<?xmlversion="1.0" encoding="US-ASCII"?> <!-- This is built from a template for a generic Internet Draft. Suggestions for improvement welcome - write to Brian Carpenter, brian.e.carpenter @ gmail.com This can be converted using the Web service at http://xml.resource.org/ -->version='1.0' encoding='utf-8'?> <!DOCTYPE rfc SYSTEM"rfc2629.dtd"> <?rfc toc="yes"?> <!-- You want a table of contents --> <?rfc symrefs="yes"?> <!-- Use symbolic labels for references --> <?rfc sortrefs="yes"?> <!-- This sorts the references --> <?rfc iprnotified="no" ?> <!-- Change to "yes" if someone has disclosed IPR for the draft --> <?rfc compact="yes"?> <!-- This defines the specific filename and version number of your draft (and inserts the appropriate IETF boilerplate -->"rfc2629-xhtml.ent"> <rfc xmlns:xi="http://www.w3.org/2001/XInclude" category="std" number="8990" docName="draft-ietf-anima-grasp-15"ipr="trust200902">ipr="trust200902" consensus="true" obsoletes="" updates="" submissionType="IETF" xml:lang="en" tocInclude="true" tocDepth="3" symRefs="true" sortRefs="true" version="3"> <!-- xml2rfc v2v3 conversion 2.46.0 --> <front> <titleabbrev="GRASP">A Genericabbrev="GRASP">GeneRic Autonomic Signaling Protocol (GRASP)</title> <seriesInfo name="RFC" value="8990"/> <author initials="C." surname="Bormann" fullname="Carsten Bormann"><organization>Universität<organization>Universität Bremen TZI</organization> <address> <postal> <street>Postfach 330440</street><city>D-28359 Bremen</city><city>Bremen</city> <code>D-28359</code> <country>Germany</country> </postal> <email>cabo@tzi.org</email> </address> </author> <author fullname="Brian Carpenter"initials="B. E."initials="B." surname="Carpenter" role="editor"> <organization abbrev="Univ. of Auckland"/> <address> <postal><street>Department<street>School of Computer Science</street> <street>University of Auckland</street> <street>PB 92019</street> <city>Auckland</city><region/><code>1142</code> <country>New Zealand</country> </postal> <email>brian.e.carpenter@gmail.com</email> </address> </author> <author fullname="Bing Liu" initials="B." surname="Liu" role="editor"> <organization>Huawei Technologies Co., Ltd</organization> <address> <postal><street>Q14, Huawei Campus</street><street>No.156 Beiqing Road</street><city>Hai-Dian District, Beijing</city><extaddr>Q14, Huawei Campus</extaddr> <extaddr>Hai-Dian District</extaddr> <city>Beijing</city> <code>100095</code><country>P.R. China</country><country>China</country> </postal> <email>leo.liubing@huawei.com</email> </address> </author><!----><dateday="7" month="July" year="2017"/>month="May" year="2021"/> <area>Operations and Management</area> <workgroup>ANIMA</workgroup> <keyword>autonomic networking</keyword> <keyword>autonomous operation</keyword> <keyword>self-management</keyword> <abstract> <t>This document specifies the GeneRic Autonomic Signaling Protocol (GRASP), which enables autonomic nodes andautonomic service agentsAutonomic Service Agents to dynamically discover peers, to synchronize state with each other, and to negotiate parameter settings with each other. GRASP depends on an external security environment that is described elsewhere. The technical objectives and parameters for specific application scenarios are to be described in separate documents. Appendices briefly discuss requirements for the protocol and existing protocols with comparable features.</t> </abstract> </front> <middle> <section anchor="intro"title="Introduction">numbered="true" toc="default"> <name>Introduction</name> <t>The success of the Internet has made IP-based networks bigger and more complicated. Large-scale ISP and enterprise networks have become more and more problematic forhuman basedhuman-based management. Also, operational costs are growing quickly. Consequently, there are increased requirements for autonomic behavior in the networks. General aspects ofautonomic networksAutonomic Networks are discussed in <xreftarget="RFC7575"/>target="RFC7575" format="default"/> and <xreftarget="RFC7576"/>.target="RFC7576" format="default"/>. </t> <t>One approach is to largely decentralize the logic of network management by migrating it into network elements. A reference model forautonomic networkingAutonomic Networking on this basis is given in <xreftarget="I-D.ietf-anima-reference-model"/>.target="RFC8993" format="default"/>. The reader should consult this document to understand how various autonomic components fit together. In order tofulfillachieve autonomy, devices that embody Autonomic Service Agents (ASAs, <xreftarget="RFC7575"/>)target="RFC7575" format="default"/>) have specific signaling requirements. Inparticularparticular, they need to discover each other, to synchronize state with each other, and to negotiate parameters and resources directly with each other. There is no limitation on the types of parameters and resources concerned, which can include very basic information needed for addressing and routing, as well as anything else that might be configured in a conventional non-autonomic network. The atomic unit of discovery,synchronizationsynchronization, or negotiation is referred to as a technical objective,i.e,i.e., a configurable parameter or set of parameters (defined more precisely in <xreftarget="terms"/>).</t>target="terms" format="default"/>).</t> <t> Negotiation is an iterative process, requiring multiple message exchanges forming a closed loop between the negotiating entities. In fact, these entities are ASAs, normally but not necessarily in different network devices. State synchronization, when needed, can be regarded as a special case ofnegotiation,negotiation without iteration. Both negotiation and synchronization must logically follow discovery. More details of the requirements are found in <xreftarget="reqts"/>.target="reqts" format="default"/>. <xreftarget="highlevel"/>target="highlevel" format="default"/> describes a behavior model for a protocol intended to support discovery,synchronizationsynchronization, and negotiation. The design of GeneRic Autonomic Signaling Protocol (GRASP) in <xreftarget="Overview"/> of this documenttarget="Overview" format="default"/> is based on this behavior model. The relevant capabilities of various existing protocols are reviewed in <xreftarget="current"/>.</t>target="current" format="default"/>.</t> <t>The proposed discovery mechanism is oriented towards synchronization and negotiation objectives. It is based on a neighbor discovery process on the local link, but it also supports diversion to peers on other links. There is no assumption of any particular form of network topology. When a device starts up with nopre-configuration,preconfiguration, it has no knowledge of the topology. The protocol itself is capable of being used in a small and/or flat network structure such as a small office or home network as well as in alargelarge, professionally managed network. Therefore, the discovery mechanism needs to be able to allow a device to bootstrap itself without making any prior assumptions about network structure. </t> <t>Because GRASP can be used as part of a decision process among distributed devices or between networks, it must run in a secure and strongly authenticated environment. </t> <t>In realistic deployments, not all devices will support GRASP. Therefore, someautonomic service agentsAutonomic Service Agents will directly manage a group of non-autonomic nodes, and other non-autonomic nodes will be managed traditionally. Such mixed scenarios are not discussed in this specification.</t> </section><!-- intro --><section anchor="Overview"title="GRASP Protocol Overview">numbered="true" toc="default"> <name>Protocol Overview</name> <section anchor="terms"title="Terminology"> <t>Thenumbered="true" toc="default"> <name>Terminology</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 in BCP 14 <xref target="RFC2119"/>when<xref target="RFC8174"/> when, and only when, they appear inALL CAPS. When these words are not in ALL CAPS (such as "should" or "Should"), they have their usual English meanings, and are not to be interpretedall capitals, as<xref target="RFC2119"/> key words.</t>shown here. </t> <t>This document uses terminology defined in <xreftarget="RFC7575"/>.</t>target="RFC7575" format="default"/>.</t> <t>The following additional terms are used throughout this document:<list style="symbols"> <!-- <t>Autonomic Device: identical to Autonomic Node.</t> --> <t>Discovery: a</t> <dl newline="true"> <dt>Discovery: </dt> <dd><t>A process by which an ASA discovers peers according to a specific discovery objective. The discovery results may be different according to the different discovery objectives. The discovered peers may later be used as negotiation counterparts or as sources of synchronization data. </t><t>Negotiation: a</dd> <dt>Negotiation: </dt> <dd> <t>A process by which two ASAs interact iteratively to agree on parameter settings that best satisfy the objectives of bothASAs.</t> <t>StateASAs. </t> </dd> <dt>State Synchronization:a</dt> <dd><t>A process by which ASAs interact to receive the current state of parameter values stored in other ASAs. This is a special case of negotiation in which information issentsent, but the ASAs do not request their peers to change parameter settings. All other definitions apply to both negotiation and synchronization.</t> <t>Technical</t></dd> <dt>Technical Objective (usually abbreviated as Objective):A</dt> <dd><t>A technical objective is a datastructure,structure whose main contents are a name and a value. The value consists of a single configurable parameter or a set of parameters of some kind. The exact format of an objective is defined in <xreftarget="ObjForm"/>.target="ObjForm" format="default"/>. An objective occurs in three contexts:Discovery, Negotiationdiscovery, negotiation, andSynchronization.synchronization. Normally, a given objective will not occur in negotiation and synchronization contexts simultaneously.<list style="symbols"> <t>One</t> <ul empty="true"> <li>One ASA may support multiple independentobjectives.</t> <t>Theobjectives. </li> <li> The parameter(s) in the value of a given objective apply to a specific service or function or action. They may in principle be anything that can be set to a specific logical,numericalnumerical, or string value, or a more complex data structure, by a network node. Each node is expected to contain one or more ASAs which may each manage subsidiary non-autonomicnodes.</t> <t>Discoverynodes. </li> <li> <dl> <dt>Discovery Objective:an</dt> <dd>an objective in the process of discovery. Its value may beundefined.</t> <t>Synchronizationundefined. </dd> <dt>Synchronization Objective:an</dt> <dd>an objective whose specific technical content needs to be synchronized among two or more ASAs. Thus, each ASA will maintain its own copy of theobjective.</t> <t>Negotiationobjective. </dd> <dt>Negotiation Objective:an</dt> <dd>an objective whose specific technical content needs to be decided in coordination with another ASA. Again, each ASA will maintain its own copy of theobjective.</t> </list>objective. </dd> </dl> </li> <li> A detailed discussion of objectives, including their format, is found in <xreftarget="ObjOption"/>.</t> <t>Discoverytarget="ObjOption" format="default"/>. </li> </ul> </dd> <dt>Discovery Initiator:an</dt> <dd><t>An ASA that starts discovery by sending adiscoveryDiscovery message referring to a specific discoveryobjective.</t> <t>Discoveryobjective. </t></dd> <dt>Discovery Responder:a</dt> <dd><t>A peer that either contains an ASA supporting the discovery objective indicated by the discoveryinitiator,initiator or caches the locator(s) of the ASA(s) supporting the objective. It sends a Discovery Response, as describedlater.</t> <t>Synchronizationlater. </t></dd> <dt>Synchronization Initiator:an</dt> <dd><t>An ASA that starts synchronization by sending a request message referring to a specific synchronizationobjective.</t> <t>Synchronizationobjective. </t></dd> <dt>Synchronization Responder:a</dt> <dd><t>A peer ASAwhichthat responds with the value of a synchronizationobjective.</t> <t>Negotiationobjective. </t></dd> <dt>Negotiation Initiator:an</dt> <dd><t>An ASA that starts negotiation by sending a request message referring to a specific negotiation objective.</t><t>Negotiation</dd> <dt>Negotiation Counterpart:a</dt> <dd> <t>A peer with which theNegotiation Initiatornegotiation initiator negotiates a specific negotiation objective.</t><t>GRASP</dd> <dt>GRASP Instance:This</dt> <dd> <t>This refers to an instantiation of a GRASP protocol engine, likely including multiple threads or processes as well as dynamic data structures such as a discovery cache, running in a given security environment on a single device. </t><t>GRASP</dd> <dt>GRASP Core:This</dt> <dd> <t>This refers to the code and shared data structures of a GRASP instance, which will communicate with individual ASAs via a suitable Application Programming Interface(API).</t> <t>Interface(API). </t> </dd> <dt>Interface or GRASP Interface:Unless</dt> <dd> <t>Unless otherwise stated,these referthis refers to a networkinterface -interface, which might be physical orvirtual -virtual, that a specific instance of GRASP is currently using. A device might have other interfaces that are not used by GRASP and which are outside the scope of theautonomic network.</t> </list></t>Autonomic Network. </t> </dd> </dl> </section> <section anchor="hilev"title="High Levelnumbered="true" toc="default"> <name>High-Level DeploymentModel">Model</name> <t>A GRASP implementation will be part of the Autonomic Networking Infrastructure (ANI) in an autonomic node, which must also provide an appropriate security environment. In accordance with <xreftarget="I-D.ietf-anima-reference-model"/>,target="RFC8993" format="default"/>, thisSHOULD<bcp14>SHOULD</bcp14> be the Autonomic Control Plane (ACP) <xreftarget="I-D.ietf-anima-autonomic-control-plane"/>.target="RFC8994" format="default"/>. As a result, all autonomic nodes in the ACP are able to trust each other. It is expected that GRASP will access the ACP by using a typical socket programminginterfaceinterface, and the ACP will make available only network interfaces within theautonomic network.Autonomic Network. If there is no ACP, the considerations described in <xreftarget="reqsec"/>target="reqsec" format="default"/> apply. </t> <t> There will also be one or more Autonomic Service Agents (ASAs). In the minimal case of a single-purpose device, these components might be fully integrated with GRASP and the ACP. A more common model is expected to be amulti-purposemultipurpose device capable of containing several ASAs, such as a router or large switch. In this case it is expected that the ACP, GRASP and the ASAs will be implemented as separate processes, which are able to support asynchronous and simultaneous operations, for example bymulti-threading.</t>multithreading.</t> <t>In some scenarios, a limited negotiation model might be deployed based on a limited trust relationship such as that between two administrative domains. ASAs might then exchange limited information and negotiate some particular configurations.</t> <t>GRASP is explicitly designed to operate within a single addressing realm. Its discovery and flooding mechanisms do not support autonomic operations that cross any form of address translator orupper layerupper-layer proxy.</t> <t>A suitable Application Programming Interface (API) will be needed between GRASP and the ASAs. In some implementations, ASAs would run in user space with a GRASP library providing the API, and this library would in turn communicate via system calls with core GRASP functions. Details of the API are out of scope for the present document. For further details of possible deployment models, see <xreftarget="I-D.ietf-anima-reference-model"/>.target="RFC8993" format="default"/>. </t> <t>An instance of GRASP must be aware of the network interfaces it will use, and of the appropriate global-scope and link-local addresses. In the presence of the ACP, such information will be available from the adjacency table discussed in <xreftarget="I-D.ietf-anima-reference-model"/>.target="RFC8993" format="default"/>. In other cases, GRASP must determine such information for itself. Details depend on the device and operating system. In the rest of this document, the terms 'interfaces' or 'GRASP interfaces' refers only to the set of network interfaces that a specific instance of GRASP is currently using. </t> <t>Because GRASP needs to work with very high reliability, especially during bootstrapping and during fault conditions, it is essential that every implementation continues to operate in adverse conditions. For example, discovery failures, or any kind of socket exception at any time, must not cause irrecoverable failures in GRASP itself, and must return suitable error codes through the API so that ASAs can also recover. </t> <t>GRASP must not depend uponnon-volatilenonvolatile data storage. Allrun timeruntime error conditions, and events such as address renumbering, network interface failures, and CPU sleep/wake cycles, must be handled in such a way that GRASP will still operate correctly and securely afterwards (<xreftarget="reqsec"/>) afterwards.</t>target="reqsec" format="default"/>).</t> <t>An autonomic node will normally run a single instance of GRASP, which is used by multiple ASAs. Possible exceptions are mentioned below. </t> </section> <section anchor="highlevel"title="High Level Design">numbered="true" toc="default"> <name>High-Level Design</name> <t>This section describes the behavior model and general design of GRASP, supporting discovery,synchronizationsynchronization, and negotiation, to act as a platform for different technical objectives.</t><t><list style="symbols"> <t>A<dl newline="true"> <dt>A genericplatform:<vspace blankLines="1"/> Theplatform: </dt> <dd><t>The protocol design is generic and independent of the synchronization or negotiation contents. The technical contents will vary according to the various technical objectives and the different pairs ofcounterparts.<vspace blankLines="1"/></t> <t>Normally,counterparts.</t></dd> <dt>Multiple instances: </dt> <dd><t> Normally, a single main instance of the GRASP protocol engine will exist in an autonomic node, and each ASA will run as an independent asynchronous process. However, scenarios where multiple instances of GRASP run in a single node, perhaps with different security properties, are possible (<xreftarget="secinst"/>).target="secinst" format="default"/>). In this case, each instanceMUST<bcp14>MUST</bcp14> listen independently for GRASP link-local multicasts, and all instancesMUST<bcp14>MUST</bcp14> be woken by each suchmulticast,multicast in order for discovery and flooding to work correctly.<vspace blankLines="1"/></t> <t>Security infrastructure:<vspace blankLines="1"/> As</t></dd> <dt>Security infrastructure: </dt> <dd><t>As noted above, the protocol itself has no built-in securityfunctionality,functionality and relies on a separate secureinfrastructure.<vspace blankLines="1"/></t> <t>Discovery, synchronizationinfrastructure.</t> </dd> <dt>Discovery, synchronization, and negotiation are designedtogether:<vspace blankLines="1"/> Thetogether: </dt> <dd><t>The discovery method and the synchronization and negotiation methods are designed in the same way and can be combined when this is useful, allowing a rapid mode of operation described in <xreftarget="discmech"/>.target="discmech" format="default"/>. These processes can also be performed independently whenappropriate. <list style="symbols"> <t>Thus,appropriate.</t> <ul empty="true"> <li> <t> Thus, for some objectives, especially those concerned withapplication layerapplication-layer services, another discovery mechanism such asthe future DNSDNS-based Service Discovery <xreftarget="RFC7558"/> MAYtarget="RFC7558" format="default"/> <bcp14>MAY</bcp14> be used. The choice is left to the designers of individualASAs.</t> </list> <vspace blankLines="1"/></t> <t>AASAs. </t> </li> </ul> </dd> <dt>A uniform pattern for technicalobjectives:<vspace blankLines="1"/>objectives: </dt> <dd> <t> The synchronization and negotiation objectives are defined according to a uniform pattern. The values that they contain could be carried either in a simple binary format or in a complex object format. The basic protocol design uses the Concise Binary Object Representation (CBOR) <xreftarget="RFC7049"/>,target="RFC8949" format="default"/>, which is readily extensible forunknownunknown, future requirements.<vspace blankLines="1"/></t> <t>A</t> </dd> <dt>A flexible model forsynchronization:<vspace blankLines="1"/> GRASPsynchronization: </dt> <dd> <t>GRASP supports synchronization between two nodes, which could be used repeatedly to perform synchronization among a small number of nodes. It also supports an unsolicited flooding mode when large groups of nodes, possibly including all autonomic nodes, need data for the same technical objective.<list style="symbols"> <t>There</t> <ul empty="true"> <li> <t> There may be some network parameters for which a more traditional flooding mechanism such asDNCPthe Distributed Node Consensus Protocol (DNCP) <xreftarget="RFC7787"/>target="RFC7787" format="default"/> is considered more appropriate. GRASP can coexist with DNCP. </t></list> <vspace blankLines="1"/></t> <t>A</li> </ul> </dd> <dt>A simple initiator/responder model fornegotiation:<vspace blankLines="1"/> Multi-partynegotiation: </dt> <dd> <t>Multiparty negotiations are very complicated to model and cannot readily be guaranteed to converge. GRASP uses a simple bilateral model and can supportmulti-partymultiparty negotiations by indirect steps.<vspace blankLines="1"/></t> <t>Organizing</t> </dd> <dt>Organizing of synchronization or negotiationcontent:<vspace blankLines="1"/> Thecontent: </dt> <dd> <t>The technical content transmitted by GRASP will be organized according to the relevant function or service. The objectives for different functions or services are keptseparate,separate because they may be negotiated or synchronized with different counterparts or have different response times. Thus a normal arrangementwould beis a single ASA managing a small set of closely related objectives, with a version of that ASA in each relevant autonomic node. Further discussion of this aspect is out of scope for the current document.