Netflix

ietf@tenghardt.net

ETH Zürich

cyrill.kraehenbuehl@inf.ethz.ch

IRTF PANRG Internet-Draft Path properties express information about paths across a network and the services provided via such paths. In a path-aware network, path properties may be fully or partially available to entities such as hosts. This document defines and categorizes path properties. Furthermore, the document specifies several path properties which might be useful to hosts or other entities, e.g., for selecting between paths or for invoking some of the provided services.

The current Internet architecture does not explicitly support endpoint discovery of forwarding paths through the network as well as the discovery of properties and services associated with these paths. Path-aware networking, as defined in Section 1.1 of , describes “endpoint discovery of the properties of paths they use for communication across an internetwork, and endpoint reaction to these properties that affects routing and/or data transfer”. This document provides a generic definition of path properties, addressing the first of the questions in path-aware networking . As terms related to paths have been used with different meanings in different areas of networking, first, this document provides a common terminology to define paths, path elements, and flows. Based on these terms, the document defines path properties. Then, this document provides some examples of use cases for path properties. Finally, the document lists several path properties that may be useful for the mentioned use cases. Note that this document does not assume that any of the listed path properties are actually available to any entity. The question of how entities can discover and distribute path properties in a trustworthy way is out of scope for this document.

An entity which processes packets, e.g., sends, receives, forwards, or modifies them. A node may be physical or virtual, e.g., a physical device or a service function provided as a virtual element. A node may also be the collection of multiple entities which, as a collection, processes packets, e.g., an entire Autonomous System (AS). A node that generally executes application programs on behalf of user(s), employing network and/or Internet communication services in support of this function, as defined in . Note that hosts include both client nodes (e.g., running a web browser) and server nodes (e.g., running a web server). A medium or communication facility that connects two or more nodes with each other. A link enables a node to send packets to other nodes. Links can be physical, e.g., a Wi-Fi network which connects an Access Point to stations, or virtual, e.g., a virtual switch which connects two virtual machines hosted on the same physical machine. A link is unidirectional. As such, bidirectional communication can be modeled as two links between the same nodes in opposite directions. Either a node or a link. For example, a path element can be an Abstract Network Element (ANE) as defined in . A sequence of adjacent path elements over which a packet can be transmitted, starting and ending with a node. A path is unidirectional. Paths are time-dependent, i.e., the sequence of path elements over which packets are sent from one node to another may change. A path is defined between two nodes. For multicast or broadcast, a packet may be sent by one node and received by multiple nodes. In this case, the packet is sent over multiple paths at once, one path for each combination of sending and receiving node; these paths do not have to be disjoint. Note that an entity may have only partial visibility of the path elements that comprise a path and visibility may change over time. Different entities may have different visibility of a path and/or treat path elements at different levels of abstraction. For example, a path may be given as a sequence of physical nodes and the links connecting these nodes, or it may be given as a sequence of logical nodes such as a sequence of ASes or an Explicit Route Object (ERO). Similarly, the representation of a path and its properties, as it is known to a specific entity, may be more complex and include details about the physical layer technology, or it may be more abstract and only consist of a specific source and destination which is known to be reachable from that source. The path that is used by a remote node in the context of bidirectional communication. Given a path, a subpath is a sequence of adjacent path elements of this path. An entity made of packets to which the traits of a path or set of subpaths may be applied in a functional sense. For example, a flow can consist of all packets sent within a TCP session with the same five-tuple between two hosts, or it can consist of all packets sent on the same physical link. A trait of one or a sequence of path elements, or a trait of a flow with respect to one or a sequence of path elements. An example of a link property is the maximum data rate that can be sent over the link. An example of a node property is the administrative domain that the node belongs to. An example of a property of a flow with respect to a subpath is the aggregated one-way delay of the flow being sent from one node to another node over this subpath. A property is thus described by a tuple containing the path element(s), the flow or an empty set if no packets are relevant for the property, the name of the property (e.g., maximum data rate), and the value of the property (e.g., 1Gbps). A collection of multiple values of a property into a single value, according to a function. A property can be aggregated over multiple path elements (i.e., a subpath), e.g., the MTU of a path as the minimum MTU of all links on the path, over multiple packets (i.e., a flow), e.g., the median one-way latency of all packets between two nodes, or over both, e.g., the mean of the queueing delays of a flow on all nodes along a path. The aggregation function can be numerical, e.g., median, sum, minimum, it can be logical, e.g., “true if all are true”, “true if at least 50\% of values are true”, or an arbitrary function which maps multiple input values to an output value. A property that is observed for a specific path element, subpath, or path, e.g., using measurements. For example, the one-way delay of a specific packet transmitted from one node to another node can be measured. An approximate calculation or assessment of the value of a property. An assessed property includes the reliability of the calculation or assessment. The notion of reliability depends on the property. For example, a path property based on an approximate calculation may describe the expected median one-way latency of packets sent on a path within the next second, including the confidence level and interval. A non-numerical assessment may instead include the likelihood that the property holds.

