is RECOMMENDED that the "parameters" field provides details to the system from which raw data is imported. In particular, one may notice that the set of end-to-end metrics defined in Table 1 has a large overlap with the set defined in [RFC8571], in the setting of IGP traffic engineering performance metrics for each link (i.e., unidirectional link delay, min/max unidirectional link delay, unidirectional delay variation, unidirectional link loss, unidirectional residual bandwidth, unidirectional available bandwidth, unidirectional utilized bandwidth). Hence, an ALTO server may use "import" to indicate that its end-to-end metrics are computed from link metrics imported from [RFC8571]. </t> --></figure> <t>There can be multiplechoices in decidingoptions available when choosing thecost-source category. It is"cost-source" category; the operator of an ALTO serverwho chooses the category.will make that choice. If a metric does not include a "cost-source" value, the applicationMUST<bcp14>MUST</bcp14> assume that the value of "cost-source" is the most generic source, i.e., "estimation".</t> </section> <section anchor="percentile"title="Performancenumbered="true" toc="default"> <name>Performance MetricStatistics">Statistics</name> <t>The measurement of a performance metric often yields a set of samples from an observation distribution(<xref target="Prometheus"></xref>),<xref target="Prometheus" format="default"/>, instead of a single value. A statistical operator is applied to the samples to obtain a value to be reported to the client. Multiple statistical operators (e.g., min, median, and max) are commonly being used.</t> <t>Hence, this document extends the generalUS-ASCIIASCII alphanumeric cost metric strings, formally specified as the CostMetric type defined inSection 10.6 of [RFC7285],<xref target="RFC7285" sectionFormat="of" section="10.6"/>, as follows:</t><t><list style="hanging"> <t>A<t indent="3">A cost metric string consists of a base metric identifier (or base identifier for short) string, followed by an optional statistical operator string, connected by the ASCIIcharactercolon character (':', U+003A), if the statistical operator string exists. The total length of the cost metric stringMUST NOT<bcp14>MUST NOT</bcp14> exceed 32, as required by[RFC7285].</t> </list></t> <!-- </t> <figure> <artwork> <![CDATA[ <metric-identifier> ::= <metric-base-identifier> [ '-' <stat> ] ]]> </artwork> </figure> <t>where <stat> MUST be one of the following: </t> --><xref target="RFC7285"/>.</t> <t>The statistical operator stringMUST<bcp14>MUST</bcp14> be one of the following:</t><t><list style="hanging"> <t hangText="cur:"><vspace blankLines="1" /> the<dl> <dt>cur:</dt> <dd>The instantaneous observation value of the metric from the most recent sample (i.e., the currentvalue). <vspace blankLines="1" /></t> <t hangText="percentile,value).</dd> <dt>percentile, with the letter 'p' followed by anumber:"><vspace blankLines="1" /> givesnumber:</dt> <dd>Gives the percentile specified by the number following the letter 'p'. The numberMUST<bcp14>MUST</bcp14> be a non-negative JSON number in the range [0, 100] (i.e., greater than or equal to 0 and less than or equal to 100), followed by an optional decimal part, ifahigher precision is needed. The decimal part should start with the '.' separator(U+002E),(U+002E) and be followed by a sequence of one or more ASCII numbers between '0' and '9'. Assume that this number isyy, and consider the case where the samples are coming from a random variable X.Then theThe metric then returns x, such that the probability of X is less than or equal to x, i.e., Prob(X <= x), = y/100. For example, delay-ow:p99 gives the99%99th percentile of observed one-way delay; delay-ow:p99.9 gives the99.9%99.9th percentile. Note that some systems use quantile, which is in the range [0, 1]. When there is a more common form for a given percentile, it isRECOMMENDED<bcp14>RECOMMENDED</bcp14> that the common form be used; that is, instead of p0, use min; instead of p50, use median; instead of p100, usemax. <vspace blankLines="1" /></t> <t hangText="min:"><vspace blankLines="1" /> themax.</dd> <dt>min:</dt> <dd>The minimal value of theobservations. <vspace blankLines="1" /></t> <t hangText="max:"><vspace blankLines="1" /> theobservations.</dd> <dt>max:</dt> <dd>The maximal value of theobservations. <vspace blankLines="1" /></t> <t hangText="median:"><vspace blankLines="1" /> the mid-pointobservations.</dd> <dt>median:</dt> <dd>The midpoint (i.e., p50) of theobservations. <vspace blankLines="1" /></t> <t hangText="mean:"><vspace blankLines="1" /> theobservations.</dd> <dt>mean:</dt> <dd>The arithmetic mean value of theobservations. <vspace blankLines="1" /></t> <t hangText="stddev:"><vspace blankLines="1" /> theobservations.</dd> <dt>stddev:</dt> <dd>The standard deviation of theobservations. <vspace blankLines="1" /></t> <t hangText="stdvar:"><vspace blankLines="1" /> theobservations.</dd> <dt>stdvar:</dt> <dd>The standard variance of theobservations. <vspace blankLines="1" /></t> </list></t>observations.</dd> </dl> <t>Examples of cost metric strings then include "delay-ow", "delay-ow:min", and "delay-ow:p99", where "delay-ow" is the base metric identifier string; "min" and "p99" are example statistical operator strings.</t> <t>If a cost metric string does not have the optional statistical operator string, the statistical operatorSHOULD<bcp14>SHOULD</bcp14> be interpreted as the default statistical operator in the definition of the base metric. If the definition of the base metric does not provide a definition for the default statistical operator, the metricMUST<bcp14>MUST</bcp14> be consideredasthe median value.</t> <t>Note thatRFC 7258<xref target="RFC7285"/> limits the overall cost metric identifier to 32 characters. The cost metric variants with statistical operator suffixes defined by this document are also subject to the same overall 32-character limit, so certain combinations of (long) base metricidentifieridentifiers and statisticaloperatoroperators will not be representable. If such a situation arises, it could be addressed by defining a new base metric identifier that is an "alias" of the desired base metric, with identical semantics and just a shorter name.</t> </section> </section><!-- End of metric attributes --><section anchor="secpktmetrics"title="Packetnumbered="true" toc="default"> <name>Packet PerformanceMetrics ">Metrics</name> <t>This section introduces ALTO network performance metrics onone wayone-way delay, round-trip delay, delay variation, packet loss rate, and hop count. They measure the "quality of experience" of the stream of packets sent from a resource provider to a resource consumer. Themeasuresmeasurements of each individual packet (pkt) can include the delay from the time when the packet enters the network to the time when the packet leaves the network(pkt.delay);(pkt.delay), whether the packet is dropped before reaching the destination(pkt.dropped);(pkt.dropped), and the number of network hops that the packet traverses (pkt.hopcount). The semantics of the performance metrics defined in this section are that they are statistics computed from thesemeasures;measurements; for example, the x-percentile of the one-way delay is the x-percentile of the set of delays {pkt.delay} for the packets in the stream.</t> <sectiontitle="Costanchor="oneway" numbered="true" toc="default"> <name>Cost Metric: One-Way Delay(delay-ow)">(delay-ow)</name> <sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "delay-ow".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value conforming to the numberspecification of Section 6 of [RFC8259].specifications provided in <xref target="RFC8259" sectionFormat="of" section="6"/>. The unit is expressed in microseconds. Hence, the number can be afloating pointfloating-point number to express delay that is smaller than microseconds. The numberMUST<bcp14>MUST</bcp14> be non-negative.</t> </section><!-- <t><list style="hanging"> <t hangText="Metric name:"> <vspace blankLines="1"/>One Way Delay<vspace blankLines="1"/> </t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>owdelay<vspace blankLines="1"/> </t> </list></t> --><sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt><dd>To specify the temporal and spatial aggregated delay of a stream of packets from the specified source to the specified destination. The base semantics of the metric is the Unidirectional Delay metric as defined in[RFC8571,RFC8570,RFC7471],<xref target="RFC8571"/>, <xref target="RFC8570"/>, and <xref target="RFC7471"/>, but instead of specifying the delay for a link, it is the (temporal) aggregation of the link delays from the source to the destination. A non-normative reference definition of the end-to-end one-way delay metric is provided in <xreftarget="RFC7679"></xref>.target="RFC7679" format="default"/>. The spatial aggregation level is specified in the query context, e.g., provider-defined identifier (PID) to PID, or endpoint to endpoint, where the PID is as defined inSection 5.1 of [RFC7285].</t> <t>Use: This<xref target="RFC7285" sectionFormat="of" section="5.1"/>.</dd> <dt>Use:</dt> <dd>This metric could be used as a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[Example 1: Delay valueresponse.