<vspace blankLines="1"/></t> <t>Requests</t> </dd> <dt>Requests and responses in negotiationprocedures:<vspace blankLines="1"/>procedures: </dt> <dd> <t> The initiator can negotiate a specific negotiation objective with relevant counterpart ASAs. It can request relevant information from a counterpart so that it can coordinate its local configuration. It can request the counterpart to make a matching configuration. It can request simulation or forecast results by sending somedry rundry-run conditions.<vspace blankLines="1"/>Beyond</t> <t> Beyond the traditional yes/no answer, the responder can reply with a suggested alternative value for the objective concerned. This would start abi-directionalbidirectional negotiation ending in a compromise between the twoASAs.<vspace blankLines="1"/></t> <t>ConvergenceASAs. </t> </dd> <dt>Convergence of negotiationprocedures:<vspace blankLines="1"/> Toprocedures: </dt> <dd> <t>To enableconvergence,convergence when a responder suggests a new value or condition in a negotiation step reply, it should be as close as possible to the original request or previous suggestion. The suggested value of later negotiation steps should be chosen between the suggested values from the previous two steps. GRASP provides mechanisms to guarantee convergence (or failure) in a small number of steps, namely a timeout and a maximum number of iterations.<vspace blankLines="1"/></t><t>Extensibility:<vspace blankLines="1"/> GRASP</dd> <dt>Extensibility: </dt> <dd> <t>GRASP intentionally does not have a version number, and it can be extended by adding new message types and options. The InvalidMessagemessage (M_INVALID) will be used to signal that an implementation does not recognize a message or option sent by another implementation. In normal use, new semantics will be added by defining new synchronization or negotiation objectives. </t></list></t></dd> </dl> </section> <sectiontitle="Quicknumbered="true" toc="default"> <name>Quick OperatingOverview">Overview</name> <t>An instance of GRASP is expected to run as a separate core module, providing an API (such as <xreftarget="I-D.liu-anima-grasp-api"/>)target="RFC8991" format="default"/>) to interface to various ASAs. These ASAs may operate without special privilege, unless they need it for other reasons (such as configuring IP addresses or manipulating routing tables).</t><t></t> <t> The GRASP mechanisms used by the ASA are built around GRASP objectives defined as data structures containing administrative information such as the objective's uniquename,name and its current value. The format and size of the value is not restricted by the protocol, except that it must be possible to serialize it for transmission in CBOR, which is no restriction at all in practice.</t><t></t> <t> GRASP provides the following mechanisms:<list style="symbols"> <t>A</t> <ul spacing="normal"> <li>A discovery mechanism (M_DISCOVERY,M_RESPONSE),M_RESPONSE) by which an ASA can discover other ASAs supporting a given objective.</t><t></li> <li> A negotiation request mechanism(M_REQ_NEG),(M_REQ_NEG) by which an ASA can start negotiation of an objective with a counterpart ASA. Once a negotiation has started, the process is symmetrical, and there is a negotiation step message (M_NEGOTIATE) for each ASA to use in turn. Two other functions support negotiating steps (M_WAIT, M_END).</t><t></li> <li> A synchronization mechanism(M_REQ_SYN),(M_REQ_SYN) by which an ASA can request the current value of an objective from a counterpart ASA. With this, there is a corresponding response function (M_SYNCH) for an ASA that wishes to respond to synchronization requests.</t><t></li> <li> A flood mechanism(M_FLOOD),(M_FLOOD) by which an ASA can cause the current value of an objective to be flooded throughout theautonomic networkAutonomic Network so that any ASA can receive it. One application of this is to act as an announcement, avoiding the need for discovery of a widely applicableobjective.</t> </list></t>objective.</li> </ul> <t>Some example messages and simple message flows are provided in <xreftarget="examples"/>.</t>target="examples" format="default"/>.</t> </section> <sectiontitle="GRASP Protocolnumbered="true" toc="default"> <name>GRASP Basic Properties andMechanisms">Mechanisms</name> <section anchor="reqsec"title="Requirednumbered="true" toc="default"> <name>Required External SecurityMechanism">Mechanism</name> <t>GRASP does not specify transport security because it is meant to be adapted to different environments. Every solution adopting GRASPMUST<bcp14>MUST</bcp14> specify a security and transport substrate used by GRASP in that solution.</t> <t>The substrateMUST<bcp14>MUST</bcp14> enforce sending and receiving GRASP messages only between members of a mutually trusted group running GRASP. Each group member is an instance of GRASP. The group members are nodes of a connected graph. The group and graphisare created by the security and transport substrate and are called the GRASP domain. The substrate must support unicast messages between any group members and (link-local) multicast messages between adjacent group members. It must deny messages between group members andnon groupnon-group members. With this model, security is provided by enforcing group membership, but any member of the trusted group can attack the entire network until revoked.</t> <t> SubstratesMUST<bcp14>MUST</bcp14> use cryptographic member authentication and message integrity for GRASP messages. This can beend-to-endend to end orhop-by-hophop by hop across the domain. The security and transport substrateMUST<bcp14>MUST</bcp14> provide mechanisms to remove untrusted members from the group.</t> <t>If the substrate does not mandate and enforce GRASP messageencryptionencryption, then any service using GRASP in such a solutionMUST<bcp14>MUST</bcp14> provide protection and encryption for message elements whose exposure could constitute an attack vector.</t> <t>The security and transport substrate for GRASP in the ANI is the ACP. Unless otherwise noted, we assume this security and transport substrate in the remainder of this document. The ACP does mandate the use of encryption;thereforetherefore, GRASP in the ANI can rely on GRASPmessagemessages being encrypted. The GRASP domain is the ACP: all nodes in an autonomic domain connected by encrypted virtual links formed by the ACP. The ACP uses hop-by-hop security(authentication/encryption)(authentication and encryption) of messages. Removal of nodes relies on standard PKI certificate revocation or expiry of sufficientlyshort livedshort-lived certificates. Refer to <xreftarget="I-D.ietf-anima-autonomic-control-plane"/>target="RFC8994" format="default"/> for more details.</t> <t>As mentioned in <xreftarget="highlevel"/>,target="highlevel" format="default"/>, some GRASP operations might be performed across an administrative domain boundary by mutual agreement, without the benefit of an ACP. Such operationsMUST<bcp14>MUST</bcp14> be confined to a separate instance of GRASP with its own copy of all GRASP data structures running across a separate GRASP domain with a security and transport substrate. In the most simple case, each point-to-point interdomain GRASP peering could be a separatedomaindomain, and the security and transport substrate could be built using transport ornetwork layernetwork-layer security protocols. This is subject to future specifications. </t><!-- TLS <xref target="RFC5246"/> and DTLS <xref target="RFC6347"/> based on a Public Key Infrastructure (PKI) <xref target="RFC5280"/> are RECOMMENDED for this purpose.--><t>An exception to the requirements for the security and transport substrate exists for highly constrained subsets of GRASP meant to support the establishment of a security and transport substrate, described in the following section.</t> </section> <section anchor="secinst"title="Discoverynumbered="true" toc="default"> <name>Discovery Unsolicited Link-Local (DULL)GRASP">GRASP</name> <t>Some services may need to use insecure GRASP discovery,responseresponse, and flood messages without being able to usepre-existingpreexisting security associations, forexampleexample, as part of discovery for establishing security associations such as a security substrate for GRASP.</t> <t>Such operations being intrinsically insecure, they need to be confined to link-local use to minimize the risk of malicious actions. Possible examples include discovery of candidate ACP neighbors <xreftarget="I-D.ietf-anima-autonomic-control-plane"/>,target="RFC8994" format="default"/>, discovery of bootstrap proxies <xreftarget="I-D.ietf-anima-bootstrapping-keyinfra"/>target="RFC8995" format="default"/>, or perhaps initialization services in networks using GRASP without being fully autonomic (e.g., no ACP). Such usageMUST<bcp14>MUST</bcp14> be limited to link-local operations on a single interface andMUST<bcp14>MUST</bcp14> be confined to a separate insecure instance of GRASP with its own copy of all GRASP data structures. This instance is nicknamed DULL--- Discovery Unsolicited Link-Local.</t> <t>The detailed rules for the DULL instance of GRASP are as follows:<list style="symbols"> <t>An</t> <ul spacing="normal"> <li>An initiatorMAY<bcp14>MAY</bcp14> send Discovery or Flood Synchronization link-local multicast messageswhich MUSTthat <bcp14>MUST</bcp14> have a loop count of 1, to prevent off-link operations. Other unsolicited GRASP message typesMUST NOT<bcp14>MUST NOT</bcp14> besent.</t> <t>Asent.</li> <li>A responderMUST<bcp14>MUST</bcp14> silently discard any message whose loop count is not1.</t> <t>A1.</li> <li>A responderMUST<bcp14>MUST</bcp14> silently discard any message referring to a GRASPObjectiveobjective that is not directly part of a service that requires this insecuremode.</t> <t>Amode.</li> <li>A responderMUST NOT<bcp14>MUST NOT</bcp14> relay any multicastmessages.</t> <t>Amessages.</li> <li>A Discovery ResponseMUST<bcp14>MUST</bcp14> indicate a link-localaddress.</t> <t>Aaddress.</li> <li>A Discovery ResponseMUST NOT<bcp14>MUST NOT</bcp14> include a Divertoption.</t> <t>Aoption.</li> <li>A nodeMUST<bcp14>MUST</bcp14> silently discard any message whose source address is notlink-local.</t> </list></t>link-local.</li> </ul> <t>To minimize traffic possibly observed by third parties, GRASP trafficSHOULD<bcp14>SHOULD</bcp14> be minimized by using only Flood Synchronization to announce objectives and their associated locators, rather than by using Discovery andResponse.Discovery Response messages. Further details are out of scope for thisdocument</t>document.</t> </section><!--<sectionanchor="secinst-sonn" title="Secure Only Neighbor Negotiation"> <t>Some services might use insecure on-link operationsanchor="trans" numbered="true" toc="default"> <name>Transport Layer Usage</name> <t>All GRASP messages, after they are serialized as a CBOR byte string, are transmitted as such directly over the transport protocol inDULL, but also use unicast synchronization or negotiation operations protected by TLS. A separate instance of GRASP is used, with its own copy of all GRASP data structures. This instance is nicknamed SONN - Secure Only Neighbor Negotiation.</t> <t>use. Thedetailed rulestransport protocol(s) forthe SONN instance ofa GRASP domain are specified by the security and transport substrate asfollows: <list style="symbols"> <t>All types of GRASP messageintroduced in <xref target="reqsec" format="default"/>.</t> <t>GRASP discovery and flooding messages arepermitted.</t> <t>An initiator MUST send any Discovery or Flood Synchronization link-local multicast messages with a loop count of 1.</t> <t>A responder MUST silently discard any Discovery or Flood Synchronization message whose loop count is not 1.</t> <t>A responder MUST silently discard any message referring to a GRASP Objective that is not directly part of the service concerned.</t> <t>A responder MUST NOT relay any multicast messages.</t> <t>A Discovery Response MUST indicate a link-local address.</t> <t>A Discovery Response MUST NOT include a Divert option.</t> <t>A node MUST silently discard any message whose source address is not link-local.</t> </list></t> <t>Further details are out of scope for this document.</t> </section> --> <section anchor="trans" title="Transport Layer Usage"> <t>All GRASP messages, after they are serialized as a CBOR byte string, are transmitted as such directly over the transport protocol in use. The transport protocol(s) for a GRASP domain are specified by the security and transport substrate as introduced in <xref target="reqsec"/>.</t> <t>GRASP discovery and flooding messages are designed for GRASP domain wide flooding through hop-by-hopdesigned for GRASP domain-wide flooding through hop-by-hop link-local multicast forwarding between adjacent GRASP nodes. The GRASP security and transport substrate needs to specify how theselink locallink-local multicasts are transported. This can be unreliable transport (UDP) but itSHOULD<bcp14>SHOULD</bcp14> be reliable transport (e.g., TCP).</t> <t>If the substrate specifies an unreliable transport such as UDP for discovery and flooding messages, then itMUST NOT<bcp14>MUST NOT</bcp14> use IP fragmentation because of its loss characteristic, especially in multi-hop flooding. GRASPMUST<bcp14>MUST</bcp14> then enforce at the user API level a limit to the size of discovery and flooding messages, so that no fragmentation can occur. For IPv6transporttransport, this means that the size of thosemessagesmessages' IPv6 packets must be at most 1280 bytessized IPv6 packets(unless there is a known larger minimum link MTU across the whole GRASP domain).</t> <t>All other GRASP messages are unicastbeteweenbetween group members of the GRASP domain. TheseMUST<bcp14>MUST</bcp14> use a reliable transport protocol because GRASP itself does not provide for error detection,retransmissionretransmission, or flow control. Unless otherwise specified by the security and transport substrate, TCPMUST<bcp14>MUST</bcp14> be used.</t> <t>The security and transport substrate for GRASP in the ANI is the ACP. Unless otherwise noted, we assume this security and transport substrate in the remainder of this document when describingGRASPsGRASP's message transport. In the ACP, TCP is used for GRASP unicast messages. GRASP discovery and flooding messages also use TCP:Thesethese link-local messages are forwarded by replicating them to all adjacent GRASP nodes on the link via TCP connections to those adjacent GRASP nodes. Because of this, GRASP in the ANI has no limitations on the size of discovery and flooding messages with respect to fragmentation issues.UDP is used in the ANI with GRASP only with DULL whenWhile the ACP is being built using a DULL instance of GRASP, native UDP multicast is used to discover ACP/GRASP neighbors onlinks.</t> <!-- <t>Nevertheless, when running within a secure ACP on reliable infrastructure, UDP MAY be used for unicast messages not exceeding the minimum IPv6 path MTU; however, TCP MUST be used for longer messages. In other words, IPv6 fragmentation is avoided. If a node receives a UDP message but the reply is too long, it MUST open a TCP connection to the peer for the reply. Note that when the network is under heavy load or in a fault condition, UDP might become unreliable. Since this is when autonomic functions are most necessary, automatic fallback to TCP MUST be implemented. The simplest implementation is therefore to use only TCP.</t> -->links. </t> <t>For link-local UDP multicast,theGRASPprotocollistens to the well-known GRASP Listen Port (<xreftarget="Constants"/>).target="Constants" format="default"/>). Transport connections forDiscoverydiscovery andFloodingflooding on relay nodes must terminate in GRASP instances(eg:(e.g., GRASP ASAs) so that link-local multicast, hop-by-hop flooding of M_DISCOVERY and M_FLOOD messages and hop-by-hop forwarding of M_RESPONSE responses and caching of those responses along the path work correctly.</t> <t>Unicast transport connections used for synchronization and negotiation can terminate directly in ASAs that implementobjectives and thereforeobjectives; therefore, this traffic does not need to pass through GRASP instances. For this, the ASA listens on its own dynamically assigned ports, which are communicated to its peers during discovery. Alternatively, the GRASP instance can also terminate the unicast transport connections and pass the traffic from/to the ASA if that ispreferrablepreferable in someimplementation (eg: toimplementations (e.g., to better decouple ASAs from network connections).</t> </section> <section anchor="discmech"title="Discoverynumbered="true" toc="default"> <name>Discovery Mechanism andProcedures">Procedures</name> <sectiontitle="Separated discoverynumbered="true" toc="default"> <name>Separated Discovery andnegotiation mechanisms">Negotiation Mechanisms</name> <t>Although discovery and negotiation or synchronization are defined together in GRASP, they are separate mechanisms. The discovery process could run independently from the negotiation or synchronization process. Upon receiving a Discovery message (<xreftarget="DiscoveryMessage"/>) message,target="DiscoveryMessage" format="default"/>), the recipient node should return aresponseDiscovery Response message in which it either indicates itself as a discovery responder or diverts the initiator towards another more suitable ASA. However, this response may be delayed if the recipient needs to relay thediscovery onwards,Discovery message onward, as describedbelow.</t>in <xref target="discovery-relaying" format="default"/>.</t> <t>The discovery action (M_DISCOVERY) will normally be followed by a negotiation (M_REQ_NEG) or synchronization (M_REQ_SYN) action. The discovery results could be utilized by the negotiation protocol to decide which ASA the initiator will negotiate with.</t> <t>The initiator of a discovery action for a given objective need not be capable of responding to that objective as aNegotiation Counterpart,negotiation counterpart, as aSynchronization Respondersynchronization responder, or as source for flooding. For example, an ASA might perform discovery even if it only wishes to act as aSynchronization Initiatorsynchronization initiator orNegotiation Initiator.negotiation initiator. Such an ASA does not itself need to respond todiscoveryDiscovery messages.</t> <t>It is also entirely possible to use GRASP discovery without any subsequent negotiation or synchronization action. In this case, the discovered objective is simply used as a name during the discoveryprocessprocess, and any subsequent operations between the peers are outside the scope of GRASP.</t> </section> <section anchor="discovw"title="Discovery Overview">numbered="true" toc="default"> <name>Discovery Overview</name> <t>A complete discovery process will start with a multicast(of M_DISCOVERY)Discovery message (M_DISCOVERY) on the local link. On-link neighbors supporting the discovery objective will respond directly(with M_RESPONSE).with Discovery Response (M_RESPONSE) messages. A neighbor with multiple interfaces may respond with a cacheddiscovery response.Discovery Response. If it has no cached response, it will relay thediscoveryDiscovery message on its other GRASPinterfaces<!--, for example reaching a higher-level gateway in a hierarchical network-->.interfaces. If a node receiving the relayeddiscoveryDiscovery message supports the discovery objective, it will respond to the relayeddiscovery.Discovery message. If it has a cached response, it will respond with that. If not, it will repeat the discovery process, which thereby becomes iterative. The loop count and timeout will ensure that the process ends. Further details are givenbelow.in <xref target="discovery-relaying" format="default"/>. </t> <t>A Discovery messageMAY<bcp14>MAY</bcp14> be sent unicast to a peer node, whichSHOULD<bcp14>SHOULD</bcp14> then proceed exactly as if the message had been multicast, except that when TCP is used, the response will be on the same socket as the query. However, this mode does not guarantee successful discovery in the general case. </t> </section> <section anchor="discproc"title="Discovery Procedures">numbered="true" toc="default"> <name>Discovery Procedures</name> <t>Discovery starts as an on-link operation. The Divert option can tell the discovery initiator to contact an off-link ASA for that discovery objective. If the security and transport substrate of the GRASP domain (see <xreftarget="trans"/>)target="trans" format="default"/>) uses UDP link-localmulticastmulticast, then the discovery initiator sends these to the ALL_GRASP_NEIGHBORS link-local multicast address (<xreftarget="Constants"/>) andtarget="Constants" format="default"/>), and all GRASP nodes need to listen to this address to act as discoveryresponder.responders. Because this port is unique in a device, this is a function of the GRASP instance and not of an individual ASA. As a result, each ASA will need to register the objectives that it supports with the local GRASP instance.</t> <t>If an ASA in a neighbor device supports the requested discovery objective, the deviceSHOULD<bcp14>SHOULD</bcp14> respond to the link-local multicast with a unicast Discovery Response message (<xreftarget="ResponseMessage"/>)target="ResponseMessage" format="default"/>) with locatoroption(s),option(s) (<xref target="LocatorOption" format="default"/>) unless it is temporarily unavailable. Otherwise, if the neighbor has cached information about an ASA that supports the requested discovery objective (usually because it discovered the same objective before), itSHOULD<bcp14>SHOULD</bcp14> respond with a Discovery Response message with a Divert option pointing to the appropriateDiscovery Responder.discovery responder. However, itSHOULD NOT<bcp14>SHOULD NOT</bcp14> respond with a cached response on an interface if itlearntlearned that information from the sameinterface,interface because the peer in question will answer directly if still operational.