The terminology defined in this document is intended to be general and applicable to existing and future path-aware technologies. Using this terminology, a path-aware technology can define and consider specific path elements and path properties on a specific level of abstraction. For instance, a technology may define path elements as IP routers, e.g., in source routing (). Alternatively, it may consider path elements on a different layer of the Internet Architecture () or as a collection of entities not tied to a specific layer, such as an AS or an ERO.

When a path-aware network exposes path properties to hosts or other entities, these entities may use this information to achieve different goals. This section lists several use cases for path properties. Note that this is not an exhaustive list, as with every new technology and protocol, novel use cases may emerge, and new path properties may become relevant. Moreover, for any particular technology, entities may have visibility of and control over different path elements and path properties, and consider them on different levels of abstraction. Therefore, a new technology may implement an existing use case related to different path elements or on a different level of abstraction.

Entities can choose what traffic to send over which path or subset of paths. A node might be able to select between a set of paths, either if there are several paths to the same destination (e.g., in case of a mobile device with two wireless interfaces, both providing a path), or if there are several destinations, and thus several paths, providing the same service (e.g., Application-Layer Traffic Optimization (ALTO) , an application layer peer-to-peer protocol allowing hosts a better-than-random peer selection). Care needs to be taken when selecting paths based on path properties, as path properties that were previously measured may not be helpful in predicting current or future path properties and such path selection may lead to unintended feedback loops. Entities may select their paths to fulfill a specific goal, e.g., related to security or performance. As an example of security-related path selection, an entity may allow or disallow sending traffic over paths involving specific networks or nodes to enforce traffic policies. In an enterprise network where all traffic has to go through a specific firewall, a path-aware host can implement this policy using path selection, in which case the host needs to be aware of paths involving that firewall. As an example of performance-related path selection, an entity may prefer paths with performance properties that best match its traffic requirements. For example, for sending a small delay sensitive query, a host may select a path with a short One-Way Delay, while for retrieving a large file, it may select a path with high Link Capacities on all links. Note, there may be trade-offs between path properties (e.g., One-Way Delay and Link Capacity), and entities may influence these trade-offs with their choices. As a baseline, a path selection algorithm should aim to not perform worse than the default case most of the time. Path selection can be done both by hosts and by entities within the network: A network (e.g., an AS) can adjust its path selection for internal or external routing based on path properties. In BGP, the Multi Exit Discriminator (MED) attribute is used in the decision-making process to select which path to choose among those having the same AS PATH length and origin ; in a path-aware network, instead of using this single MED value, other properties such as Link Capacity or Link Usage could additionally be used to improve load balancing or performance .

When sending traffic over a specific path, an entity may select an appropriate protocol or configure protocol parameters depending on path properties. A host may cache state on whether a path allows the use of QUIC and if so, first attempt to connect using QUIC before falling back to another protocol when connecting over this path again. A video streaming application may choose an (initial) video quality based on the achievable data rate or the monetary cost of sending data (e.g., volume-base or flat-rate cost model).

Conversely to path or protocol selection, in addition to selecting a protocol to use over a specific adjacent path element, an entity may choose to invoke additional functions influencing the nodes to be involved in the path. For example, a 0-RTT Transport Converter will be involved in a path only when invoked by a host; such invocation will lead to the use of MPTCP or TCPinc capabilities while such use is not supported via the default forwarding path. Another example is a connection which is composed of multiple streams where each stream has specific service requirements. A host may decide to invoke a given service function (e.g., transcoding) only for some streams while others are not processed by that service function.