</dd> </dl> <figure anchor="example-1"> <name>Delay Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 1)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 239 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-ow" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 247 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-ow" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 10, "ipv4:198.51.100.34": 20 } } }]]></artwork>]]></sourcecode> </figure> <t>Note that since the "cost-type" does not include the "cost-source" field, the values are based on "estimation". Since the identifier does not include the statistical operator string component, the values will represent median values.</t><t>Example 1a below<t><xref target="example-1a"/> shows an example that is similar to Example1,1 (<xref target="example-1"/>), but for IPv6.</t><figure> <artwork><![CDATA[Example 1a: Delay value<figure anchor="example-1a"> <name>Delay Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs for IPv6 (Example 1a)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 252 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-ow" }, "endpoints": { "srcs": [ "ipv6:2001:db8:100::1" ], "dsts": [ "ipv6:2001:db8:100::2", "ipv6:2001:db8:100::3" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 257 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-ow" } }, "endpoint-cost-map": { "ipv6:2001:db8:100::1": { "ipv6:2001:db8:100::2": 10, "ipv6:2001:db8:100::3": 20 } } }]]></artwork>]]></sourcecode> </figure> </section> <section anchor="ccspec-ow"title="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt><dd>Typically, network one-way delay does not have a nominalvalue.</t> <t>"sla": Manyvalue.</dd> <dt>"sla":</dt><dd>Many networks provide delay-related parameters in their application-level SLAs. It isRECOMMENDED<bcp14>RECOMMENDED</bcp14> that the "parameters" field of an "sla" one-way delay metric include a link (i.e., a field named "link") providingana URItofor the specification of SLA details, if available. Such a specification can be eitherfree(1) free text for possible presentation to theuser,user ora(2) a formal specification. The format of the specification isout ofoutside the scope of thisdocument.</t> <!-- <t>"import": There can be multiple sources to import one-way delay. For example, if the import is from [RFC8571] (by using unidirectional link delay, min/max unidirectional link delay), it is RECOMMENDED that "parameters" provides "protocol" as a field and "RFC8571" as the value. During import, the server should be cognizant of potential issues when computing an end-to-end summary statistic from link statistics. Another example of an import source is the IPPM framework. For IPPM, it is RECOMMENDED that "parameters" provides "protocol" as a field and "ippm" as the value; see Section 4 of [I-D.ietf-ippm-initial-registry] for additional fields which can be specified for "ippm" in "parameters". </t> --> <t>"estimation": Thedocument.</dd> <dt>"estimation":</dt><dd>The exact estimation method isout ofoutside the scope of this document. There can be multiple sourcesto estimatefor estimating one-way delay. For example, the ALTO server may estimate the end-to-end delay by aggregation of routing protocol link metrics; the server may also estimate the delay using active, end-to-endmeasurements,measurements -- for example, using the IPPM framework <xreftarget="RFC2330"></xref>.</t>target="RFC2330" format="default"/>.</dd> </dl> <t>If the estimation is computed by aggregation of routing protocol link metrics (e.g., Unidirectional Link Delay metrics for OSPF <xreftarget="RFC7471"></xref>,target="RFC7471" format="default"/>, IS-IS <xreftarget="RFC8570"></xref>,target="RFC8570" format="default"/>, or BGP-LS <xreftarget="RFC8571"></xref>) Unidirectional Delay link metrics,target="RFC8571" format="default"/>), it isRECOMMENDED<bcp14>RECOMMENDED</bcp14> that the "parameters" field of an "estimation" one-way delay metric include the following information: (1) the RFC defining the routing protocol metrics (e.g.,https://www.rfc-editor.org/info/rfc7471see <xref target="RFC7471"/> forRFC7471derivedmetrics);metrics), (2) configurations of the routing link metrics such as configuredintervals;intervals, and (3) the aggregation method from link metrics to end-to-end metrics. During aggregation from link metrics totheend-to-endmetric,metrics, the server should be cognizant of potential issues when computing an end-to-end summary statistic from link statistics. The default end-to-end average one-way delay is the sum of average link one-way delays. If an ALTO server provides the min and max statistical operators for the one-way delay metric, the values can be computed directly from the routing link metrics, as[RFC7471,RFC8570,RFC8571]<xref target="RFC7471"/>, <xref target="RFC8570"/>, and <xref target="RFC8571"/> provide Min/Max Unidirectional Link Delay.</t> <t>If the estimation is from the IPPM measurement framework, it isRECOMMEDED<bcp14>RECOMMENDED</bcp14> that the "parameters" field of an "estimation" one-way delay metricincludes the following information:include the URItoin theURI"URI" field of the IPPM metric defined in the IPPMperformance metric <xref target="IANA-IPPM"></xref>"Performance Metrics" registry <xref target="IANA-IPPM" format="default"/> (e.g.,https://www.iana.org/assignments/performance-metrics/OWDelay_Active_IP-UDP-Poisson-Payload250B_RFC8912sec7_Seconds_95Percentile).<eref target="https://www.iana.org/assignments/performance-metrics/OWDelay_Active_IP-UDP-Poisson-Payload250B_RFC8912sec7_Seconds_95Percentile" brackets="angle"/>). The IPPM metricMUST<bcp14>MUST</bcp14> be one-way delay (i.e., IPPM OWDelay* metrics). The statistical operator of the ALTO metricMUST<bcp14>MUST</bcp14> be consistent with the IPPM statistical property (e.g.,95-th95th percentile).</t><!-- <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>See section 8.3 of [I-D.ietf-ippm-initial-registry] for potential measurement method.<vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1, Data sources for potential data sources.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See section 8.3.5 of [I-D.ietf-ippm-initial-registry] for potential measurement timing considerations.<vspace blankLines="1"/></t> </list></t> --></section> </section> <sectiontitle="Costanchor="delayrt" numbered="true" toc="default"> <name>Cost Metric:Round-tripRound-Trip Delay(delay-rt)">(delay-rt)</name> <sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "delay-rt".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value conforming to the numberspecification of Section 6 of [RFC8259].specifications provided in <xref target="RFC8259" sectionFormat="of" section="6"/>. The numberMUST<bcp14>MUST</bcp14> be non-negative. The unit is expressed in microseconds.</t> </section><!-- <t><list style="hanging"> <t hangText="Metric name:"> <vspace blankLines="1"/>Round Trip Time<vspace blankLines="1"/> </t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>rtt<vspace blankLines="1"/> </t> </list></t> --><sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt><dd><t>To specify temporal and spatial aggregated round-trip delay between the specified source and specified destination. The base semantics is that it is the sum of the one-way delay from the source to the destination and the one-way delay from the destination back to the source, where the one-way delay is as defined inSection 4.1.<xref target="oneway"/>. A non-normative reference definition of the end-to-end round-trip delay metric is provided in <xreftarget="RFC2681"></xref>.target="RFC2681" format="default"/>. The spatial aggregation level is specified in the query context (e.g., PID to PID, or endpoint to endpoint).</t> <t>Note that it is possible for a client to query two one-waydelaysdelay (delay-ow) items and then compute the round-trip delay. The server should be cognizant of the consistency ofvalues.</t> <t>Use: Thisvalues.</t></dd> <dt>Use:</dt> <dd>This metric could be usedeitheras a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[ Example 2: Round-tripresponse.</dd> </dl> <figure anchor="example-2"> <name>Round-Trip Delay ofsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 2)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 238 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-rt" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 245 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-rt" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 4, "ipv4:198.51.100.34": 3 } } }]]></artwork>]]></sourcecode> </figure> </section><!--<sectiontitle="Measurement Considerations"> <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>See section 4.3 of [I-D.ietf-ippm-initial-registry] for potential measurement method. <vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See section 4.1, Data sources.