</t> <t>If a device has no information about the requested discoveryobjective,objective and is not acting as a discovery relay (seebelow)<xref target="discovery-relaying" format="default"/>), itMUST<bcp14>MUST</bcp14> silently discard the Discovery message.</t> <t>The discovery initiatorMUST<bcp14>MUST</bcp14> set a reasonable timeout on the discovery process. A suggested value is 100 milliseconds multiplied by the loop count embedded in the objective.</t> <t>If nodiscovery responseDiscovery Response is received within the timeout,<!-- a reasonable timeout (default GRASP_DEF_TIMEOUT milliseconds, <xref target="Constants"/>),-->the Discovery messageMAY<bcp14>MAY</bcp14> berepeated,repeated with a newly generated Session ID (<xreftarget="SessionID"/>).target="SessionID" format="default"/>). An exponential backoffSHOULD<bcp14>SHOULD</bcp14> be used for subsequentrepetitions,repetitions to limit the load during busy periods. The details of the backoff algorithm will depend on the use case for the objective concerned butMUST<bcp14>MUST</bcp14> be consistent with the recommendations in <xreftarget="RFC8085"/>target="RFC8085" format="default"/> for low data-volume multicast. Frequent repetition might be symptomatic of adenial of servicedenial-of-service attack.</t> <t>After a GRASP device successfully discovers a locator for aDiscovery Responderdiscovery responder supporting a specific objective, itSHOULD<bcp14>SHOULD</bcp14> cache this information, including the interface index <xreftarget="RFC3493"/>target="RFC3493" format="default"/> via which it was discovered. This cache recordMAY<bcp14>MAY</bcp14> be used for future negotiation or synchronization, and the locatorSHOULD<bcp14>SHOULD</bcp14> be passed on when appropriate as a Divert option to anotherDiscovery Initiator.</t>discovery initiator.</t> <t>The cache mechanismMUST<bcp14>MUST</bcp14> include a lifetime for each entry. The lifetime is derived from a time-to-live (ttl) parameter in each Discovery Response message. Cached entriesMUST<bcp14>MUST</bcp14> be ignored or deleted after their lifetime expires. In some environments, unplanned address renumbering might occur. In such cases, the lifetimeSHOULD<bcp14>SHOULD</bcp14> be short compared to the typical addresslifetime<!-- and a mechanism to flush the discovery cache MUST be implemented-->.lifetime. The discovery mechanism needs to track the node's current address to ensure that Discovery Responses always indicate the correct address.</t> <t>If multipleDiscovery Respondersdiscovery responders are found for the same objective, theySHOULD<bcp14>SHOULD</bcp14> all becached,cached unless this creates a resource shortage. The method of choosing between multiple responders is an implementation choice. This choiceMUST<bcp14>MUST</bcp14> be available to eachASAASA, but the GRASP implementationSHOULD<bcp14>SHOULD</bcp14> provide a default choice.</t> <t>BecauseDiscovery Respondersdiscovery responders will be cached in a finite cache, they might be deleted at any time. In this case, discovery will need to be repeated. If an ASA exits for any reason, its locator might still be cached for some time, and attempts to connect to it will fail. ASAs need to be robust in these circumstances. </t> </section> <sectiontitle="Discovery Relaying">anchor="discovery-relaying" numbered="true" toc="default"> <name>Discovery Relaying</name> <t>A GRASP instance with multiple link-layer interfaces (typically running in a router)MUST<bcp14>MUST</bcp14> support discovery on all GRASP interfaces. We refer to this as a 'relaying instance'.</t> <t>DULLInstancesinstances (<xreftarget="secinst"/>)target="secinst" format="default"/>) are always single-interface instances and thereforeMUST NOT<bcp14>MUST NOT</bcp14> perform discovery relaying.</t> <t>If a relaying instance receives a Discovery message on a given interface for a specific objective that it does not support and for which it has not previously cached aDiscovery Responder,discovery responder, itMUST<bcp14>MUST</bcp14> relay the query byre-issuingreissuing a new Discovery message as a link-local multicast on its other GRASP interfaces.</t> <t> The relayeddiscoveryDiscovery messageMUST<bcp14>MUST</bcp14> have the same Session ID andInitiator'initiator' field as the incoming message (see <xreftarget="DiscoveryMessage"/>).target="DiscoveryMessage" format="default"/>). TheInitiatorIP address in the 'initiator' field is only used toallow for disambiguation ofdisambiguate the Session ID and is never used to address Response packets. Response packets are sent back to the relaying instance, not the original initiator.</t> <t>The M_DISCOVERY message does not encode the transport address of the originator or relay. Response packets must therefore be sent to thetransport layertransport-layer address of the connection on which the M_DISCOVERY message was received. If the M_DISCOVERY was relayed via a reliable hop-by-hop transport connection, the response is simply sent back via the same connection.</t> <t>If the M_DISCOVERY was relayed via link-local(eg:(e.g., UDP) multicast, the response is sent back via a reliable hop-by-hop transport connection with the same port number as the source port of the link-local multicast. Therefore, if link-local multicast is used and M_RESPONSE messages are required (which is the case in almost all GRASP instances except for the limited use of DULL instances in the ANI), GRASP needs to be able to bind to one port number on UDP from which to originate the link-local multicast M_DISCOVERY messages and the same port number on the reliable hop-by-hop transport(eg:(e.g., TCP by default) to be able to respond to transport connections from responders that want to send M_RESPONSE messages back. Note that this port does not need to be the GRASP_LISTEN_PORT.</t> <t>The relaying instanceMUST<bcp14>MUST</bcp14> decrement the loop count within the objective, andMUST NOT<bcp14>MUST NOT</bcp14> relay the Discovery message if the result is zero. Also, itMUST<bcp14>MUST</bcp14> limit the total rate at which it relaysdiscoveryDiscovery messages to a reasonablevalue,value in order to mitigate possibledenial of servicedenial-of-service attacks. For example, the rate limit could be set to a small multiple of the observed rate ofdiscoveryDiscovery messages during normal operation. The relaying instanceMUST<bcp14>MUST</bcp14> cache the Session ID value and initiator address of each relayed Discovery message until any Discovery Responses have arrived or the discovery process has timed out. To prevent loops, itMUST NOT<bcp14>MUST NOT</bcp14> relay a Discovery messagewhichthat carries a given cached Session ID and initiator address more than once. These precautions avoid discovery loops and mitigate potential overload.</t> <t>Since the relay device is unaware of the timeout set by the originalinitiatorinitiator, itSHOULD<bcp14>SHOULD</bcp14> set a suitable timeout for the relayeddiscovery. <!-- significantly less than GRASP_DEF_TIMEOUT milliseconds (<xref target="Constants"/>).-->Discovery message. A suggested value is 100 milliseconds multiplied by the remaining loop count.</t> <t>The discovery results received by the relaying instanceMUST<bcp14>MUST</bcp14> in turn be sent as a Discovery Response message to the Discovery message that caused the relay action.</t> </section> <section anchor="rapid"title="Rapidnumbered="true" toc="default"> <name>Rapid Mode (Discovery with Negotiation orSynchronization )">Synchronization)</name> <t>A Discovery messageMAY<bcp14>MAY</bcp14> include anObjectiveobjective option. This allows a rapid mode of negotiation (<xreftarget="rapidneg"/>)target="rapidneg" format="default"/>) or synchronization (<xreftarget="rapidsynch"/>).target="rapidsynch" format="default"/>). Rapid mode is currently limited to a single objective for simplicity of design and implementation. A possible future extension is to allow multiple objectives in rapid mode for greater efficiency. </t> </section> </section> <section anchor="negproc"title="Negotiation Procedures">numbered="true" toc="default"> <name>Negotiation Procedures</name> <t>A negotiation initiator opens a transport connection to a counterpart ASA using the address,protocolprotocol, and port obtained during discovery. It then sends a negotiation request (using M_REQ_NEG) to the counterpart, including a specific negotiation objective. It may request the negotiation counterpart to make a specific configuration. Alternatively, it may request a certain simulation or forecast result by sending adry rundry-run configuration. The details, including the distinction between a dry run and a live configuration change, will be defined separately for each type of negotiation objective. Any state associated with adry rundry-run operation, such as temporarily reserving a resource for subsequent use in a live run, is entirely a matter for the designer of the ASA concerned.</t> <t>Each negotiation session as a whole is subject to a timeout (default GRASP_DEF_TIMEOUT milliseconds, <xreftarget="Constants"/>), initialisedtarget="Constants" format="default"/>), initialized when the request is sent (see <xreftarget="RequestMessage"/>).target="RequestMessage" format="default"/>). If no reply message of any kind is received within the timeout, the negotiation requestMAY<bcp14>MAY</bcp14> berepeated,repeated with a newly generated Session ID (<xreftarget="SessionID"/>).target="SessionID" format="default"/>). An exponential backoffSHOULD<bcp14>SHOULD</bcp14> be used for subsequent repetitions. The details of the backoff algorithm will depend on the use case for the objective concerned.</t><t/><t>If the counterpart can immediately apply the requested configuration, it will give an immediate positive (O_ACCEPT) answer(using M_END).using the Negotiation End (M_END) message. This will end the negotiation phase immediately. Otherwise, it will negotiate (using M_NEGOTIATE). It will reply with a proposed alternative configuration that it can apply (typically, a configuration that uses fewer resources than requested by the negotiation initiator). This will start abi-directionalbidirectional negotiation(using M_NEGOTIATE)using the Negotiate (M_NEGOTIATE) message to reach a compromise between the two ASAs.</t> <t>The negotiation procedure is ended when one of the negotiation peers sends a NegotiationEndingEnd (M_END) message, which contains anacceptAccept (O_ACCEPT) ordeclineDecline (O_DECLINE) option and does not need a response from the negotiation peer. Negotiation may also end in failure (equivalent to a decline) if a timeout is exceeded or a loop count is exceeded. When the procedure ends for whatever reason, the transport connectionSHOULD<bcp14>SHOULD</bcp14> be closed. A transport session failure is treated as a negotiation failure.</t> <t>A negotiation procedure concerns one objective and one counterpart. Both the initiator and the counterpart may take part in simultaneous negotiations with various otherASAs,ASAs or in simultaneous negotiations about different objectives. Thus, GRASP is expected to be used in amulti-threadedmultithreaded mode or its logical equivalent. Certain negotiation objectives may have restrictions onmulti-threading,multithreading, for example to avoid over-allocating resources. </t> <t>Some configuration actions, forexampleexample, wavelength switching in optical networks, might take considerable time to execute. The ASA concerned needs to allow for this by design, but GRASP does allow for a peer to insert latency in a negotiation process if necessary (<xreftarget="ConfirmWaitingMessage"/>,target="ConfirmWaitingMessage" format="default"/>, M_WAIT).</t> <section anchor="rapidneg"title="Rapidnumbered="true" toc="default"> <name>Rapid Mode (Discovery/NegotiationLinkage)">Linkage)</name> <t>A Discovery messageMAY<bcp14>MAY</bcp14> include a Negotiation Objective option. In thiscasecase, it is as if the initiator sent the sequenceM_DISCOVERY,M_DISCOVERY immediately followed by M_REQ_NEG. This has implications for the construction of the GRASP core, as it must carefully pass the contents of the Negotiation Objective option to the ASA so that it may evaluate the objective directly. When a Negotiation Objective option ispresentpresent, the ASA replies with an M_NEGOTIATE message (or M_END with O_ACCEPT if it is immediately satisfied with theproposal),proposal) rather than with an M_RESPONSE. However, if the recipient node does not support rapid mode, discovery will continue normally.</t> <t>It is possible that a Discovery Response will arrive from a responder that does not support rapidmode,mode before such a Negotiation message arrives. In this case, rapid mode will not occur.</t> <t>This rapid mode could reduce the interactions between nodes so that a higher efficiency could be achieved. However, a network in which some nodes support rapid mode and others do not will have complex timing-dependent behaviors. Therefore, the rapid negotiation functionSHOULD<bcp14>SHOULD</bcp14> be disabled by default. </t> </section> </section> <section anchor="synchproc"title="Synchronizationnumbered="true" toc="default"> <name>Synchronization and FloodingProcedures">Procedures</name> <section anchor="synch"title="Unicast Synchronization">numbered="true" toc="default"> <name>Unicast Synchronization</name> <t>A synchronization initiator opens a transport connection to a counterpart ASA using the address,protocolprotocol, and port obtained during discovery. It then sends asynchronization request (using M_REQ_SYN)Request Synchronization message (M_REQ_SYN, <xref target="RequestMessage" format="default"/>) to the counterpart, including a specific synchronization objective. The counterpart responds with a Synchronization message (M_SYNCH, <xreftarget="SynchMessage"/>)target="SynchMessage" format="default"/>) containing the current value of the requested synchronization objective. No further messages areneededneeded, and the transport connectionSHOULD<bcp14>SHOULD</bcp14> be closed. A transport session failure is treated as a synchronization failure.</t> <t>If no reply message of any kind is received within a given timeout (default GRASP_DEF_TIMEOUT milliseconds, <xreftarget="Constants"/>),target="Constants" format="default"/>), the synchronization requestMAY<bcp14>MAY</bcp14> berepeated,repeated with a newly generated Session ID (<xreftarget="SessionID"/>).target="SessionID" format="default"/>). An exponential backoffSHOULD<bcp14>SHOULD</bcp14> be used for subsequent repetitions. The details of the backoff algorithm will depend on the use case for the objective concerned.</t> </section> <section anchor="flooding"title="Flooding">numbered="true" toc="default"> <name>Flooding</name> <t>In the case just described, the message exchange is unicast and concerns only one synchronization objective. For large groups of nodes requiring the same data, synchronization flooding is available. For this, a flooding initiatorMAY<bcp14>MAY</bcp14> send an unsolicited Flood Synchronization message (<xref target="FloodMessage" format="default"/>) containing one or more Synchronization Objective option(s), if and only if the specification of those objectives permits it. This is sent as a multicast message to the ALL_GRASP_NEIGHBORS multicast address (<xreftarget="Constants"/>).</t>target="Constants" format="default"/>).</t> <t>Receiving flood multicasts is a function of the GRASP core, as in the case of discovery multicasts (<xreftarget="discproc"/>).</t>target="discproc" format="default"/>).</t> <t>To ensure that flooding does not result in a loop, the originator of the Flood Synchronization messageMUST<bcp14>MUST</bcp14> set the loop count in the objectives to a suitable value (the default is GRASP_DEF_LOOPCT). Also, a suitable mechanism is needed to avoid excessive multicast traffic. This mechanismMUST<bcp14>MUST</bcp14> be defined as part of the specification of the synchronization objective(s) concerned. It might be a simple rate limit or a more complex mechanism such as the Trickle algorithm <xreftarget="RFC6206"/>.</t>target="RFC6206" format="default"/>.</t> <t>A GRASP device with multiple link-layer interfaces (typically a router)MUST<bcp14>MUST</bcp14> support synchronization flooding on all GRASP interfaces. If it receives a multicast Flood Synchronization message on a given interface, itMUST<bcp14>MUST</bcp14> relay it byre-issuingreissuing a Flood Synchronization message as a link-local multicast on its other GRASP interfaces. The relayed messageMUST<bcp14>MUST</bcp14> have the same Session ID as the incoming message andMUST<bcp14>MUST</bcp14> be tagged with the IP address of its original initiator. </t> <t>Link-layerFloodingflooding is supported by GRASP by setting the loop count to1,1 and sending with a link-local source address. Floods with link-local source addresses and a loop count other than 1 are invalid, and such messagesMUST<bcp14>MUST</bcp14> be discarded.</t> <t>The relaying deviceMUST<bcp14>MUST</bcp14> decrement the loop count within the firstobjective,objective andMUST NOT<bcp14>MUST NOT</bcp14> relay the Flood Synchronization message if the result is zero. Also, itMUST<bcp14>MUST</bcp14> limit the total rate at which it relays Flood Synchronization messages to a reasonable value, in order to mitigate possibledenial of servicedenial-of-service attacks. For example, the rate limit could be set to a small multiple of the observed rate of flood messages during normal operation. The relaying deviceMUST<bcp14>MUST</bcp14> cache the Session ID value and initiator address of each relayed Flood Synchronization message for a time not less than twice GRASP_DEF_TIMEOUT milliseconds. To prevent loops, itMUST NOT<bcp14>MUST NOT</bcp14> relay a Flood Synchronization messagewhichthat carries a given cached Session ID and initiator address more than once. These precautions avoid synchronization loops and mitigate potential overload.</t> <t>Note that this mechanism is unreliable in the case of sleeping nodes, or new nodes that join the network, or nodes that rejoin the network after a fault. An ASA that initiates a floodSHOULD<bcp14>SHOULD</bcp14> repeat the flood at a suitable frequency, whichMUST<bcp14>MUST</bcp14> be consistent with the recommendations in <xreftarget="RFC8085"/>target="RFC8085" format="default"/> for low data-volume multicast. The ASASHOULD<bcp14>SHOULD</bcp14> also act as a synchronization responder for the objective(s) concerned. Thus nodes that require an objective subject to flooding can either wait for the next flood or request unicast synchronization for that objective. </t> <t>The multicast messages for synchronization flooding are subject to the security rules in <xreftarget="reqsec"/>.target="reqsec" format="default"/>. Inpracticepractice, this means that theyMUST NOT<bcp14>MUST NOT</bcp14> be transmitted andMUST<bcp14>MUST</bcp14> be ignored on receipt unless there is an operational ACP or equivalent strong security in place. However, because of the security weakness of link-local multicast (<xreftarget="security"/>),target="security" format="default"/>), synchronization objectives that are floodedSHOULD NOT<bcp14>SHOULD NOT</bcp14> contain unencrypted private information andSHOULD<bcp14>SHOULD</bcp14> be validated by the recipient ASA.</t> </section> <section anchor="rapidsynch"title="Rapidnumbered="true" toc="default"> <name>Rapid Mode (Discovery/SynchronizationLinkage)">Linkage)</name> <t>A Discovery messageMAY<bcp14>MAY</bcp14> include a Synchronization Objective option. In thiscasecase, the Discovery message also acts as a Request Synchronization message to indicate to theDiscovery Responderdiscovery responder that it could directly reply to theDiscovery Initiatordiscovery initiator with a Synchronization message<xref target="SynchMessage"/>(<xref target="SynchMessage" format="default"/>) with synchronization data for rapid processing, if the discovery target supports the corresponding synchronization objective. The design implications are similar to those discussed in <xreftarget="rapidneg"/>.</t>target="rapidneg" format="default"/>.</t> <t>It is possible that a Discovery Response will arrive from a responder that does not support rapidmode,mode before such a Synchronization message arrives. In this case, rapid mode will not occur.</t> <t>This rapid mode could reduce the interactions between nodes so that a higher efficiency could be achieved. However, a network in which some nodes support rapid mode and others do not will have complex timing-dependent behaviors. Therefore, the rapid synchronization functionSHOULD<bcp14>SHOULD</bcp14> be configured off by default andMAY<bcp14>MAY</bcp14> be configured on or off by Intent.</t> </section> </section> </section> <section anchor="Constants"title="GRASP Constants"> <t><list style="symbols"> <t>ALL_GRASP_NEIGHBORS<vspace blankLines="1"/>Anumbered="true" toc="default"> <name>GRASP Constants</name> <dl newline="true"> <dt>ALL_GRASP_NEIGHBORS </dt> <dd> <t>A link-local scope multicast address used by a GRASP-enabled device to discover GRASP-enabled neighbor (i.e., on-link) devices. All devices that support GRASP are members of this multicastgroup.<list style="symbols"> <t>IPv6group.</t> <ul spacing="normal"> <li>IPv6 multicast address:TBD1</t> <t>IPv4ff02::13</li> <li>IPv4 multicast address:TBD2</t> </list></t> <t>GRASP_LISTEN_PORT (TBD3)<vspace blankLines="1"/>A224.0.0.119</li> </ul> </dd> <dt>GRASP_LISTEN_PORT (7017) </dt> <dd> <t>A well-known UDP user port that every GRASP-enabled network deviceMUST<bcp14>MUST</bcp14> listen to for link-local multicasts when UDP is used for M_DISCOVERY or M_FLOOD messages in the GRASPinstanceinstance. This user portMAY<bcp14>MAY</bcp14> also be used to listen for TCP or UDP unicast messages in a simple implementation of GRASP (<xreftarget="trans"/>).</t> <t>GRASP_DEF_TIMEOUTtarget="trans" format="default"/>).