This Section gives some examples of path properties which may be useful, e.g., for the use cases described in . Within the context of any particular technology, available path properties may differ as entities have insight into and are able to influence different path elements and path properties. For example, a host may have some visibility into path elements that are on a low level of abstraction and close, e.g., individual nodes within the first few hops, or it may have visibility into path elements that are far away and/or on a higher level of abstraction, e.g., the list of ASes traversed. This visibility may depend on factors such as the physical or network distance or the existence of trust or contractual relationships between the host and the path element(s). Path properties may be relatively dynamic, e.g., the one-way delay of a packet sent over a specific path, or non-dynamic, e.g., the MTU of an Ethernet link which only changes infrequently. Usefulness over time differs depending on how dynamic a property is: The merit of a momentary measurement of a dynamic path property diminishes greatly as time goes on, e.g. the merit of an RTT measurement from a few seconds ago is quite small, while a non-dynamic path property might stay relevant for a longer period of time, e.g. a NAT typically stays on a specific path during the lifetime of a connection involving packets sent over this path. The physical or link layer technology used for transmitting or receiving a flow on one or multiple path elements. If known, the Access Technology may be given as an abstract link type, e.g., as Wi-Fi, Wired Ethernet, or Cellular. It may also be given as a specific technology used on a link, e.g., 2G, 3G, 4G, or 5G cellular, or 802.11a, b, g, n, or ac Wi-Fi. Other path elements relevant to the access technology may include nodes related to processing packets on the physical or link layer, such as elements of a cellular backbone network. Note that there is no common registry of possible values for this property. The price to be paid to transmit or receive a specific flow across a network to which one or multiple path elements belong. A service function that a path element applies to a flow, see . Examples of abstract service functions include firewalls, Network Address Translation (NAT), and TCP optimizers. Some stateful service functions, such as NAT, need to observe the same flow in both directions, e.g., by being an element of both the path and the reverse path. A node is transparent with respect to a protocol header, payload, or both for a specific action if the node performs this action independently of the given (meta-)information. Actions can for example be blocking packets or reading and modifying (other protocol) headers or payloads. An IP router could be transparent for transport protocol headers such as TCP and UDP, in contrast to a NAT that actively modifies TCP and UDP header information. The administrative domain, e.g., the IGP area, AS, or Service provider network to which a path element belongs. For a set of two paths or subpaths, the number of shared path elements can be a measure of intersection (e.g., Jaccard coefficient, which is the number of shared elements divided by the total number of elements). Conversely, the number of non-shared path elements can be a measure of disjointness (e.g., 1 - Jaccard coefficient). A multipath protocol might use disjointness as a metric to reduce the number of single points of failure. Two paths are symmetric if the path and its reverse path consist of the same path elements on the same level of abstraction, but in reverse order. For example, a path which consists of layer 3 switches and links between them and a reverse path with the same path elements but in reverse order are considered “routing” symmetric, as the same path elements on the same level of abstraction (IP forwarding) are traversed in the opposite direction. The maximum size, in octets, of an IP packet that can be transmitted without fragmentation. Whether a specific transport protocol can be used to establish a connection over a path or subpath, e.g., whether the path is QUIC-capable or MPTCP-capable, based on cached knowledge. Whether a specific protocol feature is available over a path or subpath, e.g., Explicit Congestion Notification (ECN), or TCP Fast Open. Some path properties express the performance of the transmission of a packet or flow over a link or subpath. Such transmission performance properties can be measured or approximated, e.g., by hosts or by path elements on the path, or they may be available as cost metrics, see . Transmission performance properties may be made available in an aggregated form, such as averages or minimums. Properties related to a path element which constitutes a single layer 2 domain are abstracted from the used physical and link layer technology, similar to . The link capacity is the maximum data rate at which data that was sent over a link can correctly be received at the node adjacent to the link. This property is analogous to the link capacity defined in but not restricted to IP-layer traffic. The link usage is the actual data rate at which data that was sent over a link is correctly received at the node adjacent to the link. This property is analogous to the link usage defined in but not restricted to IP-layer traffic. The one-way delay is the delay between a node sending a packet and another node on the same path receiving the packet. This property is analogous to the one-way delay defined in but not restricted to IP-layer traffic. The variation of the one-way delays within a flow. This property is similar to the one-way delay variation defined in but not restricted to IP-layer traffic and defined for packets on the same flow instead of packets sent between a source and destination IP address. Packets sent by a node but not received by another node on the same path after a certain time interval are considered lost. This property is analogous to the one-way loss defined in but not restricted to IP-layer traffic. Metrics such as loss patterns and loss episodes can be expressed as aggregated properties.