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See section 4.3.5 of [I-D.ietf-ippm-initial-registry] for Measurement Timing. <vspace blankLines="1"/></t> </list></t> </section> <section title="Measurement Considerations and Parameters"> <t>See Section 4 of [I-D.ietf-ippm-initial-registry] for measurement considerations and parameters which may be specified in "parameters". Note that the "parameters" field is an optional field providing non-normative information. </t> </section> --> <section title="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt><dd>Typically, network round-trip delay does not have a nominalvalue.</t> <t>"sla": Seevalue.</dd> <dt>"sla":</dt><dd>See the "sla" entry in <xreftarget="ccspec-ow"></xref>.</t> <!-- <t>"import": There can be multiple sources to import round-trip delay. If the import is from [RFC8571] (by using unidirectional link delay, min/max unidirectional link delay), it is RECOMMENDED that "parameters" provides "protocol" as a field and "RFC8571" as the value; see <xreftarget="ccspec-ow"/> for discussions on summing up link metrics to obtain end-to-end metrics. If the import is from the IPPM framework, it is RECOMMENDED that "parameters" provides "protocol" as a field and "ippm" as the value; see Section 4 of [I-D.ietf-ippm-initial-registry] for additional fields which can be specified for "ippm" in "parameters". </t> --> <t>"estimation": Seeformat="default"/>.</dd> <dt>"estimation":</dt><dd>See the "estimation" entry in <xreftarget="ccspec-ow"></xref>.target="ccspec-ow" format="default"/>. For estimation by aggregation of routing protocol link metrics, the aggregation should include all links from the source to the destination and then back to the source; for estimation using IPPM, the IPPM metricMUST<bcp14>MUST</bcp14> be round-trip delay (i.e., IPPM RTDelay* metrics). The statistical operator of the ALTO metricMUST<bcp14>MUST</bcp14> be consistent with the IPPM statistical property (e.g.,95-th percentile).</t> <!-- <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>See section 8.3 of [I-D.ietf-ippm-initial-registry] for potential measurement method.<vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1, Data sources for potential data sources.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See section 8.3.5 of [I-D.ietf-ippm-initial-registry] for potential measurement timing considerations.<vspace blankLines="1"/></t> </list></t> -->95th percentile).</dd> </dl> </section> </section> <sectiontitle="Costanchor="delayvar" numbered="true" toc="default"> <name>Cost Metric: Delay Variation(delay-variation)"> <!-- <t><list style="hanging"> <t hangText="Metric name:"> <vspace blankLines="1"/>Packet Delay Variation<vspace blankLines="1"/> </t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>pdv<vspace blankLines="1"/> </t> </list></t> -->(delay-variation)</name> <sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "delay-variation".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value conforming to the numberspecification of Section 6 of [RFC8259].specifications provided in <xref target="RFC8259" sectionFormat="of" section="6"/>. The numberMUST<bcp14>MUST</bcp14> be non-negative. The unit is expressed in microseconds.</t> </section> <sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt><dd><t>To specify temporal and spatial aggregated delay variation (also called delayjitter))jitter) with respect to the minimum delay observed on the stream over the one-way delay from the specified source and destination, where the one-way delay is as defined inSection 4.1.<xref target="oneway"/>. A non-normative reference definition of the end-to-end one-way delay variation metric is provided in <xreftarget="RFC3393"></xref>.target="RFC3393" format="default"/>. Note that <xreftarget="RFC3393"></xref>target="RFC3393" format="default"/> allows the specification of a generic selection function F to unambiguously define the two packets selected to compute delay variations. This document defines the specific casethatwhere F selectsasthe"first"packetthe onewith the smallest one-waydelay.delay as the "first" packet. The spatial aggregation level is specified in the query context (e.g., PID to PID, or endpoint to endpoint).</t> <t>Note that in statistics,variations arevariation is typically evaluated by the distance from samples relative to the mean. Innetworking context,the context of networking, it is more commonly defined from samples relative to the min. This definition follows the networking convention.</t><t>Use: This</dd> <dt>Use:</dt> <dd>This metric could be usedeitheras a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[Example 3: Delay variation valueresponse.</dd> </dl> <figure anchor="example-3"> <name>Delay Variation Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 3)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 245 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-variation" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } } HTTP/1.1 200 OK Content-Length: 252 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "delay-variation" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 0, "ipv4:198.51.100.34": 1 } } }]]></artwork>]]></sourcecode> </figure> </section><!-- <section title="Measurement Considerations"> <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>See Section 5.3 of [I-D.ietf-ippm-initial-registry] for potential measurement method.<vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1, Data sources for potential data sources.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See Section 5.3.5 of [I-D.ietf-ippm-initial-registry] for Measurement Timing.<vspace blankLines="1"/></t> </list></t> </section> <section title="Measurement Considerations and Parameters"> <t>See Section 5 of [I-D.ietf-ippm-initial-registry] for measurement considerations and parameters which may be specified in "parameters". Note that the "parameters" field is an optional field providing non-normative information. </t> </section> --><sectiontitle="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt><dd>Typically, network delay variation does not have a nominalvalue.</t> <t>"sla": Seevalue.</dd> <dt>"sla":</dt><dd>See the "sla" entry in <xreftarget="ccspec-ow"></xref>.</t> <!-- <t>"import": There can be multiple sources to import delay variation. If the import is from [RFC8571] (by using unidirectional delay variation), it is RECOMMENDED that "parameters" provides "protocol" as a field and "RFC8571" as the value; see <xreftarget="ccspec-ow"/> for discussions on summing up link metrics to obtain end-to-end metrics. If the import is from the IPPM framework, it is RECOMMENDED that "parameters" provides "protocol" as a field and "ippm" as the value; see Section 4 of [I-D.ietf-ippm-initial-registry] for additional fields which can be specified for "ippm" in "parameters". </t> --> <t>"estimation": Seeformat="default"/>.</dd> <dt>"estimation":</dt><dd>See the "estimation" entry in <xreftarget="ccspec-ow"></xref>.target="ccspec-ow" format="default"/>. For estimation by aggregation of routing protocol link metrics, the default aggregation of the average of delay variations is the sum of the link delay variations; for estimation using IPPM, the IPPM metricMUST<bcp14>MUST</bcp14> be delay variation (i.e., IPPM OWPDV* metrics). The statistical operator of the ALTO metricMUST<bcp14>MUST</bcp14> be consistent with the IPPM statistical property (e.g.,95-th percentile).</t> <!-- <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>See section 8.3 of [I-D.ietf-ippm-initial-registry] for potential measurement method.<vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1, Data sources for potential data sources.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See section 8.3.5 of [I-D.ietf-ippm-initial-registry] for potential measurement timing considerations.<vspace blankLines="1"/></t> </list></t> -->95th percentile).</dd> </dl> </section> </section> <sectiontitle="Costanchor="lossrate" numbered="true" toc="default"> <name>Cost Metric: Loss Rate(lossrate)"> <!-- <t><list style="hanging"> <t hangText="Metric name:"><vspace blankLines="1"/>Packet loss<vspace blankLines="1"/></t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>pktloss<vspace blankLines="1"/> </t> </list></t> -->(lossrate)</name> <sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "lossrate".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value conforming to the numberspecification of Section 6 of [RFC8259].specifications provided in <xref target="RFC8259" sectionFormat="of" section="6"/>. The numberMUST<bcp14>MUST</bcp14> be non-negative. The value represents the percentage of packet losses.</t> </section> <sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt><dd>To specify the temporal and spatial aggregated one-way packet loss rate from the specified source and the specified destination. The base semantics of the metric is the Unidirectional Link Loss metric as defined in[RFC8571,RFC8570,RFC7471],<xref target="RFC8571"/>, <xref target="RFC8570"/>, and <xref target="RFC7471"/>, but instead of specifying the loss for a link, it is the aggregated loss of all links from the source to the destination. The spatial aggregation level is specified in the query context (e.g., PID to PID, or endpoint toendpoint).</t> <t>Use: Thisendpoint).</dd> <dt>Use:</dt> <dd>This metric could be used as a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[ Example 5: Loss rate valueresponse.