</t> </dd> <dt>GRASP_DEF_TIMEOUT (60000milliseconds)<vspace blankLines="1"/>Themilliseconds) </dt> <dd><t>The default timeout used to determine that an operation has failed to complete.</t><t>GRASP_DEF_LOOPCT (6)<vspace blankLines="1"/>The</dd> <dt>GRASP_DEF_LOOPCT (6) </dt> <dd><t>The default loop count used to determine that a negotiation has failed tocomplete,complete and to avoid looping messages.</t><t>GRASP_DEF_MAX_SIZE (2048)<vspace blankLines="1"/>The</dd> <dt>GRASP_DEF_MAX_SIZE (2048) </dt> <dd><t>The default maximum message size in bytes.</t></list></t></dd> </dl> </section> <section anchor="SessionID"title="Sessionnumbered="true" toc="default"> <name>Session Identifier (SessionID)">ID)</name> <t>This is an up to 32-bit opaque value used to distinguish multiple sessions between the same two devices. A new Session IDMUST<bcp14>MUST</bcp14> be generated by the initiator for every new Discovery, FloodSynchronizationSynchronization, or Request message. All responses and follow-up messages in the same discovery,synchronizationsynchronization, or negotiation procedureMUST<bcp14>MUST</bcp14> carry the same Session ID.</t> <t>The Session IDSHOULD<bcp14>SHOULD</bcp14> have a very low collision rate locally. ItMUST<bcp14>MUST</bcp14> be generated by apseudo-randompseudorandom number generator (PRNG) using a locally generated seedwhichthat is unlikely to be used by any other device in the same network. The PRNGSHOULD<bcp14>SHOULD</bcp14> be cryptographically strong <xreftarget="RFC4086"/>.target="RFC4086" format="default"/>. When allocating a new Session ID, GRASPMUST<bcp14>MUST</bcp14> check that the value is not already in use andSHOULD<bcp14>SHOULD</bcp14> check that it has not been usedrecently,recently by consulting a cache of current and recent sessions. In the unlikely event of a clash, GRASPMUST<bcp14>MUST</bcp14> generate a new value.</t> <t>However, there is a finite probability that two nodes might generate the same Session ID value. For that reason, when a Session ID is communicated via GRASP, the receiving nodeMUST<bcp14>MUST</bcp14> tag it with the initiator's IP address to allow disambiguation. In the highly unlikely event of two peers opening sessions with the same Session ID value, this tag will allow the two sessions to be distinguished. Multicast GRASP messages and their responses, which may be relayed between links, therefore include a field that carries the initiator's global IP address.</t> <t>There is a highly unlikely race condition in which two peers start simultaneous negotiation sessions with each other using the same Session ID value. Depending on various implementation choices, this might lead to the two sessions being confused. See <xreftarget="RequestMessage"/>target="RequestMessage" format="default"/> for details of how to avoid this.</t> </section> <section anchor="GRASPMessages"title="GRASP Messages">numbered="true" toc="default"> <name>GRASP Messages</name> <sectiontitle="Message Overview">numbered="true" toc="default"> <name>Message Overview</name> <t>This section defines the GRASP message format and message types. Message types not listed here are reserved for future use. </t> <t>The messages currently defined are:<list style="bullets"> <t>Discovery</t> <ul spacing="normal" empty="true"> <li>Discovery and Discovery Response (M_DISCOVERY,M_RESPONSE).</t> <t>RequestM_RESPONSE).</li> <li>Request Negotiation, Negotiation, ConfirmWaitingWaiting, and Negotiation End (M_REQ_NEG, M_NEGOTIATE, M_WAIT,M_END).</t> <t>RequestM_END).</li> <li>Request Synchronization, Synchronization, and Flood Synchronization (M_REQ_SYN, M_SYNCH,M_FLOOD.</t> <t>NoM_FLOOD).</li> <li>No Operation and Invalid (M_NOOP,M_INVALID).</t> </list></t>M_INVALID).</li> </ul> </section> <sectiontitle="GRASPnumbered="true" toc="default"> <name>GRASP MessageFormat">Format</name> <t>GRASP messages share an identical header format and a variable format area for options. GRASP message headers and options are transmitted in Concise Binary Object Representation (CBOR) <xreftarget="RFC7049"/>.target="RFC8949" format="default"/>. In this specification, they are described usingCBOR data definition languageConcise Data Definition Language (CDDL) <xreftarget="I-D.greevenbosch-appsawg-cbor-cddl"/>.target="RFC8610" format="default"/>. Fragmentary CDDL is used to describe each item in this section. A complete and normative CDDL specification of GRASP is given in <xreftarget="cddl"/>,target="cddl" format="default"/>, including constants such as message types. </t> <t>Every GRASP message, except the No Operation message, carries a Session ID (<xreftarget="SessionID"/>).target="SessionID" format="default"/>). Options are then presentedserially in the options field.</t>serially.</t> <t>In fragmentary CDDL, every GRASP message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ grasp-message = (message .within message-structure) / noop-message message-structure = [MESSAGE_TYPE, session-id, ?initiator, *grasp-option] MESSAGE_TYPE =1..2550..255 session-id = 0..4294967295;up; up to 32 bits grasp-option = any]]></artwork> </figure></t>]]></sourcecode> <t>The MESSAGE_TYPE indicates the type of the message and thus defines the expected options. Any options received that are not consistent with the MESSAGE_TYPESHOULD<bcp14>SHOULD</bcp14> be silently discarded. </t> <t>The No Operation (noop) message is described in <xreftarget="noop"/>.</t>target="noop" format="default"/>.</t> <t>The various MESSAGE_TYPE values are defined in <xreftarget="cddl"/>.</t>target="cddl" format="default"/>.</t> <t>All other message elements are described below and formally defined in <xreftarget="cddl"/>.</t>target="cddl" format="default"/>.</t> <t>If an unrecognized MESSAGE_TYPE is received in a unicast message, an Invalid message (<xreftarget="invalid"/>) MAYtarget="invalid" format="default"/>) <bcp14>MAY</bcp14> be returned.OtherwiseOtherwise, the messageMAY<bcp14>MAY</bcp14> be logged andMUST<bcp14>MUST</bcp14> be discarded. If an unrecognized MESSAGE_TYPE is received in a multicast message, itMAY<bcp14>MAY</bcp14> be logged andMUST<bcp14>MUST</bcp14> be silently discarded.</t> </section> <sectiontitle="Message Size">numbered="true" toc="default"> <name>Message Size</name> <t>GRASP nodesMUST<bcp14>MUST</bcp14> be able to receive unicast messages of at least GRASP_DEF_MAX_SIZE bytes. GRASP nodesMUST NOT<bcp14>MUST NOT</bcp14> send unicast messages longer than GRASP_DEF_MAX_SIZE bytes unless a longer size is explicitly allowed for the objective concerned. For example, GRASP negotiation itself could be used to agree on a longer message size.</t> <t>The message parser used by GRASP should be configured to know about the GRASP_DEF_MAX_SIZE, or any larger negotiated message size, so that it may defend against overly long messages.</t> <t>The maximum size of multicast messages (M_DISCOVERY and M_FLOOD) depends on thelink layerlink-layer technology orlink adaptationthe link-adaptation layer in use.</t> </section> <section anchor="DiscoveryMessage"title="Discovery Message">numbered="true" toc="default"> <name>Discovery Message</name> <t>In fragmentary CDDL, a Discovery message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ discovery-message = [M_DISCOVERY, session-id, initiator, objective]]]></artwork> </figure></t>]]></sourcecode> <t> A discovery initiator sends a Discovery message to initiate a discovery process for a particular objective option.</t><t></t> <t> The discovery initiator sends all Discovery messages via UDP to port GRASP_LISTEN_PORT at the link-local ALL_GRASP_NEIGHBORS multicast address on each link-layer interface in use by GRASP. It then listens for unicast TCP responses on a givenport,port and stores the discoveryresults (includingresults, including responding discovery objectives and corresponding unicastlocators).locators. </t> <t>The listening port used for TCPMUST<bcp14>MUST</bcp14> be the same port as used for sending the Discovery UDP multicast, on a given interface. In an implementation with a single GRASP instance in anodenode, thisMAY<bcp14>MAY</bcp14> be GRASP_LISTEN_PORT. To support multiple instances in the same node, the GRASP discovery mechanism in each instance needs to find, for each interface, a dynamic port that it can bind to for both sending UDP link-local multicast and listening forTCP,TCP before initiating any discovery.</t> <t> The 'initiator' field in the message is a globally unique IP address of theinitiator,initiator for the sole purpose of disambiguating the Session ID in other nodes. If for some reason the initiator does not have a globally unique IP address, itMUST<bcp14>MUST</bcp14> use a link-local addressfor this purposethat is highly likely to beunique,unique forexamplethis purpose, for example, using <xreftarget="RFC7217"/>.target="RFC7217" format="default"/>. Determination of a node's globally unique IP address isimplementation-dependent. </t><t>implementation dependent. </t> <t> A Discovery messageMUST<bcp14>MUST</bcp14> include exactly one of the following:<list style="symbols"> <t>a discovery objective</t> <ul spacing="normal"> <li>A Discovery Objective option (<xreftarget="ObjForm"/>).target="ObjForm" format="default"/>). Its loop countMUST<bcp14>MUST</bcp14> be set to a suitable value to prevent discovery loops (default value is GRASP_DEF_LOOPCT). If the discovery initiator requires only on-link responses, the loop countMUST<bcp14>MUST</bcp14> be set to 1.</t> <t>a negotiation objective</li> <li>A Negotiation Objective option (<xreftarget="ObjForm"/>).target="ObjForm" format="default"/>). This is used both for the purpose of discovery and to indicate to the discovery target that itMAY<bcp14>MAY</bcp14> directly reply to the discoveryinitiatiorinitiator with a Negotiation message for rapid processing, if it could act as the corresponding negotiation counterpart. The sender of such a Discovery messageMUST<bcp14>MUST</bcp14> initialize a negotiation timer and loop count in the same way as a Request Negotiation message (<xreftarget="RequestMessage"/>). </t> <t>a synchronization objectivetarget="RequestMessage" format="default"/>). </li> <li>A Synchronization Objective option (<xreftarget="ObjForm"/>).target="ObjForm" format="default"/>). This is used both for the purpose of discovery and to indicate to the discovery target that itMAY<bcp14>MAY</bcp14> directly reply to the discovery initiator with a Synchronization message for rapid processing, if it could act as the corresponding synchronization counterpart. Its loop countMUST<bcp14>MUST</bcp14> be set to a suitable value to prevent discovery loops (default value isGRASP_DEF_LOOPCT).</t> </list></t>GRASP_DEF_LOOPCT).</li> </ul> <t>As mentioned in <xreftarget="discovw"/>,target="discovw" format="default"/>, a Discovery messageMAY<bcp14>MAY</bcp14> be sent unicast to a peer node, whichSHOULD<bcp14>SHOULD</bcp14> then proceed exactly as if the message had been multicast. </t> </section> <section anchor="ResponseMessage"title="Discoverynumbered="true" toc="default"> <name>Discovery ResponseMessage">Message</name> <t>In fragmentary CDDL, a Discovery Response message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ response-message = [M_RESPONSE, session-id, initiator, ttl, (+locator-option // divert-option),?objective)]?objective] ttl = 0..4294967295 ; in milliseconds]]></artwork> </figure></t>]]></sourcecode> <t> A nodewhichthat receives a Discovery messageSHOULD<bcp14>SHOULD</bcp14> send a Discovery Response message if and only if it can respond to the discovery.<list> <t>It MUST</t> <ul spacing="normal" empty="true"> <li>It <bcp14>MUST</bcp14> contain the same Session ID and initiator as the Discovery message.</t><t>It MUST</li> <li>It <bcp14>MUST</bcp14> contain a time-to-live (ttl) for the validity of the response, given as a positive integer value in milliseconds. Zero implies a value significantly greater than GRASP_DEF_TIMEOUT milliseconds (<xreftarget="Constants"/>).target="Constants" format="default"/>). A suggested value is ten times that amount.</t><t>It MAY</li> <li>It <bcp14>MAY</bcp14> include a copy of the discovery objective from the Discoverymessage.</t> </list>message.</li> </ul> <t> It is sent to the sender of the Discovery message via TCP at the port used to send the Discovery message (as explained in <xreftarget="DiscoveryMessage"/>).target="DiscoveryMessage" format="default"/>). In the case of a relayed Discovery message, the Discovery Response is thus sent to the relay, not the original initiator.</t><t></t> <t> In all cases, the transport sessionSHOULD<bcp14>SHOULD</bcp14> be closed after sending the Discovery Response. A transport session failure is treated as no response.</t><t></t> <t> If the responding node supports the discovery objective of the discovery, itMUST<bcp14>MUST</bcp14> include at least one kind of locator option (<xreftarget="LocatorOption"/>)target="LocatorOption" format="default"/>) to indicate its own location. A sequence of multiple kinds of locator options(e.g.(e.g., IP address option and FQDN option) is also valid.</t><t></t> <t> If the responding node itself does not support the discovery objective, but it knows the locator of the discovery objective, then itSHOULD<bcp14>SHOULD</bcp14> respond to thediscoveryDiscovery message with adivertDivert option (<xreftarget="DivertOption"/>)target="DivertOption" format="default"/>) embedding a locator option or a combination of multiple kinds of locator optionswhichthat indicate the locator(s) of the discovery objective. </t> <t>More details on the processing of Discovery Responses are given in <xreftarget="discmech"/>.</t>target="discmech" format="default"/>.</t> </section> <section anchor="RequestMessage"title="Request Messages">numbered="true" toc="default"> <name>Request Messages</name> <t>In fragmentary CDDL, Request Negotiation and Request Synchronization messages follow the patterns:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ request-negotiation-message = [M_REQ_NEG, session-id, objective] request-synchronization-message = [M_REQ_SYN, session-id, objective]]]></artwork> </figure></t>]]></sourcecode> <t> A negotiation or synchronization requesting node sends the appropriate Request message to the unicast address of the negotiation or synchronization counterpart, using the appropriate protocol and port numbers (selected from the discovery result). If the discovery result is an FQDN, it will be resolved first.</t> <t>A Request messageMUST<bcp14>MUST</bcp14> include the relevant objective option. In the case of Request Negotiation, the objective optionMUST<bcp14>MUST</bcp14> include the requested value. </t> <t>When an initiator sends a Request Negotiation message, itMUST<bcp14>MUST</bcp14> initialize a negotiation timer for the new negotiation thread. The default is GRASP_DEF_TIMEOUT milliseconds. Unless this timeout is modified by a Confirm Waiting message (<xreftarget="ConfirmWaitingMessage"/>),target="ConfirmWaitingMessage" format="default"/>), the initiator will consider that the negotiation has failed when the timer expires. </t> <t>Similarly, when an initiator sends a Request Synchronization, itSHOULD<bcp14>SHOULD</bcp14> initialize a synchronization timer. The default is GRASP_DEF_TIMEOUT milliseconds. The initiator will consider that synchronization has failed if there is no response before the timer expires.</t> <t>When an initiator sends a Request message, itMUST<bcp14>MUST</bcp14> initialize the loop count of the objective option with a value defined in the specification of the option or, if no such value is specified, with GRASP_DEF_LOOPCT. </t> <t>If a node receives a Request message for an objective for which no ASA is currently listening, itMUST<bcp14>MUST</bcp14> immediately close the relevant socket to indicate this to the initiator. This is to avoid unnecessary timeouts if, for example, an ASA exits prematurely but the GRASP core is listening on its behalf.</t> <t>To avoid the highly unlikely race condition in which two nodes simultaneously request sessions with each other using the same Session ID (<xreftarget="SessionID"/>), whentarget="SessionID" format="default"/>), a nodereceives a Request message, it MUST<bcp14>MUST</bcp14> verify that the received Session ID is not already locallyactive.active when it receives a Request message. In case of a clash, itMUST<bcp14>MUST</bcp14> discard the Request message, in which case the initiator will detect a timeout.</t> </section> <section anchor="NegotiationMessage"title="Negotiation Message">numbered="true" toc="default"> <name>Negotiation Message</name> <t>In fragmentary CDDL, a Negotiation message follows the pattern:</t><t><figure> <artwork><![CDATA[ negotiate-message<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ negotiation-message = [M_NEGOTIATE, session-id, objective]]]></artwork> </figure></t>]]></sourcecode> <t>A negotiation counterpart sends a Negotiation message in response to a Request Negotiation message, a Negotiation message, or a Discovery message inRapid Mode.rapid mode. A negotiation processMAY<bcp14>MAY</bcp14> include multiple steps.</t> <t>The Negotiation messageMUST<bcp14>MUST</bcp14> include the relevant Negotiation Objective option, with its value updated according to progress in the negotiation. The senderMUST<bcp14>MUST</bcp14> decrement the loop count by 1. If the loop count becomeszerozero, the messageMUST NOT<bcp14>MUST NOT</bcp14> be sent. In thiscasecase, the negotiation session has failed and will time out.</t> </section> <section anchor="NegotiationEndingMessage"title="Negotiationnumbered="true" toc="default"> <name>Negotiation EndMessage">Message</name> <t>In fragmentary CDDL, a Negotiation End message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ end-message = [M_END, session-id, accept-option / decline-option]]]></artwork> </figure></t>]]></sourcecode> <t> A negotiation counterpart sendsana Negotiation End message to close the negotiation. ItMUST<bcp14>MUST</bcp14> contain either anacceptAccept option or adeclineDecline option, defined in <xreftarget="AcceptOption"/>target="AcceptOption" format="default"/> and <xreftarget="DeclineOption"/>.target="DeclineOption" format="default"/>. It could be sent either by the requesting node or the responding node.</t> </section> <section anchor="ConfirmWaitingMessage"title="Confirmnumbered="true" toc="default"> <name>Confirm WaitingMessage">Message</name> <t>In fragmentary CDDL, a Confirm Waiting message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ wait-message = [M_WAIT, session-id, waiting-time] waiting-time = 0..4294967295 ; in milliseconds]]></artwork> </figure></t>]]></sourcecode> <t> A responding node sends a Confirm Waiting message to ask the requesting node to wait for a further negotiation response. It might be that the local process needs more time or that the negotiation depends on another triggered negotiation. This messageMUST NOT<bcp14>MUST NOT</bcp14> include any other options. When received, the waiting time value overwrites and restarts the current negotiation timer (<xreftarget="RequestMessage"/>).</t>target="RequestMessage" format="default"/>).</t> <t>The responding nodeSHOULD<bcp14>SHOULD</bcp14> send a Negotiation, NegotiationEndEnd, or another Confirm Waiting message before the negotiation timer expires. If not, when the initiator's timer expires, the initiatorMUST<bcp14>MUST</bcp14> treat the negotiation procedure as failed.</t> </section> <section anchor="SynchMessage"title="Synchronization Message">numbered="true" toc="default"> <name>Synchronization Message</name> <t>In fragmentary CDDL, a Synchronization message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ synch-message = [M_SYNCH, session-id, objective]]]></artwork> </figure></t>]]></sourcecode> <t>A nodewhichthat receives a Request Synchronization, or a Discovery message inRapid Mode,rapid mode, sends back a unicast Synchronization message with the synchronization data, in the form of a GRASPOptionoption for the specific synchronization objective present in the Request Synchronization.</t> </section> <section anchor="FloodMessage"title="Floodnumbered="true" toc="default"> <name>Flood SynchronizationMessage">Message</name> <t>In fragmentary CDDL, a Flood Synchronization message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ flood-message = [M_FLOOD, session-id, initiator, ttl, +[objective, (locator-option / [])]] ttl = 0..4294967295 ; in milliseconds]]></artwork> </figure></t>]]></sourcecode> <t> A nodeMAY<bcp14>MAY</bcp14> initiate flooding by sending an unsolicited Flood SynchronizationMessagemessage with synchronization data. ThisMAY<bcp14>MAY</bcp14> be sent to port GRASP_LISTEN_PORT at the link-local ALL_GRASP_NEIGHBORS multicast address, in accordance with the rules in <xreftarget="synchproc"/>. <list><t>target="synchproc" format="default"/>. </t> <ul empty="true" spacing="normal"> <li> The initiator address is provided, as described for Discovery messages (<xreftarget="DiscoveryMessage"/>),target="DiscoveryMessage" format="default"/>), only to disambiguate the Session ID.</t><t></li> <li> The messageMUST<bcp14>MUST</bcp14> contain a time-to-live (ttl) for the validity of the contents, given as a positive integer value in milliseconds. There is no default; zero indicates an indefinite lifetime.</t><t></li> <li> The synchronization data are in the form of GRASPOption(s)option(s) for specific synchronization objective(s). The loop count(s)MUST<bcp14>MUST</bcp14> be set to a suitable value to prevent flood loops (default value isGRASP_DEF_LOOPCT).</t><t>GRASP_DEF_LOOPCT).</li> <li> Each objective optionMAY<bcp14>MAY</bcp14> be followed by a locator option (<xref target="LocatorOption" format="default"/>) associated with the flooded objective. In its absence, an empty optionMUST<bcp14>MUST</bcp14> be included to indicate a null locator.