If nodes are basing policy or path selection decisions on path properties, they need to rely on the accuracy of path properties that other devices communicate to them. In order to be able to trust such path properties, nodes may need to establish a trust relationship or be able to verify the authenticity, integrity, and correctness of path properties received from another node. Security related properties such as confidentiality and integrity protection of payloads are difficult to characterize since they are only meaningful with respect to a threat model which depends on the use case, application, environment, and other factors. Likewise, properties for trust relations between nodes cannot be meaningfully defined without a concrete threat model, and defining a threat model is out of scope for this draft. Properties related to confidentiality, integrity, and trust are orthogonal to the path terminology and path properties defined in this document. Such properties are tied to the communicating entities and protocols used (e.g., client and server using HTTPS, or client and remote network node using VPN) while the path is typically oblivious to them. Intuitively, the path describes what function the network applies to packets, while confidentiality, integrity, and trust describe what function the communicating parties apply to packets.

This document has no IANA actions.

In contrast to the present Internet architecture, a path-aware internetworking architecture has two important properties: it exposes the properties of available Internet paths to endpoints, and provides for endpoints and applications to use these properties to select paths through the Internet for their traffic. While this property of "path awareness" already exists in many Internet-connected networks within single domains and via administrative interfaces to the network layer, a fully path-aware internetwork expands these concepts across layers and across the Internet. This document poses questions in path-aware networking open as of 2021, that must be answered in the design, development, and deployment of path-aware internetworks. It was originally written to frame discussions in the Path Aware Networking proposed Research Group (PANRG), and has been published to snapshot current thinking in this space. This document specifies an application proxy, called Transport Converter, to assist the deployment of TCP extensions such as Multipath TCP. A Transport Converter may provide conversion service for one or more TCP extensions. The conversion service is provided by means of the TCP Convert Protocol (Convert). This protocol provides 0-RTT (Zero Round-Trip Time) conversion service since no extra delay is induced by the protocol compared to connections that are not proxied. Also, the Convert Protocol does not require any encapsulation (no tunnels, whatsoever). This specification assumes an explicit model, where the Transport Converter is explicitly configured on hosts. As a sample applicability use case, this document specifies how the Convert Protocol applies for Multipath TCP. This document defines the core of the QUIC transport protocol. QUIC provides applications with flow-controlled streams for structured communication, low-latency connection establishment, and network path migration. QUIC includes security measures that ensure confidentiality, integrity, and availability in a range of deployment circumstances. Accompanying documents describe the integration of TLS for key negotiation, loss detection, and an exemplary congestion control algorithm. DO NOT DEPLOY THIS VERSION OF QUIC DO NOT DEPLOY THIS VERSION OF QUIC UNTIL IT IS IN AN RFC. This version is still a work in progress. For trial deployments, please use earlier versions. Note to Readers Discussion of this draft takes place on the QUIC working group mailing list (quic@ietf.org (mailto:quic@ietf.org)), which is archived at https://mailarchive.ietf.org/arch/search/?email_list=quic Working Group information can be found at https://github.com/quicwg; source code and issues list for this draft can be found at https://github.com/quicwg/base-drafts/labels/-transport. The current BGP specification doesn't use network performance metrics (e.g., network latency) in the route selection decision process. This document describes a performance-based BGP routing mechanism in which network latency metric is taken as one of the route selection criteria. This routing mechanism is useful for those server providers with global reach to deliver low-latency network connectivity services to their customers. This document is an extension to the base Application-Layer Traffic Optimization (ALTO) protocol. It extends the ALTO Cost Map service and ALTO Property Map service so that the application can decide which endpoint(s) to connect based on not only numerical/ordinal cost values but also details of the paths. This is useful for applications whose performance is impacted by specified components of a network on the end-to-end paths, e.g., they may infer that several paths share common links and prevent traffic bottlenecks by avoiding such paths. This extension introduces a new abstraction called Abstract Network Element (ANE) to represent these components and encodes a network path as a vector of ANEs. Thus, it provides a more complete but still abstract graph representation of the underlying network(s) for informed traffic optimization among endpoints. Cost metric is a basic concept in Application-Layer Traffic Optimization (ALTO), and different applications may use different cost metrics. Since the ALTO base protocol (RFC 7285) defines only a single cost metric (i.e., the generic "routingcost" metric), if an application wants to issue a cost map or an endpoint cost request to determine the resource provider that offers better delay performance, the base protocol does not define the cost metric to be used. This document addresses the issue by