</dd> </dl> <figure anchor="example-4"> <name>Loss Rate Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 4)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 238 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "lossrate" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 248 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "lossrate" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 0, "ipv4:198.51.100.34": 0.01 } } }]]></artwork> </figure> </section> <!-- <section title="Measurement Considerations and Parameters"> <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>See Section 2.6 of [RFC7680] for Measurement Method.<vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1 this document, Data sources.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See Section 2 and Section 3 of [RFC7680] for Measurement Timing.<vspace blankLines="1"/></t> </list></t> <t>See Section 4 of [I-D.ietf-ippm-initial-registry] for measurement considerations and parameters which may be specified in "parameters". Note that the "parameters" field is an optional field providing non-normative information.</t>]]></sourcecode></figure> </section>--><sectiontitle="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt> <dd>Typically, the packet loss rate does not have a nominal value, although some networks may specify zerolosses.</t> <t>"sla": Seelosses.</dd> <dt>"sla":</dt> <dd>See the "sla" entry in <xreftarget="ccspec-ow"></xref>..</t> <!-- <t>"import": There can be multiple sources to import packet loss rate. If the import is from [RFC8571] (by using unidirectional link loss), it is RECOMMENDED that "parameters" provides "protocol" as a field and "RFC8571" as the value; see <xreftarget="ccspec-ow"/> for discussions on summing up link metrics to obtain end-to-end metrics. If the import is from the IPPM framework, it is RECOMMENDED that "parameters" provides "protocol" as a field and "ippm" as the value; see Section 4 of [I-D.ietf-ippm-initial-registry] for additional fields which can be specified for "ippm" in "parameters". </t> --> <t>"estimation": Seeformat="default"/>.</dd> <dt>"estimation":</dt> <dd>See the "estimation" entry in <xreftarget="ccspec-ow"></xref>.target="ccspec-ow" format="default"/>. For estimation by aggregation of routing protocol link metrics, the default aggregation of the averageofloss rate is the sum of the linklinkloss rates. But this default aggregation is valid only if two conditions are met: (1)it is valid only whenlink loss rates arelow,low and (2)itone assumes that each link's loss events are uncorrelated with every other link's loss events. When loss rates at the links are high but independent, the general formula for aggregatinglossloss, assuming that each link isindependentindependent, is to compute end-to-end loss as one minus the product of the success rate for each link. Aggregation when losses at links are correlated can be morecomplexcomplex, and the ALTO server should be cognizant of correlated loss rates. For estimation using IPPM, the IPPM metricMUST<bcp14>MUST</bcp14> be packet loss (i.e., IPPM OWLoss* metrics). The statistical operator of the ALTO metricMUST<bcp14>MUST</bcp14> be consistent with the IPPM statistical property (e.g.,95-th percentile).</t>95th percentile).</dd> </dl> </section> </section> <sectiontitle="Costanchor="hopcount" numbered="true" toc="default"> <name>Cost Metric: Hop Count(hopcount)">(hopcount)</name> <t>Thehopcounthop count (hopcount) metric is mentioned inSection 9.2.3 of<xreftarget="RFC7285"></xref>target="RFC7285" sectionFormat="of" section="9.2.3"/> as an example. This section further clarifies its properties.</t><!-- <t><list style="hanging"> <t hangText="Metric name:"> <vspace blankLines="1"/>Hop count<vspace blankLines="1"/> </t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>hopcount<vspace blankLines="1"/> </t> </list></t> --><sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "hopcount".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value conforming to the numberspecification of Section 6 of [RFC8259].specifications provided in <xref target="RFC8259" sectionFormat="of" section="6"/>. The numberMUST<bcp14>MUST</bcp14> be a non-negative integer (greater than or equal to 0). The value represents the number of hops.</t> </section> <sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <!-- <t><list style="hanging"> <t hangText="Metric Description:"><vspace blankLines="1"/> To specify the number of hops in the path between the source endpoint and the destination endpoint. The hop count is a basic measurement of distance in a network and can be exposed as Router Hops, in direct relation to the routing protocols originating this information. </t> <t hangText="Metric Representation:"><vspace blankLines="1"/>The metric value type is a single 'JSONNumber' type value conforming to the number specification [RFC8259], Section 6. The number MUST be an integer and non-negative. <vspace blankLines="1"/></t> </list></t> --> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt> <dd>To specify the number of hops in the path from the specified source to the specified destination. The hop count is a basic measurement of distance in a network and can be exposed as the number of router hops computed from the routing protocols originating this information. A hop, however, may represent other units. The spatial aggregation level is specified in the query context (e.g., PID to PID, or endpoint toendpoint).</t> <t>Use: Thisendpoint).</dd> <dt>Use:</dt> <dd>This metric could be used as a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[ Example 4: hopcount valueresponse.</dd> </dl> <figure anchor="example-5"> <name>Hop Count Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 5)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 238 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "hopcount" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 245 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "hopcount" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 5, "ipv4:198.51.100.34": 3 } } }]]></artwork> </figure>]]></sourcecode></figure> </section><!--<sectiontitle="Measurement Considerations and Parameters"> <t>The hop count can be calculated based on the number of routers from the source endpoint through which data must pass to reach the destination endpoint. This count can be measured at the source endpoint by traceroute.</t> <t>Upon need, the traceroute can use UDP probe message or other implementations that use ICMP and TCP to discover the hop counts along the path from source endpoint to destination endpoint.</t> </section> --> <section title="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt><dd>Typically, the hop count does not have a nominalvalue.</t> <t>"sla": Typicallyvalue.</dd> <dt>"sla":</dt><dd>Typically, the hop count does not have an SLAvalue.</t> <!-- <t>"import": There can be multiple sources to import hop count, such as from IGP routing protocols. </t> --> <t>"estimation": Thevalue.</dd> <dt>"estimation":</dt><dd>The exact estimation method isout ofoutside the scope of this document. An example of estimatinghopcountshop count values is by importing from IGP routing protocols. It isRECOMMENDED<bcp14>RECOMMENDED</bcp14> that the "parameters" field of an "estimation" hop count define the meaning of ahop.</t>hop.</dd> </dl> </section> </section> </section> <section anchor="secbwmetrics"title="Throughput/Bandwidthnumbered="true" toc="default"> <name>Throughput/Bandwidth PerformanceMetrics">Metrics</name> <t>This section introducesfour throughput/bandwidththree metrics relatedmetrics.to throughput and bandwidth. Given a specified sourcetoand a specified destination, these metrics reflect the volume of traffic that the network can carry from the source to the destination.</t> <sectiontitle="Costanchor="tput" numbered="true" toc="default"> <name>Cost Metric: TCP Throughput(tput)">(tput)</name> <sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "tput".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value conforming to the numberspecification of Section 6 of [RFC8259].specifications provided in <xref target="RFC8259" sectionFormat="of" section="6"/>. The numberMUST<bcp14>MUST</bcp14> be non-negative. The unit is bytes per second.</t> </section><!-- <t><list style="hanging"> <t hangText="Metric name:"><vspace blankLines="1"/>Throughput<vspace blankLines="1"/></t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>throughput<vspace blankLines="1"/> </t> </list></t> --><sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt><dd>To give the throughput of aTCP congestion-controlcongestion control conforming TCP flow from the specified source to the specified destination. The throughputSHOULD<bcp14>SHOULD</bcp14> be interpreted as only an estimation, and the estimation is designed only for bulkflows.</t> <t>Use: Thisflows.</dd> <dt>Use:</dt> <dd>This metric could be used as a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[ Example 5: TCP throughput valueresponse.