</t> </list></li> </ul> <t> A node that receives a Flood Synchronization messageMUST<bcp14>MUST</bcp14> cache the received objectives for use by local ASAs. Each cached objectiveMUST<bcp14>MUST</bcp14> be tagged with the locator option sent with it, or with a null tag if an empty locator option was sent. If a subsequent Flood Synchronization messagecarryingcarries an objective with the same name and the same tag, the corresponding cached copy of the objectiveMUST<bcp14>MUST</bcp14> be overwritten. If a subsequent Flood Synchronization message carrying an objective with same name arrives with a different tag, a new cached entryMUST<bcp14>MUST</bcp14> be created.</t> <t>Note: the purpose of this mechanism is to allow the recipient of flooded values to distinguish between different senders of the same objective, and if necessary communicate with them using the locator,protocolprotocol, and port included in the locator option. Many objectives will not need this mechanism, so they will be flooded with a null locator.</t> <t>Cached entriesMUST<bcp14>MUST</bcp14> be ignored or deleted after their lifetime expires.</t> </section> <section anchor="invalid"title="Invalid Message">numbered="true" toc="default"> <name>Invalid Message</name> <t>In fragmentary CDDL, an Invalid message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ invalid-message = [M_INVALID, session-id, ?any]]]></artwork> </figure></t>]]></sourcecode> <t> This messageMAY<bcp14>MAY</bcp14> be sent by an implementation in response to an incoming unicast message that it considers invalid. Thesession-id MUSTSession ID value <bcp14>MUST</bcp14> be copied from the incoming message. The contentSHOULD<bcp14>SHOULD</bcp14> be diagnostic information such as a partial copy of the invalid message up to the maximum message size. An M_INVALID messageMAY<bcp14>MAY</bcp14> be silently ignored by a recipient. However, it could be used in support of extensibility, since it indicates that the remote node does not support a new or obsolete message or option.</t> <t>An M_INVALID messageMUST NOT<bcp14>MUST NOT</bcp14> be sent in response to an M_INVALID message.</t> </section> <section anchor="noop"title="Nonumbered="true" toc="default"> <name>No OperationMessage">Message</name> <t>In fragmentary CDDL, a No Operation message follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ noop-message = [M_NOOP]]]></artwork> </figure></t>]]></sourcecode> <t> This messageMAY<bcp14>MAY</bcp14> be sent by an implementation that for practical reasons needs to initialize a socket. ItMUST<bcp14>MUST</bcp14> be silently ignored by a recipient.</t> </section> </section> <section anchor="GRASPOptions"title="GRASP Options">numbered="true" toc="default"> <name>GRASP Options</name> <t>This section defines the GRASP options for the negotiation and synchronization protocol signaling. Additional options may be defined in the future.</t> <sectiontitle="Formatnumbered="true" toc="default"> <name>Format of GRASPOptions">Options</name> <t>GRASP optionsare<bcp14>SHOULD</bcp14> be CBORobjectsarrays thatMUST<bcp14>MUST</bcp14> start with an unsigned integer identifying the specific option type carried in this option. These option types are formally defined in <xreftarget="cddl"/>. Apart from that the only format requirement is that each option MUST be a well-formed CBOR object. In general a CBOR array format is RECOMMENDED to limit overhead.</t>target="cddl" format="default"/>.</t> <t>GRASP options may be defined to include encapsulated GRASP options.</t> </section> <section anchor="DivertOption"title="Divert Option">numbered="true" toc="default"> <name>Divert Option</name> <t>The Divert option is used to redirect a GRASP request to another node, which may be more appropriate for the intended negotiation or synchronization. It may redirect to an entity that is known as a specific negotiation or synchronization counterpart (on-link or off-link) or a default gateway. ThedivertDivert optionMUST<bcp14>MUST</bcp14> only be encapsulated in Discovery Response messages. If found elsewhere, itSHOULD<bcp14>SHOULD</bcp14> be silently ignored.</t> <t>A discovery initiatorMAY<bcp14>MAY</bcp14> ignore a Divert option if it only requires directdiscovery responses.Discovery Responses. </t> <t>In fragmentary CDDL, the Divert option follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ divert-option = [O_DIVERT, +locator-option]]]></artwork> </figure></t>]]></sourcecode> <t>The embeddedLocator Option(s)locator option(s) (<xreftarget="LocatorOption"/>)target="LocatorOption" format="default"/>) point to diverted destination target(s) in response to a Discovery message. </t> </section> <section anchor="AcceptOption"title="Accept Option">numbered="true" toc="default"> <name>Accept Option</name> <t>TheacceptAccept option is used to indicate to the negotiation counterpart that the proposed negotiation content is accepted.</t> <t>TheacceptAccept optionMUST<bcp14>MUST</bcp14> only be encapsulated in Negotiation End messages. If found elsewhere, itSHOULD<bcp14>SHOULD</bcp14> be silently ignored.</t> <t>In fragmentary CDDL, the Accept option follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ accept-option = [O_ACCEPT]]]></artwork> </figure></t>]]></sourcecode> </section> <section anchor="DeclineOption"title="Decline Option">numbered="true" toc="default"> <name>Decline Option</name> <t>ThedeclineDecline option is used to indicate to the negotiation counterpart the proposed negotiation content is declined and to end the negotiation process.</t> <t>ThedeclineDecline optionMUST<bcp14>MUST</bcp14> only be encapsulated in Negotiation End messages. If found elsewhere, itSHOULD<bcp14>SHOULD</bcp14> be silently ignored.</t> <t>In fragmentary CDDL, the Decline option follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ decline-option = [O_DECLINE, ?reason] reason = text;optional; optional UTF-8 error message]]></artwork> </figure></t>]]></sourcecode> <t>Note: there might be scenarios where an ASA wants to decline the proposed value and restart the negotiation process. In thiscasecase, it is an implementation choice whether to send a Decline option or to continue with aNegotiateNegotiation message, with an objective option that contains a nullvalue,value or one that contains a new value that might achieve convergence.</t> </section> <section anchor="LocatorOption"title="Locator Options">numbered="true" toc="default"> <name>Locator Options</name> <t>These locator options are used to present reachability information for an ASA, adevicedevice, or an interface. They are Locator IPv6 AddressOption,option, Locator IPv4 AddressOption,option, Locator FQDN(Fully Qualified Domain Name) Optionoption, and Locator URI(Uniform Resource Identifier) Option.</t>option.</t> <t>Since ASAs will normally run as independent user programs, locator options need to indicate thenetwork layernetwork-layer locator plus the transport protocol and port number for reaching the target. For this reason, theLocator Optionslocator options for IP addresses and FQDNs include this information explicitly. In the case of the Locator URIOption,option, this information can be encoded in the URI itself.</t> <t>Note: It is assumed that all locators used in locator options are in scope throughout the GRASP domain. As stated in <xreftarget="hilev"/>,target="hilev" format="default"/>, GRASP is not intended to work across disjoint addressing or naming realms. </t> <sectiontitle="Locatornumbered="true" toc="default"> <name>Locator IPv6address option">Address Option</name> <t>In fragmentary CDDL, the Locator IPv6addressAddress option follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ ipv6-locator-option = [O_IPv6_LOCATOR, ipv6-address, transport-proto, port-number] ipv6-address = bytes .size 16 transport-proto = IPPROTO_TCP / IPPROTO_UDP IPPROTO_TCP = 6 IPPROTO_UDP = 17 port-number = 0..65535]]></artwork> </figure></t>]]></sourcecode> <t>The content of this option is a binary IPv6 address followed by the protocol number and port number to be used.</t> <t>Note 1: The IPv6 addressMUST<bcp14>MUST</bcp14> normally have global scope. However, during initialization, a link-local addressMAY<bcp14>MAY</bcp14> be used for specific objectives only (<xreftarget="secinst"/>).target="secinst" format="default"/>). In thiscasecase, the corresponding Discovery Response messageMUST<bcp14>MUST</bcp14> be sent via the interface to which the link-local address applies.</t> <t>Note 2: A link-local IPv6 addressMUST NOT<bcp14>MUST NOT</bcp14> be used when this option is included in a Divert option.</t> <t>Note 3: The IPPROTO values are taken from the existing IANA Protocol Numbers registry in order to specify TCP or UDP. If GRASP requires future values that are not in that registry, a new registry for values outside the range 0..255 will be needed.</t> </section> <sectiontitle="Locatornumbered="true" toc="default"> <name>Locator IPv4address option">Address Option</name> <t>In fragmentary CDDL, the Locator IPv4addressAddress option follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ ipv4-locator-option = [O_IPv4_LOCATOR, ipv4-address, transport-proto, port-number] ipv4-address = bytes .size 4]]></artwork> </figure></t>]]></sourcecode> <t>The content of this option is a binary IPv4 address followed by the protocol number and port number to be used.</t> <t>Note: If an operator has internal network address translation for IPv4, this optionMUST NOT<bcp14>MUST NOT</bcp14> be used within the Divert option.</t> </section> <sectiontitle="Locatornumbered="true" toc="default"> <name>Locator FQDNoption">Option</name> <t>In fragmentary CDDL, the Locator FQDN option follows the pattern:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ fqdn-locator-option = [O_FQDN_LOCATOR, text, transport-proto, port-number]]]></artwork> </figure></t>]]></sourcecode> <t>The content of this option is theFully Qualified Domain NameFQDN of the target followed by the protocol number and port number to be used. </t> <t>Note 1: Any FQDNwhichthat might not be valid throughout the network in question, such as a Multicast DNS name <xreftarget="RFC6762"/>, MUST NOTtarget="RFC6762" format="default"/>, <bcp14>MUST NOT</bcp14> be used when this option is used within the Divert option.</t> <t>Note 2: Normal GRASP operations are not expected to use this option. It is intended for special purposes such as discovering external services.</t> </section> <sectiontitle="Locatornumbered="true" toc="default"> <name>Locator URIoption">Option</name> <t>In fragmentary CDDL, the Locator URI option follows the pattern:</t><t><figure> <artwork><![CDATA[ uri-locator<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ uri-locator-option = [O_URI_LOCATOR, text, transport-proto / null, port-number / null]]]></artwork> </figure></t>]]></sourcecode> <t>The content of this option is theUniform Resource IdentifierURI of the target followed by the protocol number and port number to be used (or by null values if not required) <xreftarget="RFC3986"/>.target="RFC3986" format="default"/>. </t> <t>Note 1: Any URI which might not be valid throughout the network in question, such as one based on a Multicast DNS name <xreftarget="RFC6762"/>, MUST NOTtarget="RFC6762" format="default"/>, <bcp14>MUST NOT</bcp14> be used when this option is used within the Divert option.</t> <t>Note 2: Normal GRASP operations are not expected to use this option. It is intended for special purposes such as discovering external services.ThereforeTherefore, its use is not further described in this specification.</t> </section> </section><!----></section> <section anchor="ObjOption"title="Objective Options">numbered="true" toc="default"> <name>Objective Options</name> <section anchor="ObjForm"title="Formatnumbered="true" toc="default"> <name>Format of ObjectiveOptions">Options</name> <t>An objective option is used to identify objectives for the purposes of discovery,negotiationnegotiation, or synchronization. All objectivesMUST<bcp14>MUST</bcp14> be in the following format, described in fragmentary CDDL:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ objective = [objective-name, objective-flags, loop-count, ?objective-value] objective-name = text objective-value = any loop-count = 0..255]]></artwork> </figure></t>]]></sourcecode> <t>All objectives are identified by a unique namewhichthat is a UTF-8 string <xreftarget="RFC3629"/>,target="RFC3629" format="default"/>, to be compared byte by byte. </t> <t>The names of generic objectivesMUST NOT<bcp14>MUST NOT</bcp14> include a colon (":") andMUST<bcp14>MUST</bcp14> be registered with IANA (<xreftarget="iana"/>).</t>target="iana" format="default"/>).</t> <t>The names of privately defined objectivesMUST<bcp14>MUST</bcp14> include at least one colon (":"). The string preceding the last colon in the nameMUST<bcp14>MUST</bcp14> be globally unique and in some way identify the entity or person defining the objective. The following three methodsMAY<bcp14>MAY</bcp14> be used to create such a globally unique string:<list style="numbers"> <t>The</t> <ol spacing="normal" type="1"> <li>The unique string is a decimal number representing a registered32 bit32-bit Private Enterprise Number (PEN) <xreftarget="RFC5612"/>target="RFC5612" format="default"/> that uniquely identifies the enterprise defining theobjective.</t> <t>Theobjective.</li> <li>The unique string is afully qualified domain nameFQDN that uniquely identifies the entity or person defining theobjective.</t> <t>Theobjective.</li> <li>The unique string is an email address that uniquely identifies the entity or person defining theobjective.</t> </list> Theobjective.</li> </ol> <t> GRASPprotocoltreats the objective name as an opaque string. For example, "EX1", "32473:EX1", "example.com:EX1","example.org:EX1"example.org:EX1", and "user@example.org:EX1"would beare five different objectives.</t> <t>The 'objective-flags' field is describedbelow.</t>in <xref target="objective_flags" format="default"/>.</t> <t>The 'loop-count' field is used for terminating negotiation as described in <xreftarget="NegotiationMessage"/>.target="NegotiationMessage" format="default"/>. It is also used for terminating discovery as described in <xreftarget="discmech"/>,target="discmech" format="default"/> and for terminating flooding as described in <xreftarget="flooding"/>.target="flooding" format="default"/>. It is placed in the objective rather than in the GRASP message format because, as far as the ASA is concerned, it is a property of the objective itself. </t> <t> The 'objective-value' fieldis to expressexpresses the actual value of a negotiation or synchronization objective. Its format is defined in the specification of the objective and may be a simple value or a data structure of any kind, as long as it can be represented in CBOR. It is optionalbecause it is optionalonly in a Discovery or Discovery Response message.</t> </section> <sectiontitle="Objective flags">anchor="objective_flags" numbered="true" toc="default"> <name>Objective Flags</name> <t>An objective may be relevant for discovery only, for discovery and negotiation, or for discovery and synchronization. This is expressed in the objective by logical flag bits:</t><t><figure> <artwork><![CDATA[<sourcecode type="cddl" name="grasp-fragments.cddl"><![CDATA[ objective-flags = uint .bits objective-flag objective-flag = &( F_DISC: 0 ; valid for discovery F_NEG: 1 ; valid for negotiation F_SYNCH: 2 ; valid for synchronization F_NEG_DRY: 3 ; negotiation isdry-run ) ]]></artwork> </figure></t>a dry run )]]> </sourcecode> <t>These bits are independent and may be combined appropriately,e.g.e.g., (F_DISC and F_SYNCH) or (F_DISC and F_NEG) or (F_DISC and F_NEG and F_NEG_DRY).</t> <t>Note that for a given negotiation session, an objective must beeitherused either fornegotiation,negotiation or for dry-run negotiation. Mixing the two modes in a single negotiation is not possible.</t> </section> <section anchor="ConsOption"title="Generalnumbered="true" toc="default"> <name>General Considerations for ObjectiveOptions">Options</name> <t>As mentioned above,Objective Options MUSTobjective options <bcp14>MUST</bcp14> be assigned a unique name. As long as privately definedObjective Optionsobjective options obey the rules above, this document does not restrict their choice of name, but the entity or person concernedSHOULD<bcp14>SHOULD</bcp14> publish the names in use. </t> <t>Names are expressed as UTF-8 strings for convenience in designingObjective Optionsobjective options for localized use. For generic usage, names expressed in the ASCII subset of UTF-8 areRECOMMENDED.<bcp14>RECOMMENDED</bcp14>. Designers planning to use non-ASCII names are strongly advised to consult <xreftarget="RFC7564"/>target="RFC8264" format="default"/> or its successor to understand the complexities involved. SincetheGRASPprotocolcompares names byte by byte, all issues of Unicode profiling and canonicalizationMUST<bcp14>MUST</bcp14> be specified in the design of theObjective Option.</t>objective option.</t> <t>AllObjective Options MUSTobjective options <bcp14>MUST</bcp14> respect the CBOR patterns defined above as "objective" andMUST<bcp14>MUST</bcp14> replace the"any"'any' field with a valid CBOR data definition for the relevant use case and application. </t> <t>AnObjective Optionobjective option that contains no additional fields beyond its"loop-count"'loop-count' can only be a discovery objective andMUST<bcp14>MUST</bcp14> only be used in Discovery and Discovery Response messages.</t> <t>The Negotiation ObjectiveOptionsoptions contain negotiation objectives, which vary according to differentfunctions/services.functions and/or services. TheyMUST<bcp14>MUST</bcp14> be carried by Discovery, RequestNegotiationNegotiation, or Negotiation messages only. The negotiation initiatorMUST<bcp14>MUST</bcp14> set the initial"loop-count"'loop-count' to a value specified in the specification of the objective or, if no such value is specified, to GRASP_DEF_LOOPCT.</t> <t>For most scenarios, there should be initial values in the negotiation requests. Consequently, the Negotiation Objective optionsMUST<bcp14>MUST</bcp14> always be completely presented in a Request Negotiation message, or in a Discovery message in rapid mode. If there is no initial value, thevalue'value' fieldSHOULD<bcp14>SHOULD</bcp14> be set to the 'null' value defined by CBOR.</t> <t>Synchronization ObjectiveOptionsoptions are similar, butMUST<bcp14>MUST</bcp14> be carried by Discovery, Discovery Response, Request Synchronization, or Flood Synchronization messages only. They includevalue'value' fields only in Synchronization or Flood Synchronization messages. </t> <t>The design of an objective interacts in various ways with the design of the ASAs that will use it. ASA design considerations are discussed in <xreftarget="I-D.carpenter-anima-asa-guidelines"/>.</t>target="I-D.ietf-anima-asa-guidelines" format="default"/>.</t> </section> <sectiontitle="Organizingnumbered="true" toc="default"> <name>Organizing of ObjectiveOptions">Options</name> <t>Generic objective optionsMUST<bcp14>MUST</bcp14> be specified in documents available to the public andSHOULD<bcp14>SHOULD</bcp14> be designed to use either the negotiation or the synchronization mechanism described above. </t> <t>As noted earlier, one negotiation objective is handled by each GRASP negotiation thread. Therefore, a negotiation objective, which is based on a specific function or action,SHOULD<bcp14>SHOULD</bcp14> be organized as a single GRASP option. It isNOT RECOMMENDED<bcp14>NOT RECOMMENDED</bcp14> to organize multiple negotiation objectives into a singleoption,option nor to split a single function or action into multiple negotiation objectives. </t> <t>It is important to understand that GRASP negotiation does not support transactional integrity. If transactional integrity is needed for a specific objective, this must be ensured by the ASA. For example, an ASA might need to ensure that it only participates in one negotiation thread at the same time. Such an ASA would need to stop listening for incoming negotiation requests before generating an outgoing negotiation request.</t> <t>A synchronization objectiveSHOULD<bcp14>SHOULD</bcp14> be organized as a single GRASP option.</t> <t>Some objectives will support more than one operational mode. An example is a negotiation objective with both a"dry run"dry-run mode (where the negotiation is tofind outdetermine whether the other endcancan, infactfact, make the requested change without problems) and a"live"live mode, as explained in <xreftarget="negproc"/>.target="negproc" format="default"/>. The semantics of such modes will be defined in the specification of the objectives. These objectivesSHOULD<bcp14>SHOULD</bcp14> include flags indicating the applicable mode(s).</t> <t>An issue requiring particular attention is that GRASP itself is not a transactionally safe protocol. Any state associated with adry rundry-run operation, such as temporarily reserving a resource for subsequent use in a live run, is entirely a matter for the designer of the ASA concerned.</t> <t>As indicated in <xreftarget="terms"/>,target="terms" format="default"/>, an objective's value may include multiple parameters. Parameters might be categorized into two classes: the obligatory ones presented as fixedfields;fields and the optional ones presented in some other form of data structure embedded in CBOR. The format might be inherited from an existing management or configuration protocol, with the objective option acting as a carrier for that format. The data structure might be defined in a formal language, but that is a matter for the specifications of individual objectives. There are many candidates, according to the context, such as ABNF, RBNF, XML Schema, YANG, etc.TheGRASPprotocolitself is agnostic on these questions. The only restriction is that the format can be mapped into CBOR.</t> <t>It isNOT RECOMMENDED<bcp14>NOT RECOMMENDED</bcp14> to mix parameters that have significantly differentresponse timeresponse-time characteristics in a single objective. Separate objectives are more suitable for such a scenario.