introducing network performance metrics, including network delay, jitter, packet loss rate, hop count, and bandwidth. There are multiple sources (e.g., estimation based on measurements or service-level agreement) to derive 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. This document discusses the Border Gateway Protocol (BGP), which is an inter-Autonomous System routing protocol.The primary function of a BGP speaking system is to exchange network reachability information with other BGP systems. This network reachability information includes information on the list of Autonomous Systems (ASes) that reachability information traverses. This information is sufficient for constructing a graph of AS connectivity for this reachability from which routing loops may be pruned, and, at the AS level, some policy decisions may be enforced.BGP-4 provides a set of mechanisms for supporting Classless Inter-Domain Routing (CIDR). These mechanisms include support for advertising a set of destinations as an IP prefix, and eliminating the concept of network "class" within BGP. BGP-4 also introduces mechanisms that allow aggregation of routes, including aggregation of AS paths.This document obsoletes RFC 1771. [STANDARDS-TRACK] Measuring capacity is a task that sounds simple, but in reality can be quite complex. In addition, the lack of a unified nomenclature on this subject makes it increasingly difficult to properly build, test, and use techniques and tools built around these constructs. This document provides definitions for the terms 'Capacity' and 'Available Capacity' related to IP traffic traveling between a source and destination in an IP network. By doing so, we hope to provide a common framework for the discussion and analysis of a diverse set of current and future estimation techniques. This memo provides information for the Internet community. Distributed applications -- such as file sharing, real-time communication, and live and on-demand media streaming -- prevalent on the Internet use a significant amount of network resources. Such applications often transfer large amounts of data through connections established between nodes distributed across the Internet with little knowledge of the underlying network topology. Some applications are so designed that they choose a random subset of peers from a larger set with which to exchange data. Absent any topology information guiding such choices, or acting on suboptimal or local information obtained from measurements and statistics, these applications often make less than desirable choices.This document discusses issues related to an information-sharing service that enables applications to perform better-than-random peer selection. This memo provides information for the Internet community. The IETF has developed a one-way packet loss metric that measures the loss rate on a Poisson and Periodic probe streams between two hosts. However, the impact of packet loss on applications is, in general, sensitive not just to the average loss rate but also to the way in which packet losses are distributed in loss episodes (i.e., maximal sets of consecutively lost probe packets). This document defines one-way packet loss episode metrics, specifically, the frequency and average duration of loss episodes and a probing methodology under which the loss episode metrics are to be measured. [STANDARDS-TRACK] This document describes an architecture for the specification, creation, and ongoing maintenance of Service Function Chains (SFCs) in a network. It includes architectural concepts, principles, and components used in the construction of composite services through deployment of SFCs, with a focus on those to be standardized in the IETF. This document does not propose solutions, protocols, or extensions to existing protocols. This memo defines a metric for one-way delay of packets across Internet paths. It builds on notions introduced and discussed in the IP Performance Metrics (IPPM) Framework document, RFC 2330; the reader is assumed to be familiar with that document. This memo makes RFC 2679 obsolete. This memo defines a metric for one-way loss of packets across Internet paths. It builds on notions introduced and discussed in the IP Performance Metrics (IPPM) Framework document, RFC 2330; the reader is assumed to be familiar with that document. This memo makes RFC 2680 obsolete. When routing devices rely on modems to effect communications over wireless links, they need timely and accurate knowledge of the characteristics of the link (speed, state, etc.) in order to make routing decisions. In mobile or other environments where these characteristics change frequently, manual configurations or the inference of state through routing or transport protocols does not allow the router to make the best decisions. This document introduces a new protocol called the Dynamic Link Exchange Protocol (DLEP), which provides a bidirectional, event-driven communication channel between the router and the modem to facilitate communication of changing link characteristics. This RFC is an official specification for the Internet community. It incorporates by reference, amends, corrects, and supplements the primary protocol standards documents relating to hosts. [STANDARDS-TRACK] The purpose of SDRP is to support source-initiated selection of routes to complement the route selection provided by existing routing protocols for both inter-domain and intra-domain routes. This memo provides information for the Internet community. This memo does not specify an Internet standard of any kind.

Thanks to the Path-Aware Networking Research Group for the discussion and feedback. Specifically, thanks to Mohamed Boudacair for the detailed review and various text suggestions, thanks to Brian Trammell for suggesting the flow definition, and thanks to Adrian Perrig and Matthias Rost for the detailed feedback. Thanks to Paul Hoffman for the editorial changes.