</dd> </dl> <figure anchor="example-6"> <name>TCP Throughput Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 6)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 234 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "tput" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 251 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "tput" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 256000, "ipv4:198.51.100.34": 128000 } } }]]></artwork> </figure> </section> <!-- <section title="Measurement Considerations and Parameters"> <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>See Section 3.3 of [RFC6349] for Measurement Method.<vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1 of this document.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>Similar to RTT. See Section 4.3.5 of [I-D.ietf-ippm-initial-registry] for Measurement Timing. <vspace blankLines="1"/></t> </list></t> <t>See Section 3.3 of [RFC6349] for measurement method and parameters which may be specified in "parameters". Note that the "parameters" field is an optional field providing non-normative information.</t>]]></sourcecode></figure> </section>--><sectiontitle="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt><dd>Typically, TCP throughput does not have a nominalvalue,value andSHOULD NOT<bcp14>SHOULD NOT</bcp14> begenerated.</t> <t>"sla": Typicallygenerated.</dd> <dt>"sla":</dt><dd>Typically, TCP throughput does not have an SLAvalue, and SHOULD NOT be generated.</t> <!-- <t>"import": Typically there is not a routing protocol through which one can import TCP throughput. If the import is from the IPPM framework, it is RECOMMENDED that "parameters" provides "protocol" as a fieldvalue and"ippm" as the value; see Section 4 of [I-D.ietf-ippm-initial-registry] for additional fields which can<bcp14>SHOULD NOT</bcp14> bespecified for "ippm" in "parameters". </t> --> <t>"estimation": Thegenerated.</dd> <dt>"estimation":</dt><dd>The exact estimation method isout ofoutside the scope of this document. It isRECOMMENDED<bcp14>RECOMMENDED</bcp14> that the "parameters" field of an "estimation" TCP throughput metric include the following information: (1) thecongestion-control algorithm;congestion control algorithm and (2) the estimation methodology. To specify (1), it isRECOMMENDED<bcp14>RECOMMENDED</bcp14> that the "parameters" field (object) include a field named "congestion-control-algorithm", which provides a URI for the specification of the algorithm; for example, for an ALTO server to provide estimationtoof the throughput of aCubic CongestionCUBIC congestion control flow, its "parameters"includes afield"congestion-control-algorithm",includes the "congestion-control-algorithm" field, with value being set to the URI for <xreftarget="I-D.ietf-tcpm-rfc8312bis"></xref>;target="RFC9438" format="default"/>; for an ongoing congestion control algorithm such as BBR, aalink to itsspecification.specification can be added. To specify (2), the "parameters" field includes as many details as possible; for example, for the TCP Cubic throughout estimation, the "parameters" field specifies that the throughput is estimated by setting _C_ to 0.4, and theEquationequation in <xref target="RFC9438" sectionFormat="comma" section="5.1"/>, Figure 8of <xref target="I-D.ietf-tcpm-rfc8312bis"></xref>is applied; as an alternative, the methodology may be based on the NUM model <xreftarget="Prophet"></xref>,target="Prophet" format="default"/> or theG2model described in <xreftarget="G2"></xref>.target="G2" format="default"/>. The exact specification of theparameters"parameters" field isout ofoutside the scope of thisdocument.</t>document.</dd> </dl> </section> </section><!-- TCP Throughput --><sectiontitle="Costanchor="bwresidual" numbered="true" toc="default"> <name>Cost Metric: Residual Bandwidth(bw-residual)"> <!-- <t><list style="hanging"> <t hangText="Metric name:"><vspace blankLines="1"/>Residual Bandwidth<vspace blankLines="1"/></t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>residualbw<vspace blankLines="1"/> </t> </list></t> -->(bw-residual)</name> <sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "bw-residual".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value that is non-negative. The unit of measurement is bytes per second.</t> </section> <sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt><dd><t>To specify temporal and spatial residual bandwidth from the specified sourceandto the specified destination. The base semantics of the metric is the Unidirectional Residual Bandwidth metric as defined in[RFC8571,RFC8570,RFC7471],<xref target="RFC8571"/>, <xref target="RFC8570"/>, and <xref target="RFC7471"/>, but instead of specifying the residual bandwidth for a link, it is the residual bandwidth of the path from the source to the destination. Hence, it is the minimal residual bandwidth among all links from the source to the destination. When the max statistical operator is defined for the metric, it typically provides the minimum of the link capacities along the path, as the default value of the residual bandwidth of a link is its link capacity[RFC8571,8570,7471].<xref target="RFC8571"/> <xref target="RFC8570"/> <xref target="RFC7471"/>. The spatial aggregation unit is specified in the query context (e.g., PID to PID, or endpoint to endpoint).</t> <t>The default statistical operator for residual bandwidth is the current instantaneous sample; that is, the default is assumed to be "cur".</t><t>Use: This</dd> <dt>Use:</dt> <dd>This metric could be usedeitheras a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[ Example 7: bw-residual valueresponse.</dd> </dl> <figure anchor="example-7"> <name>Residual Bandwidth Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 7)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 241 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "bw-residual" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 255 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "bw-residual" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 0, "ipv4:198.51.100.34": 2000 } } }]]></artwork> </figure> </section> <!-- <section title="Measurement Considerations and Parameters"> <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>residual Bandwidth is the Unidirectional residual bandwidth measured between two directly connected IS-IS neighbors or OSPF neighbors. See Section 4.5 of [RFC7810] for Measurement Method. <vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1 of this document.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See Section 5 of [RFC7810] for Measurement Timing.<vspace blankLines="1"/></t> </list></t>]]></sourcecode></figure> </section>--><sectiontitle="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt><dd>Typically, residual bandwidth does not have a nominalvalue.</t> <t>"sla": Typicallyvalue.</dd> <dt>"sla":</dt><dd>Typically, residual bandwidth does not have an"sla" value.</t> <!-- <t>"import": There can be multiple sources to import residual bandwidth. If the import is from [RFC8571] (by using unidirectional residual bandwidth), it is RECOMMENDED that "parameters" provides "protocol" as a field and "RFC8571" as the value. The server should be cognizant of issues when computing end-to-end summary statistics from link statistics. For example, the min of the end-to-end path residual bandwidth is the min of all links on the path. </t> --> <t>"estimation": SeeSLA value.</dd> <dt>"estimation":</dt><dd>See the "estimation" entry inSection 4.1.4 on aggregation of routing protocol link metrics.<xref target="ccspec-ow"/>. The current ("cur") residual bandwidth of a path is the minimalof theresidual bandwidth of all links on thepath.</t>path.</dd> </dl> </section> </section><!-- residual bandwidth --><sectiontitle="Costanchor="bwavailable" numbered="true" toc="default"> <name>Cost Metric: Available Bandwidth(bw-available)"> <!-- <t><list style="hanging"> <t hangText="Metric name:"><vspace blankLines="1"/>Maximum Reservable Bandwidth<vspace blankLines="1"/></t> <t hangText="Metric Identifier:"> <vspace blankLines="1"/>maxresbw<vspace blankLines="1"/> </t> </list></t> -->(bw-available)</name> <sectiontitle="Base Identifier">numbered="true" toc="default"> <name>Base Identifier</name> <t>The base identifier for this performance metric is "bw-available".</t> </section> <sectiontitle="Value Representation">numbered="true" toc="default"> <name>Value Representation</name> <t>The metric value type is a single 'JSONNumber' type value that is non-negative. The unit of measurement is bytes per second.</t> </section> <sectiontitle="Intendednumbered="true" toc="default"> <name>Intended Semantics andUse"> <t>Intended Semantics: ToUse</name> <dl> <dt>Intended Semantics:</dt><dd><t>To specify temporal and spatial available bandwidth from the specified source to the specified destination. The base semantics of the metric is the Unidirectional Available Bandwidth metric as defined in[RFC8571,RFC8570,RFC7471],<xref target="RFC8571"/>, <xref target="RFC8570"/>, and <xref target="RFC7471"/>, but instead of specifying the available bandwidth for a link, it is the available bandwidth of the path from the source to the destination. Hence, it is the minimal available bandwidth among all links from the source to thedestination.Thedestination. The spatial aggregation unit is specified in the query context (e.g., PID to PID, or endpoint to endpoint).