</t> <t>All objectivesMUST<bcp14>MUST</bcp14> support GRASP discovery. However, as mentioned in <xreftarget="highlevel"/>,target="highlevel" format="default"/>, it is acceptable for an ASA to use an alternative method of discovery. </t> <t>Normally, a GRASP objective will refer to specific technical parameters as explained in <xreftarget="terms"/>.target="terms" format="default"/>. However, it is acceptable to define an abstract objective for the purpose of managing or coordinating ASAs. It is also acceptable to define a special-purpose objective for purposes such as trust bootstrapping or formation of the ACP.</t> <t> To guarantee convergence, a limited number of rounds or a timeout is needed for each negotiation objective. Therefore, the definition of each negotiation objectiveSHOULD<bcp14>SHOULD</bcp14> clearly specify this, forexampleexample, a default loop count and timeout, so that the negotiation can always be terminated properly. If not, the GRASP defaults will apply. </t> <t> There must be a well-defined procedure for concluding that a negotiation cannot succeed, and ifsoso, deciding what happens next (e.g., deadlock resolution, tie-breaking, orrevertreversion to best-effort service). ThisMUST<bcp14>MUST</bcp14> be specified for individual negotiation objectives. </t> </section> <sectiontitle="Experimentalnumbered="true" toc="default"> <name>Experimental and Example ObjectiveOptions">Options</name> <t>The names "EX0" through "EX9" have been reserved for experimental options. Multiple names have been assigned because a single experiment may use multiple options simultaneously. These experimental options are highly likely to have different meanings when used for different experiments. Therefore, theySHOULD NOT<bcp14>SHOULD NOT</bcp14> be used without an explicit human decision andMUST NOT<bcp14>MUST NOT</bcp14> be used in unmanaged networks such as home networks.</t> <t>These names are alsoRECOMMENDED<bcp14>RECOMMENDED</bcp14> for use in documentation examples.</t> </section> </section> </section> <sectiontitle="Implementation Status [RFC Editor: please remove]"> <t>Two prototype implementations of GRASP have been made.</t> <section title="BUPT C++ Implementation"> <t><list style="symbols"> <t>Name: BaseNegotiator.cpp, msg.cpp, Client.cpp, Server.cpp</t> <t>Description: C++ implementation of GRASP core and API</t> <t>Maturity: Prototype code, interoperable between Ubuntu.</t> <t>Coverage: Corresponds to draft-carpenter-anima-gdn-protocol-03. Since it was implemented based on the old version draft, the most significant limitations comparing to current protocol design include: <list style="symbols"> <t>Not support CBOR</t> <t>Not support Flooding</t> <t>Not support loop avoidance</t> <t>only coded for IPv6, any IPv4 is accidental</t></list></t> <t>Licensing: Huawei License.</t> <t>Experience: https://github.com/liubingpang/IETF-Anima-Signaling-Protocol/blob/master/README.md</t> <t>Contact: https://github.com/liubingpang/IETF-Anima-Signaling-Protocol</t> </list></t> </section> <section title="Python Implementation"> <t><list style="symbols"> <t>Name: graspy</t> <t>Description: Python 3 implementation of GRASP core and API.</t> <t>Maturity: Prototype code, interoperable between Windows 7 and Linux.</t> <t>Coverage: Corresponds to draft-ietf-anima-grasp-13. Limitations include: <list style="symbols"> <t>insecure: uses a dummy ACP module</t> <t>only coded for IPv6, any IPv4 is accidental</t> <t>FQDN and URI locators incompletely supported</t> <t>no code for rapid mode</t> <t>relay code is lazy (no rate control)</t> <t>all unicast transactions use TCP (no unicast UDP). Experimental code for unicast UDP proved to be complex and brittle.</t> <t>optional Objective option in Response messages not implemented</t> <t>workarounds for defects in Python socket module and Windows socket peculiarities</t> </list></t> <t>Licensing: Simplified BSD</t> <t>Experience: Tested on Windows, Linux and MacOS. https://www.cs.auckland.ac.nz/~brian/graspy/graspy.pdf</t> <t>Contact: https://www.cs.auckland.ac.nz/~brian/graspy/</t> </list></t> </section> </section> <sectionanchor="security"title="Security Considerations">numbered="true" toc="default"> <name>Security Considerations</name> <t>A successful attack on negotiation-enabled nodes would be extremely harmful, as such nodes might end up with a completely undesirable configuration that would also adversely affect their peers. GRASP nodes and messages therefore require full protection. As explained in <xreftarget="reqsec"/>,target="reqsec" format="default"/>, GRASPMUST<bcp14>MUST</bcp14> run within a secure environment such as theAutonomic Control PlaneACP <xreftarget="I-D.ietf-anima-autonomic-control-plane"/>,target="RFC8994" format="default"/>, except for the constrained instances described in <xreftarget="secinst"/>.</t> <t>- Authentication<list style="hanging"> <t>Atarget="secinst" format="default"/>.</t> <dl newline="true" spacing="normal"> <dt>Authentication </dt> <dd><t>A cryptographically authenticated identity for each device is needed in anautonomic network.Autonomic Network. It is not safe to assume that a large network is physically secured against interference or that all personnel are trustworthy. Each autonomic nodeMUST<bcp14>MUST</bcp14> be capable of proving its identity and authenticating its messages. GRASP relies on aseparateseparate, external certificate-based security mechanism to support authentication, data integrity protection, and anti-replay protection.</t> <t>Since GRASP must be deployed in an existing secure environment, the protocol itself specifies nothing concerning the trust anchor and certification authority. For example, in theAutonomic Control PlaneACP <xreftarget="I-D.ietf-anima-autonomic-control-plane"/>,target="RFC8994" format="default"/>, all nodes can trust each other and the ASAs installed in them.</t> <t>If GRASP is used temporarily without an external security mechanism, forexampleexample, during system bootstrap (<xreftarget="reqsec"/>),target="reqsec" format="default"/>), the Session ID (<xreftarget="SessionID"/>)target="SessionID" format="default"/>) will act as a nonce to provide limited protection againstthird partiesthe injectingresponses.of responses by third parties. A full analysis of the secure bootstrap process is in <xreftarget="I-D.ietf-anima-bootstrapping-keyinfra"/>. </t> </list></t> <t>- Authorizationtarget="RFC8995" format="default"/>.</t> </dd> <dt>Authorization andRoles<list style="hanging"> <t>The GRASP protocolroles</dt> <dd><t>GRASP is agnostic about the roles and capabilities of individual ASAs and about which objectives a particular ASA is authorized to support. An implementation might support precautions such as allowing only one ASA in a given node to modify a given objective, but this may not be appropriate in all cases. For example, it might be operationally useful to allow an old and a new version of the same ASA to run simultaneously during an overlap period. These questions are out of scope for the presentspecification.</t> </list></t> <t>- Privacyspecification.</t></dd> <dt>Privacy andconfidentiality<list style="hanging"> <t>GRASPconfidentiality </dt> <dd><t>GRASP is intended fornetwork managementnetwork-management purposes involving network elements, not end hosts. Therefore, no personal information is expected to be involved in the signaling protocol, so there should be no direct impact on personal privacy. Nevertheless, applications that do convey personal information cannot be excluded. Also, traffic flow paths, VPNs,etc.etc., could be negotiated, which could be of interest for traffic analysis. Operators generally want to conceal details of their network topology and traffic density from outsiders. Therefore, since insider attacks cannot be excluded in a large network, the security mechanism for the protocolMUST<bcp14>MUST</bcp14> provide message confidentiality. This is why <xreftarget="reqsec"/>target="reqsec" format="default"/> requires either an ACP or an alternative securitymechanism.</t> </list></t> <t>- Link-localmechanism.</t></dd> <dt>Link-local multicastsecurity<list style="hanging"> <t>GRASPsecurity </dt> <dd><t>GRASP has no reasonable alternative to using link-local multicast for Discovery or Flood Synchronizationmessagesmessages, and these messages are sent in the clear and with no authentication. They are only sent on interfaces within theautonomic networkAutonomic Network (see <xreftarget="terms"/>target="terms" format="default"/> and <xreftarget="reqsec"/>).target="reqsec" format="default"/>). Theyare howeverare, however, available to on-linkeavesdroppers,eavesdroppers and could be forged by on-link attackers. In the case ofDiscovery,discovery, the Discovery Responses are unicast and will therefore be protected (<xreftarget="reqsec"/>),target="reqsec" format="default"/>), and an untrusted forger will not be able to receive responses. In the case ofFlood Synchronization,flood synchronization, an on-link eavesdropper will be able to receive the floodedobjectivesobjectives, but there is no response message to consider. Some precautions for Flood Synchronization messages are suggested in <xreftarget="flooding"/>.</t> </list></t> <t>- DoS Attack Protection<list style="hanging"> <t>GRASPtarget="flooding" format="default"/>.</t></dd> <dt>DoS attack protection </dt> <dd><t>GRASP discovery partly relies on insecure link-local multicast. Since routers participating in GRASP sometimes relaydiscoveryDiscovery messages from one link to another, this could be a vector fordenial of servicedenial-of-service attacks. Some mitigations are specified in <xreftarget="discmech"/>.target="discmech" format="default"/>. However, malicious code installed inside theAutonomic Control PlaneACP could always launch DoS attacks consisting of either spuriousdiscovery messages,Discovery messages orofspuriousdiscovery responses.Discovery Responses. It is important that firewalls prevent any GRASP messages from entering the domain from an unknownsource. </t> </list></t> <t>- Securitysource.</t></dd> <dt>Security during bootstrap anddiscovery<list style="hanging"> <t>Adiscovery </dt> <dd><t>A node cannot trust GRASP traffic from other nodes until the security environment (such as the ACP) has identified the trust anchor and can authenticate traffic by validating certificates for other nodes. Also, until it hassuccesfullysuccessfully enrolled <xreftarget="I-D.ietf-anima-bootstrapping-keyinfra"/>target="RFC8995" format="default"/>, a node cannot assume that other nodes are able to authenticate its own traffic. Therefore, GRASP discovery during the bootstrap phase for a new device will inevitably be insecure. Secure synchronization and negotiation will be impossible until enrollment is complete. Further details are given in <xreftarget="secinst"/>.</t> </list></t> <t>- Securitytarget="secinst" format="default"/>.</t></dd> <dt>Security of discoveredlocators<list style="hanging"> <t>Whenlocators </dt> <dd><t>When GRASP discovery returns an IP address, itMUST<bcp14>MUST</bcp14> be that of a node within the secure environment (<xreftarget="reqsec"/>).target="reqsec" format="default"/>). If it returns an FQDN or a URI, the ASA that receives itMUST NOT<bcp14>MUST NOT</bcp14> assume that the target of the locator is within the secureenvironment.</t> </list></t>environment.</t></dd> </dl> </section> <section anchor="cddl"title="CDDLnumbered="true" toc="default"> <name>CDDL Specification ofGRASP"> <t><figure> <artwork><![CDATA[ <CODE BEGINS>GRASP</name> <sourcecode name="grasp.cddl" type="cddl" markers="true"><![CDATA[ grasp-message = (message .within message-structure) / noop-message message-structure = [MESSAGE_TYPE, session-id, ?initiator, *grasp-option] MESSAGE_TYPE = 0..255 session-id = 0..4294967295;up; up to 32 bits grasp-option = any message /= discovery-message discovery-message = [M_DISCOVERY, session-id, initiator, objective] message /= response-message;response; response to Discovery response-message = [M_RESPONSE, session-id, initiator, ttl, (+locator-option // divert-option), ?objective] message /= synch-message;response; response to Synchronization request synch-message = [M_SYNCH, session-id, objective] message /= flood-message flood-message = [M_FLOOD, session-id, initiator, ttl, +[objective, (locator-option / [])]] message /= request-negotiation-message request-negotiation-message = [M_REQ_NEG, session-id, objective] message /= request-synchronization-message request-synchronization-message = [M_REQ_SYN, session-id, objective] message /= negotiation-message negotiation-message = [M_NEGOTIATE, session-id, objective] message /= end-message end-message = [M_END, session-id, accept-option /decline-option ]decline-option] message /= wait-message wait-message = [M_WAIT, session-id, waiting-time] message /= invalid-message invalid-message = [M_INVALID, session-id, ?any] noop-message = [M_NOOP] divert-option = [O_DIVERT, +locator-option] accept-option = [O_ACCEPT] decline-option = [O_DECLINE, ?reason] reason = text;optional; optional UTF-8 error message waiting-time = 0..4294967295 ; in milliseconds ttl = 0..4294967295 ; in milliseconds locator-option /= [O_IPv4_LOCATOR, ipv4-address, transport-proto, port-number] ipv4-address = bytes .size 4 locator-option /= [O_IPv6_LOCATOR, ipv6-address, transport-proto, port-number] ipv6-address = bytes .size 16 locator-option /= [O_FQDN_LOCATOR, text, transport-proto, port-number] locator-option /= [O_URI_LOCATOR, text, transport-proto / null, port-number / null] transport-proto = IPPROTO_TCP / IPPROTO_UDP IPPROTO_TCP = 6 IPPROTO_UDP = 17 port-number = 0..65535 initiator = ipv4-address / ipv6-address objective-flags = uint .bits objective-flag objective-flag = &( F_DISC: 0 ; valid for discovery F_NEG: 1 ; valid for negotiation F_SYNCH: 2 ; valid for synchronization F_NEG_DRY: 3 ; negotiation isdry-run )a dry run ) objective = [objective-name, objective-flags, loop-count, ?objective-value] objective-name = text;see; see section "Format of Objective Options" objective-value = any loop-count = 0..255 ; Constants for message types and option types M_NOOP = 0 M_DISCOVERY = 1 M_RESPONSE = 2 M_REQ_NEG = 3 M_REQ_SYN = 4 M_NEGOTIATE = 5 M_END = 6 M_WAIT = 7 M_SYNCH = 8 M_FLOOD = 9 M_INVALID = 99 O_DIVERT = 100 O_ACCEPT = 101 O_DECLINE = 102 O_IPv6_LOCATOR = 103 O_IPv4_LOCATOR = 104 O_FQDN_LOCATOR = 105 O_URI_LOCATOR = 106<CODE ENDS> ]]></artwork> </figure></t>]]></sourcecode> </section> <section anchor="iana"title="IANA Considerations">numbered="true" toc="default"> <name>IANA Considerations</name> <t>This document defines the GeneRic Autonomic Signaling Protocol (GRASP).</t> <t><xreftarget="Constants"/>target="Constants" format="default"/> explains the following link-local multicastaddresses, whichaddresses that IANAis requested to assignhas assigned for use byGRASP:</t> <t><list style="hanging"> <t hangText="ALL_GRASP_NEIGHBORS multicast address">(IPv6): (TBD1). AssignedGRASP.</t> <t>Assigned in theIPv6 Link-Local"Link-Local Scope MulticastAddresses registry.</t> <t hangText="ALL_GRASP_NEIGHBORS multicast address">(IPv4): (TBD2). Assigned inAddresses" subregistry of theIPv4"IPv6 MulticastLocalAddress Space Registry":</t> <dl newline="false" spacing="compact"> <dt>Address(es):</dt><dd>ff02::13</dd> <dt>Description:</dt><dd>ALL_GRASP_NEIGHBORS</dd> <dt>Reference:</dt><dd>RFC 8990</dd> </dl> <t>Assigned in the "Local Network ControlBlock. <!-- <vspace blankLines="1"/> (Note in draft: alternatively, we could use 224.0.0.1, currently defined as All Systems on this Subnet.)--></t> </list></t>Block (224.0.0.0 - 224.0.0.255 (224.0.0/24))" subregistry of the "IPv4 Multicast Address Space Registry":</t> <dl newline="false" spacing="compact"> <dt>Address(es):</dt><dd>224.0.0.119</dd> <dt>Description:</dt><dd>ALL_GRASP_NEIGHBORS</dd> <dt>Reference:</dt><dd>RFC 8990</dd> </dl> <t><xreftarget="Constants"/>target="Constants" format="default"/> explains the following UserPort,Port (GRASP_LISTEN_PORT), which IANAis requested to assignhas assigned for use by GRASP for both UDP and TCP:</t><t>GRASP_LISTEN_PORT: (TBD3) <vspace blankLines="0"/> Service Name: Generic<dl spacing="compact"> <dt>Service Name:</dt> <dd>grasp</dd> <dt>Port Number:</dt> <dd>7017</dd> <dt>Transport Protocol:</dt> <dd>udp, tcp</dd> <dt>Description</dt><dd>GeneRic Autonomic SignalingProtocol (GRASP) <vspace blankLines="0"/> Transport Protocols: UDP, TCP <vspace blankLines="0"/> Assignee: iesg@ietf.org <vspace blankLines="0"/> Contact: chair@ietf.org <vspace blankLines="0"/> Description: See <xref target="Constants"/> <vspace blankLines="0"/> Reference: RFC XXXX (this document)</t>Protocol</dd> <dt>Assignee:</dt> <dd>IESG <iesg@ietf.org></dd> <dt>Contact:</dt> <dd>IETF Chair <chair@ietf.org></dd> <dt>Reference:</dt> <dd>RFC 8990</dd> </dl> <t>The IANAis requested to create a GRASP Parameter Registry including two registry tables. These arehas created theGRASP"GeneRic Autonomic Signaling Protocol (GRASP) Parameters" registry, which includes two subregistries: "GRASP Messages andOptions TableOptions" andthe GRASP"GRASP ObjectiveNames Table.</t> <t>GRASP Messages and Options Table. TheNames".</t> <t>The values inthis tablethe "GRASP Messages and Options" subregistry are names paired with decimal integers. Future valuesMUST<bcp14>MUST</bcp14> be assigned using the Standards Action policy defined by <xreftarget="RFC8126"/>.target="RFC8126" format="default"/>. The following initial values are assigned by this document:</t><t><figure> <artwork><![CDATA[M_NOOP = 0 M_DISCOVERY = 1 M_RESPONSE = 2 M_REQ_NEG = 3 M_REQ_SYN = 4 M_NEGOTIATE = 5 M_END = 6 M_WAIT = 7 M_SYNCH = 8 M_FLOOD = 9 M_INVALID = 99 O_DIVERT = 100 O_ACCEPT = 101 O_DECLINE = 102 O_IPv6_LOCATOR = 103 O_IPv4_LOCATOR = 104 O_FQDN_LOCATOR = 105 O_URI_LOCATOR = 106 ]]></artwork> </figure> </t> <t>GRASP Objective Names Table. The<table anchor="msg-options"> <name>Initial Values of the "GRASP Messages and Options" Subregistry</name> <thead> <tr><th>Value</th><th>Message/Option</th></tr> </thead> <tbody> <tr> <td>0</td> <td>M_NOOP</td> </tr> <tr> <td>1</td> <td>M_DISCOVERY</td> </tr> <tr> <td>2</td> <td>M_RESPONSE</td> </tr> <tr> <td>3</td> <td>M_REQ_NEG</td> </tr> <tr> <td>4</td> <td>M_REQ_SYN</td> </tr> <tr> <td>5</td> <td>M_NEGOTIATE</td> </tr> <tr> <td>6</td> <td>M_END</td> </tr> <tr> <td>7</td> <td>M_WAIT</td> </tr> <tr> <td>8</td> <td>M_SYNCH</td> </tr> <tr> <td>9</td> <td>M_FLOOD</td> </tr> <tr> <td>99</td> <td>M_INVALID</td> </tr> <tr> <td>100</td> <td>O_DIVERT</td> </tr> <tr> <td>101</td> <td>O_ACCEPT</td> </tr> <tr> <td>102</td> <td>O_DECLINE</td> </tr> <tr> <td>103</td> <td>O_IPv6_LOCATOR</td> </tr> <tr> <td>104</td> <td>O_IPv4_LOCATOR</td> </tr> <tr> <td>105</td> <td>O_FQDN_LOCATOR</td> </tr> <tr> <td>106</td> <td>O_URI_LOCATOR</td> </tr> </tbody> </table> <t>The values inthis tablethe "GRASP Objective Names" subregistry are UTF-8 stringswhich MUST NOTthat <bcp14>MUST NOT</bcp14> include a colon (":"), according to <xreftarget="ObjForm"/>.target="ObjForm" format="default"/>. Future valuesMUST<bcp14>MUST</bcp14> be assigned using the Specification Required policy defined by <xreftarget="RFC8126"/>.</t>target="RFC8126" format="default"/>.</t> <t>To assist expert review of a new objective, the specification should include a precise description of the format of the new objective, with sufficientexplanation of its semantics to allow independent implementations. See <xref target="ConsOption"/> for more details. If the new objective is similar in name or purpose to a previously registered objective, the specification should explain why a new objective is justified. </t> <t>The following initial values are assigned by this document:</t> <t><figure> <artwork><![CDATA[ EX0 EX1 EX2 EX3 EX4 EX5 EX6 EX7 EX8 EX9 ]]></artwork> </figure> </t> </section> <section anchor="ack" title="Acknowledgements"> <t>A major contribution to the original version of this document was made by Sheng Jiang and significant contributions were made by Toerless Eckert. Significant early review inputs were received from Joel Halpern, Barry Leiba, Charles E. Perkins, and Michael Richardson. William Atwood provided important assistance in debugging a prototype implementation.</t> <t>Valuable comments were received from Michael Behringer, Jeferson Campos Nobre, Laurent Ciavaglia, Zongpeng Du, Yu Fu, Joel Jaeggli, Zhenbin Li, Dimitri Papadimitriou, Pierre Peloso, Reshad Rahman, Markus Stenberg, Martin Stiemerling, Rene Struik, Martin Thomson, Dacheng Zhang, and participants in the NMRG research group, the ANIMA working group, and the IESG.</t> </section> </middle> <back> <references title="Normative References"> <?rfc include='reference.RFC.2119'?> <!-- <?rfc include='reference.RFC.5280'?> --> <?rfc include='reference.RFC.4086'?> <!-- <?rfc include='reference.RFC.5246'?> --> <!-- <?rfc include='reference.RFC.6347'?> --> <?rfc include='reference.RFC.3986'?> <?rfc include='reference.RFC.7049'?> <?rfc include='reference.RFC.7217'?> <?rfc include='reference.RFC.3629'?> <?rfc include='reference.RFC.8085'?> <?rfc include='reference.I-D.ietf-anima-autonomic-control-plane'?> <?rfc include='reference.I-D.greevenbosch-appsawg-cbor-cddl'?> </references> <references title="Informative References"> <?rfc include='reference.RFC.2334'?> <?rfc include='reference.RFC.3493'?> <?rfc include='reference.RFC.8126'?> <?rfc include='reference.RFC.6733'?> <?rfc include='reference.RFC.2865'?> <?rfc include='reference.RFC.4861'?> <?rfc include='reference.RFC.5971'?> <?rfc include='reference.RFC.6241'?> <!-- <?rfc include='reference.RFC.3209'?> --> <?rfc include='reference.RFC.2205'?> <?rfc include='reference.RFC.3416'?> <?rfc include='reference.RFC.3315'?> <?rfc include='reference.RFC.6887'?> <?rfc include='reference.RFC.6762'?> <?rfc include='reference.RFC.6763'?> <?rfc include='reference.RFC.2608'?> <?rfc include='reference.RFC.6206'?> <?rfc include='reference.RFC.7564'?> <?rfc include='reference.RFC.7575'?> <?rfc include='reference.RFC.7576'?> <?rfc include='reference.RFC.7558'?> <?rfc include='reference.RFC.7787'?> <?rfc include='reference.RFC.7788'?> <?rfc include='reference.RFC.8040'?> <?rfc include='reference.I-D.liu-anima-grasp-api'?> <?rfc include='reference.I-D.stenberg-anima-adncp'?> <?rfc include='reference.I-D.chaparadza-intarea-igcp'?> <?rfc include='reference.I-D.ietf-anima-reference-model'?> <?rfc include='reference.I-D.ietf-anima-bootstrapping-keyinfra'?> <?rfc include='reference.I-D.ietf-anima-stable-connectivity'?> <?rfc include='reference.RFC.5612'?> <?rfc include='reference.I-D.carpenter-anima-asa-guidelines'?> </references> <section title="Open Issues [RFC Editor: This section should be empty. Please remove]"> <t><list style="symbols"> <t>68. (Placeholder)</t> </list></t> </section> <section title="Closed Issues [RFC Editor: Please remove]"> <t> <list style="symbols"> <t>1. UDP vs TCP: For now, this specification suggests UDP and TCP as message transport mechanisms. This is not clarified yet. UDP is good for short conversations, is necessary for multicast discovery, and generally fits the discovery and divert scenarios well. However, it will cause problems with large messages. TCP is good for stable and long sessions, with a little bit of time consumption during the session establishment stage. If messages exceed a reasonable MTU, a TCP mode will be required in any case. This question may be affected by the security discussion. <vspace blankLines="1"/> RESOLVED by specifying UDP for short message and TCP for longer one. </t> <t>2. DTLS or TLS vs built-in security mechanism. For now, this specification has chosen a PKI based built-in security mechanism based on asymmetric cryptography. However, (D)TLS might be chosen as security solution to avoid duplication of effort. It also allows essentially similar security for short messages over UDP and longer ones over TCP. The implementation trade-offs are different. The current approach requires expensive asymmetric cryptographic calculations for every message. (D)TLS has startup overheads but cheaper crypto per message. DTLS is less mature than TLS. <vspace blankLines="1"/> RESOLVED by specifying external security (ACP or (D)TLS). </t> <t>The following open issues applied only if the original security model was retained: <list style="symbols"> <t>2.1. For replay protection, GRASP currently requires every participant to have an NTP-synchronized clock. Is this OK for low-end devices, and how does it work during device bootstrapping? We could take the Timestamp out of signature option, to become an independent and OPTIONAL (or RECOMMENDED) option.</t> <t>2.2. The Signature Option states that this option could be any place in a message. Wouldn't it be better to specify a position (such as the end)? That would be much simpler to implement. </t> </list>RESOLVED by changing security model.</t> <t>3. DoS Attack Protection needs work. <vspace blankLines="1"/> RESOLVED by adding text.</t> <t>4. Should we consider preferring a text-based approach to discovery (after the initial discovery needed for bootstrapping)? This could be a complementary mechanism for multicast based discovery, especially for a very large autonomic network. Centralized registration could be automatically deployed incrementally. At the very first stage, the repository could be empty; then it could be filled in by the objectives discovered by different devices (for example using Dynamic DNS Update). The more records are stored in the repository, the less the multicast-based discovery is needed. However, if we adopt such a mechanism, there would be challenges: stateful solution, and security. <vspace blankLines="1"/> RESOLVED for now by adding optional use of DNS-SD by ASAs. Subsequently removed by editors as irrelevant to GRASP istelf. </t> <t>5. Need to expand description of the minimum requirements for the specification of an individual discovery, synchronization or negotiation objective. <vspace blankLines="1"/> RESOLVED for now by extra wording.</t> <t>6. Use case and protocol walkthrough. A description of how a node starts up, performs discovery, and conducts negotiation and synchronisation for a sample use case would help readers to understand the applicability of this specification. Maybe it should be an artificial use case or maybe a simple real one, based on a conceptual API. However, the authors have not yet decided whether to have a separate document or have it in the protocol document. <vspace blankLines="1"/> RESOLVED: recommend a separate document.</t> <t>7. Cross-check against other ANIMA WG documents for consistency and gaps. <vspace blankLines="1"/> RESOLVED: Satisfied by WGLC.</t> <t>8. Consideration of ADNCP proposal. <vspace blankLines="1"/> RESOLVED by adding optional use of DNCP for flooding-type synchronization.</t> <t>9. Clarify how a GDNP instance knows whether it is running inside the ACP. (Sheng) <vspace blankLines="1"/> RESOLVED by improved text.</t> <t>10. Clarify how a non-ACP GDNP instance initiates (D)TLS. (Sheng) <vspace blankLines="1"/> RESOLVED by improved text and declaring DTLS out of scope for this draft. </t> <t>11. Clarify how UDP/TCP choice is made. (Sheng) [Like DNS? - Brian] <vspace blankLines="1"/> RESOLVED by improved text.</t> <t>12. Justify that IP address within ACP or (D)TLS environment is sufficient to prove AN identity; or explain how Device Identity Option is used. (Sheng) <vspace blankLines="1"/> RESOLVED for now: we assume that all ASAs in a device are trusted as soon as the device is trusted, so they share credentials. In that case the Device Identity Option is useless. This needs to be reviewed later.</t> <t>13. Emphasise that negotiation/synchronization are independent from discovery, although the rapid discovery mode includes the first step of a negotiation/synchronization. (Sheng) <vspace blankLines="1"/> RESOLVED by improved text. </t> <t>14. Do we need an unsolicited flooding mechanism for discovery (for discovery results that everyone needs), to reduce scaling impact of flooding discovery messages? (Toerless) <vspace blankLines="1"/> RESOLVED: Yes, added to requirements and solution. </t> <t>15. Do we need flag bits in Objective Options to distinguish distinguish Synchronization and Negotiation "Request" or rapid mode "Discovery" messages? (Bing) <vspace blankLines="1"/> RESOLVED: yes, work on the API showed that these flags are essential. </t> <t>16. (Related to issue 14). Should we revive the "unsolicited Response" for flooding synchronisation data? This has to be done carefully due to the well-known issues with flooding, but it could be useful, e.g. for Intent distribution, where DNCP doesn't seem applicable. <vspace blankLines="1"/> RESOLVED: Yes, see #14. </t> <t>17. Ensure that the discovery mechanism is completely proof against loops and protected against duplicate responses. <vspace blankLines="1"/> RESOLVED: Added loop count mechanism. </t> <t>18. Discuss the handling of multiple valid discovery responses. <vspace blankLines="1"/> RESOLVED: Stated that the choice must be available to the ASA but GRASP implementation should pick a default. </t> <t>19. Should we use a text-oriented format such as JSON/CBOR instead of native binary TLV format? <vspace blankLines="1"/> RESOLVED: Yes, changed to CBOR. </t> <t>20. Is the Divert option needed? If a discovery response provides a valid IP address or FQDN, the recipient doesn't gain any extra knowledge from the Divert. On the other hand, the presence of Divert informs the receiver that the target is off-link, which might be useful sometimes. <vspace blankLines="1"/> RESOLVED: Decided to keep Divert option. </t> <t>21. Rename the protocol as GRASP (GeneRic Autonomic Signaling Protocol)? <vspace blankLines="1"/> RESOLVED: Yes, name changed.</t> <t>22. Does discovery mechanism scale robustly as needed? Need hop limit on relaying? <vspace blankLines="1"/> RESOLVED: Added hop limit.</t> <t>23. Need more details on TTL for caching discovery responses. <vspace blankLines="1"/> RESOLVED: Done.</t> <t>24. Do we need "fast withdrawal" of discovery responses? <vspace blankLines="1"/> RESOLVED: This doesn't seem necessary. If an ASA exits or stops supporting a given objective, peers will fail to start future sessions and will simply repeat discovery. </t> <t>25. Does GDNP discovery meet the needs of multi-hop DNS-SD? <vspace blankLines="1"/> RESOLVED: Decided not to consider this further as a GRASP protocol issue. GRASP objectives could embed DNS-SD formats if needed.</t> <t>26. Add a URL type to the locator options (for security bootstrap etc.) <vspace blankLines="1"/> RESOLVED: Done, later renamed as URI. </t> <t>27. Security of Flood multicasts (<xref target="flooding"/>). <vspace blankLines="1"/> RESOLVED: added text.</t> <t>28. Does ACP support secure link-local multicast? <vspace blankLines="1"/> RESOLVED by new text in the Security Considerations.</t> <t>29. PEN is used to distinguish vendor options. Would it be better to use a domain name? Anything unique will do. <vspace blankLines="1"/> RESOLVED: Simplified this by removing PEN field and changing naming rules for objectives.</t> <t>30. Does response to discovery require randomized delays to mitigate amplification attacks? <vspace blankLines="1"/> RESOLVED: WG feedback is that it's unnecessary.</t> <t>31. We have specified repeats for failed discovery etc. Is that sufficient to deal with sleeping nodes? <vspace blankLines="1"/> RESOLVED: WG feedback is that it's unnecessary to say more.</t> <t>32. We have one-to-one synchronization and flooding synchronization. Do we also need selective flooding to a subset of nodes? <vspace blankLines="1"/> RESOLVED: This will be discussed as a protocol extension in a separate draft (draft-liu-anima-grasp-distribution).</t> <t>33. Clarify if/when discovery needs to be repeated. <vspace blankLines="1"/> RESOLVED: Done.</t> <t>34. Clarify what is mandatory for running in ACP, expand discussion of security boundary when running with no ACP - might rely on the local PKI infrastructure. <vspace blankLines="1"/> RESOLVED: Done.</t> <t>35. State that role-based authorization of ASAs is out of scope for GRASP. GRASP doesn't recognize/handle any "roles". <vspace blankLines="1"/> RESOLVED: Done.</t> <t>36. Reconsider CBOR definition for PEN syntax. ( objective-name = text / [pen, text] ; pen = uint ) <vspace blankLines="1"/> RESOLVED: See issue 29.</t> <t>37. Are URI locators really needed? <vspace blankLines="1"/> RESOLVED: Yes, e.g. for security bootstrap discovery, but added note that addresses are the normal case (same for FQDN locators).</t> <t>38. Is Session ID sufficient to identify relayed responses? Isn't the originator's address needed too? <vspace blankLines="1"/> RESOLVED: Yes, this is needed for multicast messages and their responses.</t> <t>39. Clarify that a node will contain one GRASP instance supporting multiple ASAs. <vspace blankLines="1"/> RESOLVED: Done.</t> <t>40. Add a "reason" code to the DECLINE option? <vspace blankLines="1"/> RESOLVED: Done.</t> <t>41. What happens if an ASA cannot conveniently use one of the GRASP mechanisms? Do we (a) add a message type to GRASP, or (b) simply pass the discovery results to the ASA so that it can open its own socket?<vspace blankLines="1"/> RESOLVED: Both would be possible, but (b) is preferred.</t> <t>42. Do we need a feature whereby an ASA can bypass the ACP and use the data plane for efficiency/throughput? This would require discovery to return non-ACP addresses and would evade ACP security.<vspace blankLines="1"/> RESOLVED: This is considered out of scope for GRASP, but a comment has been added in security considerations. </t> <t>43. Rapid mode synchronization and negotiation is currently limited to a single objective for simplicity of design and implementation. A future consideration is to allow multiple objectives in rapid mode for greater efficiency. <vspace blankLines="1"/> RESOLVED: This is considered out of scope for this version.</t> <t>44. In requirement T9, the words that encryption "may not be required in all deployments" were removed. Is that OK?.<vspace blankLines="1"/> RESOLVED: No objections.</t> <t>45. Device Identity Option is unused. Can we remove it completely?.<vspace blankLines="1"/> RESOLVED: No objections. Done.</t> <t>46. The 'initiator' field in DISCOVER, RESPONSE and FLOOD messages is intended to assist in loop prevention. However, we also have the loop count for that. Also, if we create a new Session ID each time a DISCOVER or FLOOD is relayed, that ID can be disambiguated by recipients. It would be simpler to remove the initiator from the messages, making parsing more uniform. Is that OK?<vspace blankLines="1"/> RESOLVED: Yes. Done.</t> <t>47. REQUEST is a dual purpose message (request negotiation or request synchronization). Would it be better to split this into two different messages (and adjust various message names accordingly)?<vspace blankLines="1"/> RESOLVED: Yes. Done.</t> <t>48. Should the Appendix "Capability Analysis of Current Protocols" be deleted before RFC publication?<vspace blankLines="1"/> RESOLVED: No (per WG meeting at IETF 96).</t> <t>49. <xref target="reqsec"/> Should say more about signaling between two autonomic networks/domains. <vspace blankLines="1"/> RESOLVED: Description of separate GRASP instance added.</t> <t>50. Is Rapid mode limited to on-link only? What happens if first discovery responder does not support Rapid Mode? <xref target="negproc"/>, <xref target="synchproc"/>) <vspace blankLines="1"/> RESOLVED: Not limited to on-link. First responder wins.</t> <t>51. Should flooded objectives have a time-to-live before they are deleted from the flood cache? And should they be tagged in the cache with their source locator? <vspace blankLines="1"/> RESOLVED: TTL added to Flood (and Discovery Response) messages. Cached flooded objectives must be tagged with their originating ASA locator, and multiple copies must be kept if necessary.</t> <t>52. Describe in detail what is allowed and disallowed in an insecure instance of GRASP. <vspace blankLines="1"/> RESOLVED: Done.</t> <t>53. Tune IANA Considerations to support early assignment request.<vspace blankLines="1"/></t> <t>54. Is there a highly unlikely race condition if two peers simultaneously choose the same Session ID and send each other simultaneous M_REQ_NEG messages? <vspace blankLines="1"/> RESOLVED: Yes. Enhanced text on Session ID generation, and added precaution when receiving a Request message.</t> <t>55. Could discovery be performed over TCP?<vspace blankLines="1"/> RESOLVED: Unicast discovery added as an option.</t> <t>56. Change Session-ID to 32 bits?<vspace blankLines="1"/> RESOLVED: Done.</t> <t>57. Add M_INVALID message?<vspace blankLines="1"/> RESOLVED: Done.</t> <t>58. Maximum message size? <vspace blankLines="1"/> RESOLVED by specifying default maximum message size (2048 bytes).</t> <t>59. Add F_NEG_DRY flag to specify a "dry run" objective?.<vspace blankLines="1"/> RESOLVED: Done.</t> <t>60. Change M_FLOOD syntax to associate a locator with each objective?<vspace blankLines="1"/> RESOLVED: Done.</t> <t>61. Is the SONN constrained instance really needed?<vspace blankLines="1"/> RESOLVED: Retained but only as an option.</t> <t>62. Is it helpful to tag descriptive text with message names (M_DISCOVER etc.)?<vspace blankLines="1"/> RESOLVED: Yes, done in various parts of the text.</t> <t>63. Should encryption be MUST instead of SHOULD in <xref target="reqsec"/> and <xref target="reqsec"/>? <vspace blankLines="1"/> RESOLVED: Yes, MUST implement in both cases.</t> <t>64. Should more security text be moved from the main text into the Security Considerations? <vspace blankLines="1"/> RESOLVED: No, on AD advice.</t> <t>65. Do we need to formally restrict Unicode characters allowed in objective names?<vspace blankLines="1"/> RESOLVED: No, but need to point to guidance from PRECIS WG.</t> <t>66. Split requirements into separate document?<vspace blankLines="1"/> RESOLVED: No, on AD advice.</t> <t>67. Remove normative dependency on draft-greevenbosch-appsawg-cbor-cddl?<vspace blankLines="1"/> RESOLVED: No, on AD advice. In worst case, fix at AUTH48.</t> </list></t> </section> <section anchor="changes" title="Change log [RFC Editor: Please remove]"> <t>draft-ietf-anima-grasp-15, 2017-07-07: <vspace blankLines="1"/> Updates following additional IESG comments: <vspace blankLines="1"/> Security (Eric Rescorla): missing brittleness of group security concept, attack via compromised member. <vspace blankLines="1"/> TSV (Mirja Kuehlewind): clarification on the use of UDP, TCP, mandate use of TCP (or other reliable transport). <vspace blankLines="1"/> Clarified that in ACP, UDP is not used at all. <vspace blankLines="1"/> Clarified that GRASP itself needs TCP listen port (was previously written as if this was optional). </t> <t>draft-ietf-anima-grasp-14, 2017-07-02: <vspace blankLines="1"/> Updates following additional IESG comments: <vspace blankLines="1"/> Updated 2.5.1 and 2.5.2 based on IESG security feedback (specify dependency against security substrate). <vspace blankLines="1"/> Strengthened requirement for reliable transport protocol. </t> <t>draft-ietf-anima-grasp-13, 2017-06-06: <vspace blankLines="1"/> Updates following additional IESG comments: <vspace blankLines="1"/> Removed all mention of TLS, including SONN, since it was under-specified. <vspace blankLines="1"/> Clarified other text about trust and security model. <vspace blankLines="1"/> Banned Rapid Mode when multicast is insecure. <vspace blankLines="1"/> Explained use of M_INVALID to support extensibility <vspace blankLines="1"/> Corrected details on discovery cache TTL and discovery timeout. <vspace blankLines="1"/> Improved description of multicast UDP w.r.t. RFC8085. <vspace blankLines="1"/> Clarified when transport connections are opened or closed. <vspace blankLines="1"/> Noted that IPPROTO values come from the Protocol Numbers registry <vspace blankLines="1"/> Protocol change: Added protocol and port numbers to URI locator. <vspace blankLines="1"/> Removed inaccurate text about routing protocols <vspace blankLines="1"/> Moved Requirements section to an Appendix. <vspace blankLines="1"/> Other editorial and technical clarifications. </t> <t>draft-ietf-anima-grasp-12, 2017-05-19: <vspace blankLines="1"/> Updates following IESG comments: <vspace blankLines="1"/> Clarified that GRASP runs in a single addressing realm <vspace blankLines="1"/> Improved wording about FQDN resolution, clarified that URI usage is out of scope. <vspace blankLines="1"/> Clarified description of negotiation timeout. <vspace blankLines="1"/> Noted that 'dry run' semantics are ASA-dependent <vspace blankLines="1"/> Made the ACP a normative reference <vspace blankLines="1"/> Clarified that LL multicasts are limited to GRASP interfaces <vspace blankLines="1"/> Unicast UDP moved out of scope <vspace blankLines="1"/> Editorial clarifications </t> <t>draft-ietf-anima-grasp-11, 2017-03-30: <vspace blankLines="1"/> Updates following IETF 98 discussion: <vspace blankLines="1"/> Encryption changed to a MUST implement. <vspace blankLines="1"/> Pointed to guidance on UTF-8 names. </t> <t>draft-ietf-anima-grasp-10, 2017-03-10: <vspace blankLines="1"/> Updates following IETF Last call: <vspace blankLines="1"/> Protocol change: Specify that an objective with no initial value should have its value field set to CBOR 'null'. <vspace blankLines="1"/> Protocol change: Specify behavior on receiving unrecognized message type. <vspace blankLines="1"/> Noted that UTF-8 names are matched byte-for-byte. <vspace blankLines="1"/> Added brief guidance for Expert Reviewer of new generic objectives. <vspace blankLines="1"/> Numerous editorial improvements and clarifications and minor text rearrangements, none intended to change the meaning. </t> <t>draft-ietf-anima-grasp-09, 2016-12-15: <vspace blankLines="1"/> Protocol change: Add F_NEG_DRY flag to specify a "dry run" objective. <vspace blankLines="1"/> Protocol change: Change M_FLOOD syntax to associate a locator with each objective. <vspace blankLines="1"/> Concentrated mentions of TLS in one section, with all details out of scope. <vspace blankLines="1"/> Clarified text around constrained instances of GRASP. <vspace blankLines="1"/> Strengthened text restricting LL addresses in locator options. <vspace blankLines="1"/> Clarified description of rapid mode processsing. <vspace blankLines="1"/> Specified that cached discovery results should not be returned on the same interface where they were learned. <vspace blankLines="1"/> Shortened text in "High Level Design Choices" <vspace blankLines="1"/> Dropped the word 'kernel' to avoid confusion with o/s kernel mode. <vspace blankLines="1"/> Editorial improvements and clarifications. </t> <t>draft-ietf-anima-grasp-08, 2016-10-30: <vspace blankLines="1"/> Protocol change: Added M_INVALID message. <vspace blankLines="1"/> Protocol change: Increased Session ID space to 32 bits. <vspace blankLines="1"/> Enhanced rules to avoid Session ID clashes. <vspace blankLines="1"/> Corrected and completed description of timeouts for Request messages. <vspace blankLines="1"/> Improved wording about exponential backoff and DoS. <vspace blankLines="1"/> Clarified that discovery relaying is not done by limited security instances. <vspace blankLines="1"/> Corrected and expanded explanation of port used for Discovery Response. <vspace blankLines="1"/> Noted that Discovery message could be sent unicast in special cases. <vspace blankLines="1"/> Added paragraph on extensibility. <vspace blankLines="1"/> Specified default maximum message size. <vspace blankLines="1"/> Added Appendix for sample messages. <vspace blankLines="1"/> Added short protocol overview. <vspace blankLines="1"/> Editorial fixes, including minor re-ordering for readability. </t> <t>draft-ietf-anima-grasp-07, 2016-09-13: <vspace blankLines="1"/> Protocol change: Added TTL field to Flood message (issue 51). <vspace blankLines="1"/> Protocol change: Added Locator option to Flood message (issue 51). <vspace blankLines="1"/> Protocol change: Added TTL field to Discovery Response message (corrollary to issue 51). <vspace blankLines="1"/> Clarified details of rapid mode (issues 43 and 50). <vspace blankLines="1"/> Description of inter-domain GRASP instance added (issue 49). <vspace blankLines="1"/> Description of limited security GRASP instances added (issue 52). <vspace blankLines="1"/> Strengthened advice