</t> <t>The default statistical operator for available bandwidth is the current instantaneous sample; that is, the default is assumed to be "cur".</t><t>Use: This</dd> <dt>Use:</dt><dd>This metric could be usedeitheras a cost metric constraint attribute or as a returned cost metric in theresponse.</t> <figure> <artwork><![CDATA[ Example 8: bw-available valueresponse.</dd> </dl> <figure anchor="example-8"> <name>Available Bandwidth Value onsource-destination endpoint pairsSource-Destination Endpoint Pairs (Example 8)</name> <sourcecode type="json"><![CDATA[ POST /endpointcost/lookup HTTP/1.1 Host: alto.example.com Content-Length: 244 Content-Type: application/alto-endpointcostparams+json Accept: application/alto-endpointcost+json,application/alto-error+json { "cost-type": { "cost-mode": "numerical", "cost-metric": "bw-available" }, "endpoints": { "srcs": [ "ipv4:192.0.2.2" ], "dsts": [ "ipv4:192.0.2.89", "ipv4:198.51.100.34" ] } }]]></artwork> </figure> <figure> <artwork><![CDATA[HTTP/1.1 200 OK Content-Length: 255 Content-Type: application/alto-endpointcost+json { "meta": { "cost-type": { "cost-mode": "numerical", "cost-metric": "bw-available" } }, "endpoint-cost-map": { "ipv4:192.0.2.2": { "ipv4:192.0.2.89": 0, "ipv4:198.51.100.34": 2000 } } }]]></artwork> </figure> </section> <!-- <section title="Measurement Considerations and Parameters"> <t><list style="hanging"> <t hangText="Method of Measurement or Calculation:"><vspace blankLines="1"/>Maximum Reservable Bandwidth is the bandwidth measured between two directly connected IS-IS neighbors or OSPF neighbors. See Section 3.5 of [RFC5305] for Measurement Method.<vspace blankLines="1"/></t> <t hangText="Measurement Point(s) with Potential Measurement Domain:"><vspace blankLines="1"/>See Section 4.1 this document for discussions.<vspace blankLines="1"/></t> <t hangText="Measurement Timing:"><vspace blankLines="1"/>See Section 3.5 of [RFC5305] and Section 5 of [RFC7810] for Measurement Timing.<vspace blankLines="1"/></t> </list></t>]]></sourcecode></figure> </section>--><sectiontitle="Cost-Contextnumbered="true" toc="default"> <name>Cost-Context SpecificationConsiderations"> <t>"nominal": TypicallyConsiderations</name> <dl> <dt>"nominal":</dt><dd>Typically, available bandwidth does not have a nominalvalue.</t> <t>"sla": Typicallyvalue.</dd> <dt>"sla":</dt><dd>Typically, available bandwidth does not have an"sla" value.</t> <!-- <t>"import": There can be multiple sources to import maximum reservable bandwidth. For example, Maximum reservable bandwidth is defined by IS-IS/OSPF TE, and measures the reservable bandwidth between two directly connected IS-IS neighbors or OSPF neighbors; see Section 3.5 of [RFC5305]. If the import is from [RFC8571] (by using unidirectional maximum reservable bandwidth), it is RECOMMENDED that "parameters" provides "protocol" as a field and "RFC8571" as the value. </t> --> <t>"estimation": SeeSLA value.</dd> <dt>"estimation":</dt><dd>See the "estimation" entry inSection 4.1.4 on aggregation of routing protocol link metrics.<xref target="ccspec-ow"/>. The current ("cur") available bandwidth of a path is the minimum of the available bandwidth of all links on thepath.</t>path.</dd> </dl> </section><!-- cc consider --></section><!-- end of available bw --></section> <section anchor="secopconsider"title="Operational Considerations">numbered="true" toc="default"> <name>Operational Considerations</name> <t>The exact measurement infrastructure, measurementcondition,conditions, and computation algorithms can varyfrombetween differentnetworks,networks and are outside the scope of this document. Both the ALTO server and the ALTO clients, however, need to be cognizant of the operational issues discussed in the followingsub-sections.</t>subsections.</t> <t>Also, the performance metrics specified in this document aresimilar,similar in that they may use similar data sources and have similar issues in their calculation. Hence, this document specifiescommonissuesunless one metric has its unique challenges.</t>that the performance metrics might have in common and also discusses challenges regarding the computation of ALTO performance metrics (<xref target="comp-consider"/>).</t> <sectiontitle="Source Considerations">numbered="true" toc="default"> <name>Source Considerations</name> <t>The addition of the "cost-source" fieldis to solvesolves a key issue:Anan ALTO server needs data sources to compute the cost metrics described in this document, and an ALTO client needs to know the data sources to better interpret the values.</t> <t>To avoid information that is toofine-grained information,fine grained, this document introduces "cost-source" to indicate only the high-leveltypetypes of data sources: "estimation","nominal""nominal", or"lsa","sla", where "estimation" is a type of measurement data source, "nominal" is a type of static configuration, and "sla" is a type that ismorebased more on policy.</t> <t>Forestimation, forexample, for "estimation", the ALTO server may use log servers or theOAMOperations, Administration, and Maintenance (OAM) system as its datasourcesource, as recommended by <xreftarget="RFC7971"></xref>.target="RFC7971" format="default"/>. In particular, the cost metrics defined in this document can be computed using routing systems as the data sources.</t><!-- Mechanisms defined in [RFC2681], [RFC3393], [RFC7679], [RFC7680], [RFC3630], [RFC3784], [RFC7471], [RFC7810], [RFC7752] and [I-D.ietf-idr-te-pm-bgp] that allow an ALTO Server to retrieve and derive the necessary information to compute the metrics that we describe in this document.</t> --></section> <sectiontitle="Metricnumbered="true" toc="default"> <name>Metric TimestampConsideration ">Considerations</name> <t>Despite the introduction of the additionalcost-context"cost-context" information, the metrics do not have a field to indicate the timestamps of the data used to compute the metrics. To indicate this attribute, the ALTO serverSHOULD<bcp14>SHOULD</bcp14> return an HTTP"Last-Modified",Last-Modified value to indicate the freshness of the data used to compute the performance metrics.</t> <t>If the ALTO client obtains updates through an incremental update mechanism <xreftarget="RFC8895"></xref>,target="RFC8895" format="default"/>, the clientSHOULD<bcp14>SHOULD</bcp14> assume that the metric is computed using a snapshot at the time that is approximated by the receiving time.</t> </section> <sectiontitle="Backward Compatibility Considerations">numbered="true" toc="default"> <name>Backward-Compatibility Considerations</name> <t>One potential issue introduced by the optional "cost-source" field is backward compatibility. Considerthatthe case where an IRDwhichdefines twocost-types"cost-type" entries with the same "cost-mode" and "cost-metric", but one with "cost-source" being "estimation" and the other being "sla".ThenIn such a case, an ALTO client that is not aware of the extension will not be able to distinguish between these two types. A similar issue can arise even with a singlecost-type,"cost-type" whose "cost-source" is "sla": an ALTO client that is not aware of this extension will ignore this field and instead consider the metric estimation.</t> <t>To address the backward-compatibility issue, if a "cost-metric" is "routingcost" and the metric contains a "cost-context" field, then itMUST<bcp14>MUST</bcp14> be "estimation"; if it is not, the clientSHOULD<bcp14>SHOULD</bcp14> reject the information as invalid.</t> </section> <sectiontitle="Computation Considerations">anchor="comp-consider" numbered="true" toc="default"> <name>Computation Considerations</name> <t>The metric values exposed by an ALTO server may result from additional processingonof measurements from data sources to compute exposed metrics. This may involve data processing tasks such as aggregating the results across multiple systems, removing outliers, and creating additional statistics.There are two challenges on theThe computation of ALTO performancemetrics.</t>metrics can present two challenges.</t> <sectiontitle="Configuration Parameters Considerations">numbered="true" toc="default"> <name>Configuration Parameter Considerations</name> <t>Performance metrics often depend on configuration parameters, and exposing such configuration parameters can help an ALTO client to better understand the exposed metrics. In particular, an ALTO server may be configured to compute a TE metric (e.g., packet loss rate)inat fixed intervals, say every T seconds. To expose this information, the ALTO server may provide the client with two pieces of additional information: (1) when the metricsarewere lastcomputed,computed and (2) when the metrics will be updated (i.e., the validity period of the exposed metric values). The ALTO server can expose these two pieces of information by using the HTTP response headers Last-Modified and Expires.</t> </section> <sectiontitle="Aggregationnumbered="true" toc="default"> <name>Aggregation ComputationConsiderations">Considerations</name> <t>An ALTO server may not be able to measure the performance metrics to be exposed. The basic issue is that the "source" information can often belink level.link-level information. For example, routing protocols often measure and report onlyper link loss,per-link loss and not end-to-end loss; similarly, routing protocols reportlink levellink-level availablebandwidth,bandwidth and not end-to-end available bandwidth. The ALTO server then needs to aggregate these data to provide an abstract and unified view that can be more useful to applications. The server shouldconsiderbe aware that different metrics may use different aggregationcomputation.computations. For example, the end-to-end latency of a path is the sum of thelatencylatencies of the links on the path; the end-to-end available bandwidth of a path is the minimum of the available bandwidth of the links on the path; in contrast, aggregating loss values is complicated by the potential for correlated loss events on different links in thepath</t>path.</t> </section> </section> </section> <section anchor="secsecconsider"title="Security Considerations">numbered="true" toc="default"> <name>Security Considerations</name> <t>The properties defined in this document present no security considerations beyond those in Section15<xref target="RFC7285" section="15" sectionFormat="bare"/> of the base ALTO specification <xreftarget="RFC7285"></xref>.</t>target="RFC7285" format="default"/>.</t> <t>However, concerns addressed inSections 15.1, 15.2,Sections <xref target="RFC7285" section="15.1" sectionFormat="bare"/>, <xref target="RFC7285" section="15.2" sectionFormat="bare"/>, and15.3<xref target="RFC7285" section="15.3" sectionFormat="bare"/> of <xreftarget="RFC7285"></xref>target="RFC7285" format="default"/> remain of utmost importance. Indeed,Traffic Engineering (TE)TE performance is highly sensitive ISP information; therefore, sharing TE metric values in numerical mode requires full mutual confidence between the entities managing the ALTO server and the ALTO client. ALTO servers will most likely distribute numerical TE performance to ALTO clients under strict and formal mutual trust agreements. On the other hand, ALTO clients must be cognizantonof the risks attached to such information that they would have acquired outside formal conditions of mutual trust.</t> <t>To mitigate confidentiality risks during information transport of TE performance metrics, the operator should address the risk of ALTO information being leaked to maliciousClientsclients or thirdparties,parties throughattackssuch attacks astheperson-in-the-middle (PITM) attacks. As specified in"Protection Strategies" (Section 15.3.2Section <xref target="RFC7285" section="15.3.2" sectionFormat="bare">"Protection Strategies"</xref> of <xreftarget="RFC7285"></xref>),target="RFC7285"/>, the ALTOServerserver should authenticate ALTOClientsclients when transmitting an ALTO information resource containing sensitive TE performance metrics."AuthenticationSection <xref target="RFC7285" section="8.3.5" sectionFormat="bare">"Authentication andEncryption" (Section 8.3.5Encryption"</xref> of <xreftarget="RFC7285"></xref>)target="RFC7285"/> specifies that"ALTO ServerALTO server implementations as well as ALTOClientclient implementationsMUST<bcp14>MUST</bcp14> support the "https" URI schemeof<xreftarget="RFC7230"></xref>target="RFC9110" format="default"/> and Transport Layer Security (TLS)of<xreftarget="RFC8446"></xref>".</t>target="RFC8446" format="default"/>.</t> </section> <section anchor="ianaconsider"title="IANA Considerations">numbered="true" toc="default"> <name>IANA Considerations</name> <section> <name>ALTO Cost Metrics Registry</name> <t>IANAhascreated and now maintains the "ALTO CostMetric"Metrics" registry, as listed inSection 14.2, Table 3 of<xreftarget="RFC7285"></xref>.target="RFC7285" section="14.2" sectionFormat="comma"/>, Table 3. This registry is located at<https://www.iana.org/assignments/alto-protocol/alto-protocol.xhtml#cost-metrics>. This document requests to add<eref target="https://www.iana.org/assignments/alto-protocol/" brackets="angle"/>. IANA has added the following entries to the“ALTO"ALTO CostMetric”Metrics" registry.</t><figure> <artwork><![CDATA[ +-----------------+----------------------------+ | Identifier | Intended Semantics | +-----------------+----------------------------+ | delay-ow | Section 4.1 of [RFCXXX] | | delay-rt | Section 4.2 of [RFCXXX] | | delay-variation | Section 4.3 of [RFCXXX] | | lossrate | Section 4.4 of [RFCXXX] | | hopcount | Section 4.5 of [RFCXXX] | | tput | Section 5.1 of [RFCXXX] | | bw-residual | Section 5.2 of [RFCXXX] | | bw-available | Section 5.3 of [RFCXXX] | +-----------------+----------------------------+ ]]></artwork> </figure> <t><list style="symbols"> <t>[Note to the RFC Editor]: Please replace RFCXXX with the RFC number assigned to this document.</t> </list></t> <t>This document requests the creation of<table> <name>ALTO Cost Metrics Registry</name> <thead> <tr> <th>Identifier</th> <th>Intended Semantics</th> <th>Reference</th> </tr> </thead> <tbody> <tr> <td>delay-ow</td> <td>See <xref target="oneway"/></td> <td>RFC 9439</td> </tr><tr> <td>delay-rt</td> <td>See <xref target="delayrt"/></td> <td>RFC 9439</td> </tr><tr> <td>delay-variation</td> <td>See <xref target="delayvar"/></td> <td>RFC 9439</td> </tr><tr> <td>lossrate</td> <td>See <xref target="lossrate"/></td> <td>RFC 9439</td> </tr><tr> <td>hopcount</td> <td>See <xref target="hopcount"/></td> <td>RFC 9439</td> </tr><tr> <td>tput</td> <td>See <xref target="tput"/></td> <td>RFC 9439</td> </tr><tr> <td>bw-residual</td> <td>See <xref target="bwresidual"/></td> <td>RFC 9439</td> </tr><tr> <td>bw-available</td> <td>See <xref target="bwavailable"/></td> <td>RFC 9439</td> </tr> </tbody> </table> </section> <section> <name>ALTO Cost Source Types Registry</name> <t>IANA has created the "ALTO CostSource"Source Types" registry. This registry serves two purposes. First, it ensures the uniqueness of identifiers referring to ALTO cost source types. Second, it provides references to particular semantics of allocated cost source types to be applied by both ALTO servers and applications utilizing ALTO clients.</t> <t>A new ALTO cost source type can be added after IETF Review <xreftarget="RFC8126"></xref>,target="RFC8126" format="default"/>, to ensure that proper documentation regarding the new ALTO cost source type and its security considerationshavehas been provided. The RFC(s) documenting the new cost source type should be detailed enough to provide guidance to both ALTO service providers and applications utilizing ALTO clients as to how values of the registered ALTO cost source type should be interpreted. Updates and deletions of ALTO cost source types follow the same procedure.</t> <t>Registered ALTO address type identifiersMUST<bcp14>MUST</bcp14> conform to the syntactical requirements specified inSection 3.1.<xref target="meta" sectionFormat="bare"/>. Identifiers are to be recorded and displayed as strings.</t> <t>Requests to add a new value to the registryMUST<bcp14>MUST</bcp14> include the following information:<list style="symbols"> <t>Identifier: The</t> <dl> <dt>Identifier:</dt><dd>The name of the desired ALTO cost sourcetype.</t> <t>Intended Semantics:type.</dd> <dt>Intended Semantics:</dt><dd> ALTO cost sourcetypetypes carry with them semantics to guide their usage by ALTO clients. Hence, a document defining a new type should provide guidance to both ALTO service providers and applications utilizing ALTO clients as to how values of the registered ALTO endpoint property should beinterpreted.</t> <t>Security Considerations:interpreted.</dd> <dt>Security Considerations:</dt><dd> ALTO cost source types expose information to ALTO clients. ALTO service providers should be made aware of the security ramifications related to the exposure of a cost sourcetype.</t> </list></t> <t>This specification requests registration oftype.</dd> </dl> <t>IANA has registered the identifiers "nominal", "sla", and "estimation" as listed in the tablebelow. Semantics for the these are documented in Section 3.1, and security considerations are documented in Section 7.</t> <figure> <artwork><![CDATA[ +------------+----------------------------------+----------------+ | Identifier | Intended Semantics | Security | | | | Considerations | +------------+----------------------------------+----------------+ | nominal | Valuesbelow.</t> <table> <name>ALTO Cost Source Types Registry</name> <thead> <tr> <th>Identifier</th> <th>Intended Semantics</th> <th>Security Considerations</th> <th>Reference</th> </tr> </thead> <tbody> <tr> <td>nominal</td> <td>Values in nominalcases; | Section 7 of | | | Section 3.1 of [RFCXXX] | [RFCXXX] | | sla | Valuescases (<xref target="meta"/>)</td> <td><xref target="secsecconsider"/></td> <td>RFC 9439</td> </tr> <tr> <td>sla</td> <td>Values reflectingservice level | Section 7 of | | | agreement; Section 3.1 of | [RFCXXX] | | | [RFCXXXX] | | | estimation | ValuesService Level Agreement (<xref target="meta"/>)</td> <td><xref target="secsecconsider"/></td> <td>RFC 9439</td> </tr> <tr> <td>estimation</td> <td>Values byestimation; | Section 7 of | | | Section 3.1 of [RFCXXX] | [RFCXXX] | +------------+----------------------------------+----------------+ ]]></artwork> </figure>estimation (<xref target="meta"/>)</td> <td><xref target="secsecconsider"/></td> <td>RFC 9439</td> </tr> </tbody> </table> </section><section title="Acknowledgments"> <t>The authors of this document would also like to thank Martin Duke for the highly informative, thorough AD reviews and comments. We thank Christian Amsüss, Elwyn Davies, Haizhou Du, Kai Gao, Geng Li, Lili Liu, Danny Alex Lachos Perez, and Brian Trammell for the reviews and comments. We thank Benjamin Kaduk, Eric Kline, Francesca Palombini, Lars Eggert, Martin Vigoureux, Murrary Kucherawy, Roman Danyliw, Zaheduzzaman Sarker, Éric Vyncke for discussions and comments that improve this document.</t></section> </middle> <back><references title="Normative References"> <!-- <?rfc include="reference.RFC.5234.xml"?> <?rfc include="reference.RFC.4627.xml"?> <?rfc include="reference.RFC.7471.xml"?> <?rfc include="reference.RFC.7752.xml"?> <?rfc include="reference.RFC.7810.xml"?> <?rfc include="reference.RFC.7680.xml"?> <?rfc include="reference.RFC.2679.xml"?> <?rfc include="reference.RFC.2681.xml"?> <?rfc include="reference.RFC.3393.xml"?> <?rfc include="reference.RFC.5305.xml"?> <?rfc include="reference.RFC.6349.xml"?> <?rfc include="reference.RFC.7679.xml"?> <?rfc include="reference.RFC.8571.xml"?> <?rfc include="reference.I-D.ietf-idr-te-pm-bgp.xml"?> <?rfc include="reference.I-D.ietf-ippm-initial-registry.xml"?> <?rfc include="reference.RFC.2818.xml"?> --> <!-- requirements words --> <?rfc include="reference.RFC.2119.xml"?> <!-- TE for OSPF --> <?rfc include="reference.RFC.3630.xml"?> <!-- TE for ISIS --> <?rfc include="reference.RFC.5305.xml"?> <!-- guidelines on new metrics --> <?rfc include="reference.RFC.6390.xml"?> <!-- https change rfc 2818 to 7230 as 2818 is informational--> <?rfc include="reference.RFC.7230.xml"?> <!-- alto base --> <?rfc include="reference.RFC.7285.xml"?> <!-- OSPF TE metrics --> <?rfc include="reference.RFC.7471.xml"?> <!-- iana --> <?rfc include="reference.RFC.8126.xml"?> <!-- requirement words --> <?rfc include="reference.RFC.8174.xml"?> <!-- JSON Data--> <?rfc include="reference.RFC.8259.xml"?> <!-- TLS 1.3 --> <?rfc include="reference.RFC.8446.xml"?> <!-- ISIS TE --> <?rfc include="reference.RFC.8570.xml"?> <!-- BGP-LS --> <?rfc include="reference.RFC.8571.xml"?><displayreference target="I-D.corre-quic-throughput-testing" to="QUIC-THROUGHPUT-TESTING"/> <references> <name>References</name> <references> <name>Normative References</name> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.2119.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.3630.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.5305.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.6390.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7285.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7471.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8126.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8174.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8259.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8446.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8570.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8571.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.8895.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.9110.xml"/> <reference anchor="IANA-IPPM" target="https://www.iana.org/assignments/performance-metrics/"> <front> <title>Performance Metrics</title> <author> <organization>IANA</organization> </author> <date/> </front> </reference> <!--ALTO SSEdraft-ietf-tcpm-rfc8312bis (RFC 9438) (AUTH48-DONE) --><?rfc include="reference.RFC.8895.xml"?><referenceanchor="IANA-IPPM">anchor='RFC9438' target='https://www.rfc-editor.org/info/rfc9438'> <front><title>Performance Metrics Registry, https://www.iana.org/assignments/performance-metrics/performance-metrics.xhtml</title><title>CUBIC for Fast and Long-Distance Networks</title> <author fullname="Lisong Xu"> </author> <author fullname="Sangtae Ha"> </author> <author fullname="Injong Rhee"> </author> <authorinitials="" surname="IANA"></author>fullname="Vidhi Goel"> </author> <author fullname="Lars Eggert" role="editor"> </author> <dateyear="" />month="August" year="2023"/> </front> <seriesInfo name="RFC" value="9438"/> <seriesInfo name="DOI" value="10.17487/RFC9438"/> </reference><?rfc include="reference.I-D.ietf-tcpm-rfc8312bis.xml"?></references><references title="Informative References"> <?rfc include="reference.RFC.2330.xml"?> <!-- IPPM Framework --> <?rfc include="reference.RFC.2681.xml"?> <!-- IPPM Round Trip Delay --> <?rfc include="reference.RFC.3393.xml"?> <!-- IPPM Packet Delay Variation --> <?rfc include="reference.RFC.5357.xml"?> <!-- TWAMP --> <?rfc include="reference.RFC.7679.xml"?> <!-- IPPM One Way Delay --> <?rfc include="reference.RFC.7971.xml"?> <?rfc include='reference.I-D.corre-quic-throughput-testing'?> <?rfc include='reference.RFC.9000'?> <!-- ALTO requirements --><references> <name>Informative References</name> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.2330.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.2681.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.3393.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.5357.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7679.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.7971.xml"/> <!-- draft-corre-quic-throughput-testing (Expired) --> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.corre-quic-throughput-testing.xml"/> <xi:include href="https://bib.ietf.org/public/rfc/bibxml/reference.RFC.9000.xml"/> <referenceanchor="G2">anchor="G2" target="https://dl.acm.org/doi/10.1145/3366707"> <front> <title>On the Bottleneck Structure of Congestion-Controlled Networks</title> <author initials="J"surname="Ros-Giralt"></author>surname="Ros-Giralt"/> <author initials="A"surname="Bohara"></author>surname="Bohara"/> <author initials="S"surname="Yellamraju"></author>surname="Yellamraju"/> <author initials="M" surname="Harper Langston"/> <author initials="R" surname="Lethin"/> <authorinitials="" surname="et. al."></author>initials="Y" surname="Jiang"/> <author initials="L" surname="Tassiulas"/> <author initials="J" surname="Li"/> <author initials="Y" surname="Tan"/> <author initials="M" surname="Veeraraghavan"/> <dateyear="2020" />month="December" year="2019"/> </front> <refcontent>Proceedings of the ACM on Measurement and Analysis of Computing Systems, Vol. 3, No. 3, Article No. 59, pp. 1-31</refcontent> <seriesInfoname="ACM SIGMETRICS" value="2019" />name="DOI" value="10.1145/3366707"/> </reference> <referenceanchor="FlowDirector">anchor="FlowDirector" target=""> <front> <title>Steering Hyper-Giants' Traffic at Scale</title> <author initials="E"surname="Pujol"></author>surname="Pujol"/> <author initials="I"surname="Poese"></author>surname="Poese"/> <author initials="J"surname="Zerwas"></author>surname="Zerwas"/> <author initials="G"surname="Smaragdakis"></author>surname="Smaragdakis"/> <author initials="A"surname="Feldmann"></author>surname="Feldmann"/> <dateyear="2020" />month="December" year="2019"/> </front><seriesInfo name="ACM CoNEXT" value="2020" /><refcontent>ACM CoNEXT '19</refcontent> </reference> <referenceanchor="Prometheus">anchor="Prometheus" target=""> <front> <title>Prometheus: A Next-Generation MonitoringSystem</title>System (Talk)</title> <author initials="J"surname="Volz"></author>surname="Volz"/> <author initials="B"surname="Rabenstein"></author>surname="Rabenstein"/> <dateyear="2015" />month="May" year="2015"/> </front> <refcontent>SREcon15 Europe</refcontent> </reference> <referenceanchor="Prophet">anchor="Prophet" target="https://dl.acm.org/doi/10.1109/TNET.2020.3016838"> <front> <title>Prophet: Toward Fast,AccurateError-Tolerant Model-Based Throughput Predictionwithfor ReactiveFlows</title> <author initials="K" surname="Gao"></author>Flows in DC Networks</title> <author initials="J"surname="Zhang"></author>surname="Zhang"/> <author initials="K" surname="Gao"/> <author initials="YR"surname="Yang"></author>surname="Yang"/> <author initials="J" surname="Bi"/> <dateyear="2020" />month="December" year="2020"/> </front><seriesInfo name="ACM/IEEE<refcontent>IEEE/ACM Transactions onNetworking" value="July" />Networking, Volume 28, Issue 601, pp. 2475-2488</refcontent> </reference> </references> </references> <section numbered="false" toc="default"> <name>Acknowledgments</name> <t>The authors of this document would like to thank <contact fullname="Martin Duke"/> for the highly informative, thorough AD reviews and comments. We thank <contact fullname="Christian Amsüss"/>, <contact fullname="Elwyn Davies"/>, <contact fullname="Haizhou Du"/>, <contact fullname="Kai Gao"/>, <contact fullname="Geng Li"/>, <contact fullname="Lili Liu"/>, <contact fullname="Danny Alex Lachos Perez"/>, and <contact fullname="Brian Trammell"/> for their reviews and comments. We thank <contact fullname="Benjamin Kaduk"/>, <contact fullname="Erik Kline"/>, <contact fullname="Francesca Palombini"/>, <contact fullname="Lars Eggert"/>, <contact fullname="Martin Vigoureux"/>, <contact fullname="Murray Kucherawy"/>, <contact fullname="Roman Danyliw"/>, <contact fullname="Zaheduzzaman Sarker"/>, and <contact fullname="Éric Vyncke"/> for discussions and comments that improved this document.</t> </section> </back> </rfc>