to use TCP rather than UDP. <vspace blankLines="1"/> Updated IANA considerations and text about well-known port usage (issue 53). <vspace blankLines="1"/> Amended text about ASA authorization and roles to allow for overlapping ASAs. <vspace blankLines="1"/> Added text recommending that Flood should be repeated periodically. <vspace blankLines="1"/> Editorial fixes. </t> <t>draft-ietf-anima-grasp-06, 2016-06-27: <vspace blankLines="1"/> Added text on discovery cache timeouts. <vspace blankLines="1"/> Noted that ASAs that are only initiators do not need to respond to discovery message. <vspace blankLines="1"/> Added text on unexpected address changes. <vspace blankLines="1"/> Added text on robust implementation. <vspace blankLines="1"/> Clarifications and editorial fixes for numerous review comments <vspace blankLines="1"/> Added open issues for some review comments. </t> <t>draft-ietf-anima-grasp-05, 2016-05-13: <vspace blankLines="1"/> Noted in requirement T1 that it should be possible to implement ASAs independently as user space programs. <vspace blankLines="1"/> Protocol change: Added protocol number and port to discovery response. Updated protocol description, CDDL and IANA considerations accordingly. <vspace blankLines="1"/> Clarified that discovery and flood multicasts are handled by the GRASP core, not directly by ASAs. <vspace blankLines="1"/> Clarified that a node may discover an objective without supporting it for synchronization or negotiation. <vspace blankLines="1"/> Added Implementation Status section. <vspace blankLines="1"/> Added reference to SCSP. <vspace blankLines="1"/> Editorial fixes. </t> <t>draft-ietf-anima-grasp-04, 2016-03-11: <vspace blankLines="1"/> Protocol change: Restored initiator field in certain messages and adjusted relaying rules to provide complete loop detection. <vspace blankLines="1"/> Updated IANA Considerations. </t> <t>draft-ietf-anima-grasp-03, 2016-02-24: <vspace blankLines="1"/> Protocol change: Removed initiator field from certain messages and adjusted relaying requirement to simplify loop detection. Also clarified narrative explanation of discovery relaying. <vspace blankLines="1"/> Protocol change: Split Request message into two (Request Negotiation and Request Synchronization) and updated other message names for clarity. <vspace blankLines="1"/> Protocol change: Dropped unused Device ID option. <vspace blankLines="1"/> Further clarified text on transport layer usage. <vspace blankLines="1"/> New text about multicast insecurity in Security Considerations. <vspace blankLines="1"/> Various other clarifications and editorial fixes, including moving some material to Appendix. </t> <t>draft-ietf-anima-grasp-02, 2016-01-13: <vspace blankLines="1"/> Resolved numerous issues according to WG discussions. <vspace blankLines="1"/> Renumbered requirements, added D9. <vspace blankLines="1"/> Protocol change: only allow one objective in rapid mode. <vspace blankLines="1"/> Protocol change: added optional error string to DECLINE option. <vspace blankLines="1"/> Protocol change: removed statement that seemed to say that a Request not preceded by a Discovery should cause a Discovery response. That made no sense, because there is no way the initiator would know where to send the Request. <vspace blankLines="1"/> Protocol change: Removed PEN option from vendor objectives, changed naming rule accordingly. <vspace blankLines="1"/> Protocol change: Added FLOOD message to simplify coding. <vspace blankLines="1"/> Protocol change: Added SYNCH message to simplify coding. <vspace blankLines="1"/> Protocol change: Added initiator id to DISCOVER, RESPONSE and FLOOD messages. But also allowed the relay process for DISCOVER and FLOOD to regenerate a Session ID. <vspace blankLines="1"/> Protocol change: Require that discovered addresses must be global (except during bootstrap). <vspace blankLines="1"/> Protocol change: Receiver of REQUEST message must close socket if no ASA is listening for the objective. <vspace blankLines="1"/> Protocol change: Simplified Waiting message. <vspace blankLines="1"/> Protocol change: Added No Operation message. <vspace blankLines="1"/> Renamed URL locator type as URI locator type. <vspace blankLines="1"/> Updated CDDL definition. <vspace blankLines="1"/> Various other clarifications and editorial fixes. </t> <t>draft-ietf-anima-grasp-01, 2015-10-09: <vspace blankLines="1"/> Updated requirements after list discussion. <vspace blankLines="1"/> Changed from TLV to CBOR format - many detailed changes, added co-author. <vspace blankLines="1"/> Tightened up loop count and timeouts for various cases. <vspace blankLines="1"/> Noted that GRASP does not provide transactional integrity. <vspace blankLines="1"/> Various other clarifications and editorial fixes. </t> <t>draft-ietf-anima-grasp-00, 2015-08-14: <vspace blankLines="1"/> File name and protocol name changed following WG adoption. <vspace blankLines="1"/> Added URL locator type. </t> <t>draft-carpenter-anima-gdn-protocol-04, 2015-06-21: <vspace blankLines="1"/> Tuned wording around hierarchical structure. <vspace blankLines="1"/> Changed "device" to "ASA" in many places. <vspace blankLines="1"/> Reformulated requirements to be clear that the ASA is the main customer for signaling. <vspace blankLines="1"/> Added requirement for flooding unsolicited synch, and added it to protocol spec. Recognized DNCP as alternative for flooding synch data. <vspace blankLines="1"/> Requirements clarified, expanded and rearranged following design team discussion. <vspace blankLines="1"/> Clarified that GDNP discovery must not be a prerequisite for GDNP negotiation or synchronization (resolved issue 13). <vspace blankLines="1"/> Specified flag bits for objective options (resolved issue 15). <vspace blankLines="1"/> Clarified usage of ACP vs TLS/DTLS and TCP vs UDP (resolved issues 9,10,11). <vspace blankLines="1"/> Updated DNCP description from latest DNCP draft. <vspace blankLines="1"/> Editorial improvements.</t> <t>draft-carpenter-anima-gdn-protocol-03, 2015-04-20: <vspace blankLines="1"/> Removed intrinsic security, required external security <vspace blankLines="1"/> Format changes to allow DNCP co-existence <vspace blankLines="1"/> Recognized DNS-SD as alternative discovery method. <vspace blankLines="1"/> Editorial improvements</t> <t>draft-carpenter-anima-gdn-protocol-02, 2015-02-19: <vspace blankLines="1"/> Tuned requirements to clarify scope, <vspace blankLines="1"/> Clarified relationship between types of objective, <vspace blankLines="1"/> Clarified that objectives may be simple values or complex data structures, <vspace blankLines="1"/> Improved description of objective options, <vspace blankLines="1"/> Added loop-avoidance mechanisms (loop count and default timeout, limitations on discovery relaying and on unsolicited responses), <vspace blankLines="1"/> Allow multiple discovery objectives in one response, <vspace blankLines="1"/> Provided for missing or multiple discovery responses, <vspace blankLines="1"/> Indicated how modes such as "dry run" should be supported, <vspace blankLines="1"/> Minor editorial and technical corrections and clarifications, <vspace blankLines="1"/> Reorganized future work list. </t> <t>draft-carpenter-anima-gdn-protocol-01, restructured the logical flowexplanation of its semantics to allow independent implementations. See <xref target="ConsOption" format="default"/> for more details. If thedocument, updatednew objective is similar in name or purpose todescribe synchronization completely, add unsolicited responses, numerous corrections and clarifications, expanded future work list, 2015-01-06.a previously registered objective, the specification should explain why a new objective is justified. </t><t>draft-carpenter-anima-gdn-protocol-00, combination<t>The following initial values are assigned by this document:</t> <table anchor="obj-names"> <name>Initial Values ofdraft-jiang-config-negotiation-ps-03 and draft-jiang-config-negotiation-protocol-02, 2014-10-08.</t>the "GRASP Objective Names" Subregistry</name> <thead> <tr><th>Objective Name</th><th>Reference</th></tr> </thead> <tbody> <tr> <td>EX0</td> <td>RFC 8990</td> </tr> <tr> <td>EX1</td> <td>RFC 8990</td> </tr> <tr> <td>EX2</td> <td>RFC 8990</td> </tr> <tr> <td>EX3</td> <td>RFC 8990</td> </tr> <tr> <td>EX4</td> <td>RFC 8990</td> </tr> <tr> <td>EX5</td> <td>RFC 8990</td> </tr> <tr> <td>EX6</td> <td>RFC 8990</td> </tr> <tr> <td>EX7</td> <td>RFC 8990</td> </tr> <tr> <td>EX8</td> <td>RFC 8990</td> </tr> <tr> <td>EX9</td> <td>RFC 8990</td> </tr> </tbody> </table> </section> </middle> <back> <displayreference target="I-D.stenberg-anima-adncp" to="ADNCP"/> <displayreference target="I-D.chaparadza-intarea-igcp" to="IGCP"/> <displayreference target="I-D.ietf-anima-asa-guidelines" to="ASA-GUIDELINES"/> <references> <name>References</name> <references> <name>Normative References</name> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2119.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8174.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4086.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3986.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8949.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7217.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3629.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8085.xml"/> <reference anchor="RFC8994" target="https://www.rfc-editor.org/info/rfc8994"> <front> <title>An Autonomic Control Plane (ACP)</title> <author initials="T" surname="Eckert" fullname="Toerless Eckert" role="editor"> <organization/> </author> <author initials="M" surname="Behringer" fullname="Michael H. Behringer" role="editor"> <organization/> </author> <author initials="S" surname="Bjarnason" fullname="Steinthor Bjarnason"> <organization/> </author> <date month="May" year="2021"/> </front> <seriesInfo name="RFC" value="8994"/> <seriesInfo name="DOI" value="10.17487/RFC8994"/> </reference> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8610.xml"/> </references> <references> <name>Informative References</name> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2334.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3493.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8126.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6733.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2865.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.4861.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5971.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6241.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2205.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.3416.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8415.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.5612.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6887.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6762.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6763.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.2608.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.6206.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8264.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7575.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7576.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7558.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7787.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.7788.xml"/> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8040.xml"/> <reference anchor="RFC8991" target="https://www.rfc-editor.org/info/rfc8991"> <front> <title>GeneRic Autonomic Signaling Protocol Application Program Interface (GRASP API)</title> <author initials="B" surname="Carpenter" fullname="Brian Carpenter"> <organization/> </author> <author initials="B" surname="Liu" fullname="Bing Liu" role="editor"> <organization/> </author> <author initials="W" surname="Wang" fullname="Wendong Wang"> <organization/> </author> <author initials="X" surname="Gong" fullname="Xiangyang Gong"> <organization/> </author> <date month="May" year="2021"/> </front> <seriesInfo name="RFC" value="8991"/> <seriesInfo name="DOI" value="10.17487/RFC8991"/> </reference> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.stenberg-anima-adncp.xml"/> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.chaparadza-intarea-igcp.xml"/> <reference anchor="RFC8993" target="https://www.rfc-editor.org/info/rfc8993"> <front> <title>A Reference Model for Autonomic Networking</title> <author initials="M" surname="Behringer" fullname="Michael H. Behringer" role="editor"> <organization/> </author> <author initials="B" surname="Carpenter" fullname="Brian Carpenter"> <organization/> </author> <author initials="T" surname="Eckert" fullname="Toerless Eckert"> <organization/> </author> <author initials="L" surname="Ciavaglia" fullname="Laurent Ciavaglia"> <organization/> </author> <author initials="J" surname="Nobre" fullname="Jéferson Campos Nobre"> <organization/> </author> <date month="May" year="2021"/> </front> <seriesInfo name="RFC" value="8993"/> <seriesInfo name="DOI" value="10.17487/RFC8993"/> </reference> <reference anchor="RFC8995" target="https://www.rfc-editor.org/info/rfc8995"> <front> <title>Bootstrapping Remote Secure Key Infrastructure (BRSKI)</title> <author initials="M" surname="Pritikin" fullname="Max Pritikin"> <organization/> </author> <author initials="M" surname="Richardson" fullname="Michael C. Richardson"> <organization/> </author> <author initials="T" surname="Eckert" fullname="Toerless Eckert"> <organization/> </author> <author initials="M" surname="Behringer" fullname="Michael H. Behringer"> <organization/> </author> <author initials="K" surname="Watsen" fullname="Kent Watsen"> <organization/> </author> <date month="May" year="2021"/> </front> <seriesInfo name="RFC" value="8995"/> <seriesInfo name="DOI" value="10.17487/RFC8995"/> </reference> <xi:include href="https://xml2rfc.ietf.org/public/rfc/bibxml/reference.RFC.8368.xml"/> <xi:include href="https://datatracker.ietf.org/doc/bibxml3/reference.I-D.ietf-anima-asa-guidelines.xml"/> </references> </references> <section anchor="examples"title="Examplenumbered="true" toc="default"> <name>Example MessageFormats">Formats</name> <t>For readers unfamiliar with CBOR, this appendix shows a number of example GRASP messages conforming to the CDDL syntax given in <xreftarget="cddl"/>.target="cddl" format="default"/>. Each message is shown three times in the following formats:<list style="numbers"> <t>CBOR</t> <ol spacing="normal" type="1"> <li>CBOR diagnosticnotation.</t> <t>Similar,notation.</li> <li>Similar, but showing the names of the constants. (Details of the flag bit encoding are omitted.)</t> <t>Hexadecimal</li> <li>Hexadecimal version of the CBOR wireformat.</t> </list>format.</li> </ol> <t> Long lines are split for display purposes only.</t> <sectiontitle="Discovery Example">numbered="true" toc="default"> <name>Discovery Example</name> <t>The initiator (2001:db8:f000:baaa:28cc:dc4c:9703:6781) multicasts adiscoveryDiscovery message looking for objective EX1:</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [1, 13948744, h'20010db8f000baaa28ccdc4c97036781', ["EX1", 5, 2, 0]] [M_DISCOVERY, 13948744, h'20010db8f000baaa28ccdc4c97036781', ["EX1", F_SYNCH_bits, 2, 0]] h'84011a00d4d7485020010db8f000baaa28ccdc4c970367818463455831050200' ]]></artwork></figure></t><t>A peer (2001:0db8:f000:baaa:f000:baaa:f000:baaa) responds with a locator:</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [2, 13948744, h'20010db8f000baaa28ccdc4c97036781', 60000, [103, h'20010db8f000baaaf000baaaf000baaa', 6, 49443]] [M_RESPONSE, 13948744, h'20010db8f000baaa28ccdc4c97036781', 60000, [O_IPv6_LOCATOR, h'20010db8f000baaaf000baaaf000baaa', IPPROTO_TCP, 49443]] h'85021a00d4d7485020010db8f000baaa28ccdc4c9703678119ea6084186750 20010db8f000baaaf000baaaf000baaa0619c123' ]]></artwork></figure></t></section> <sectiontitle="Flood Example">numbered="true" toc="default"> <name>Flood Example</name> <t>The initiator multicasts afloodFlood Synchronization message. The single objective has a null locator. There is no response:</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [9, 3504974, h'20010db8f000baaa28ccdc4c97036781', 10000, [["EX1", 5, 2, ["Example 1 value=", 100]],[] ] ] [M_FLOOD, 3504974, h'20010db8f000baaa28ccdc4c97036781', 10000, [["EX1", F_SYNCH_bits, 2, ["Example 1 value=", 100]],[] ] ]h'86091a00357b4e5020010db8f000baaa28ccdc4c97036781192710h'85091a00357b4e5020010db8f000baaa28ccdc4c97036781192710 828463455831050282704578616d706c6520312076616c75653d186480' ]]></artwork></figure></t></section> <sectiontitle="Synchronization Example">numbered="true" toc="default"> <name>Synchronization Example</name> <t>Following successful discovery of objective EX2, the initiator unicasts arequest:</t> <t><figure> <artwork><![CDATA[Request Synchronization message:</t> <artwork name="grasp-examples.txt" align="left"><![CDATA[ [4, 4038926, ["EX2", 5, 5, 0]] [M_REQ_SYN, 4038926, ["EX2", F_SYNCH_bits, 5, 0]] h'83041a003da10e8463455832050500' ]]></artwork></figure></t><t>The peer responds with a value:</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [8, 4038926, ["EX2", 5, 5, ["Example 2 value=", 200]]] [M_SYNCH, 4038926, ["EX2", F_SYNCH_bits, 5, ["Example 2 value=", 200]]] h'83081a003da10e8463455832050582704578616d706c6520322076616c75653d18c8' ]]></artwork></figure></t></section> <sectiontitle="Simplenumbered="true" toc="default"> <name>Simple NegotiationExample">Example</name> <t>Following successful discovery of objective EX3, the initiator unicasts arequest:</t> <t><figure> <artwork><![CDATA[Request Negotiation message:</t> <artwork name="grasp-examples.txt" align="left"><![CDATA[ [3, 802813, ["EX3", 3, 6, ["NZD", 47]]] [M_REQ_NEG, 802813, ["EX3", F_NEG_bits, 6, ["NZD", 47]]] h'83031a000c3ffd8463455833030682634e5a44182f' ]]></artwork></figure></t><t>The peer responds with immediate acceptance. Note that no objective isneeded,needed because the initiator's request was accepted without change:</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [6, 802813, [101]] [M_END , 802813, [O_ACCEPT]] h'83061a000c3ffd811865' ]]></artwork></figure></t></section> <sectiontitle="Completenumbered="true" toc="default"> <name>Complete NegotiationExample">Example</name> <t>Again the initiator unicasts arequest:</t> <t><figure> <artwork><![CDATA[Request Negotiation message:</t> <artwork name="grasp-examples.txt" align="left"><![CDATA[ [3, 13767778, ["EX3", 3, 6, ["NZD", 410]]] [M_REQ_NEG, 13767778, ["EX3", F_NEG_bits, 6, ["NZD", 410]]] h'83031a00d214628463455833030682634e5a4419019a' ]]></artwork></figure></t><t>The responder starts to negotiate (making an offer):</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [5, 13767778, ["EX3", 3, 6, ["NZD", 80]]] [M_NEGOTIATE, 13767778, ["EX3", F_NEG_bits, 6, ["NZD", 80]]] h'83051a00d214628463455833030682634e5a441850' ]]></artwork></figure></t><t>The initiator continues to negotiate (reducing its request, and note that the loop count is decremented):</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [5, 13767778, ["EX3", 3, 5, ["NZD", 307]]] [M_NEGOTIATE, 13767778, ["EX3", F_NEG_bits, 5, ["NZD", 307]]] h'83051a00d214628463455833030582634e5a44190133' ]]></artwork></figure></t><t>The responder asks for more time:</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [7, 13767778, 34965] [M_WAIT, 13767778, 34965] h'83071a00d21462198895' ]]></artwork></figure></t><t>The responder continues to negotiate (increasing its offer):</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [5, 13767778, ["EX3", 3, 4, ["NZD", 120]]] [M_NEGOTIATE, 13767778, ["EX3", F_NEG_bits, 4, ["NZD", 120]]] h'83051a00d214628463455833030482634e5a441878' ]]></artwork></figure></t><t>The initiator continues to negotiate (reducing its request):</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [5, 13767778, ["EX3", 3, 3, ["NZD", 246]]] [M_NEGOTIATE, 13767778, ["EX3", F_NEG_bits, 3, ["NZD", 246]]] h'83051a00d214628463455833030382634e5a4418f6' ]]></artwork></figure></t><t>The responder refuses to negotiate further:</t><t><figure> <artwork><![CDATA[<artwork name="grasp-examples.txt" align="left"><![CDATA[ [6, 13767778, [102, "Insufficient funds"]] [M_END , 13767778, [O_DECLINE, "Insufficient funds"]] h'83061a00d2146282186672496e73756666696369656e742066756e6473' ]]></artwork></figure></t><t>This negotiation has failed. If either side had sent [M_END, 13767778, [O_ACCEPT]] it would have succeeded, converging on the objective value in the preceding M_NEGOTIATE. Note that apart from the initial M_REQ_NEG, the process is symmetrical.</t> </section> </section> <section anchor="reqts"title="Requirementnumbered="true" toc="default"> <name>Requirement Analysis of Discovery,SynchronizationSynchronization, andNegotiation">Negotiation</name> <t>This section discusses the requirements for discovery,negotiationnegotiation, and synchronization capabilities. The primary user of the protocol is anautonomic service agentAutonomic Service Agent (ASA), so the requirements are mainly expressed as the features needed by an ASA. A single physical device might contain several ASAs, and a single ASA might manage several technical objectives. If a technical objective is managed by several ASAs, any necessary coordination is outside the scope ofthe GRASP signaling protocol.GRASP. Furthermore, requirements for ASAs themselves, such as the processing of Intent <xreftarget="RFC7575"/>,target="RFC7575" format="default"/>, are out of scope for the present document.</t> <sectiontitle="Requirements for Discovery"> <t>D1. ASAsnumbered="true" toc="default"> <name>Requirements for Discovery</name> <ol type="D%d." indent="6"> <li> <t>ASAs may be designed to manage any type of configurable device or software, as required in <xreftarget="synchreq"/>.target="synchreq" format="default"/>. A basic requirement is therefore that the protocol can represent and discover any kind of technical objective (as defined in <xreftarget="terms"/>)target="terms" format="default"/>) among arbitrary subsets of participating nodes.</t> <t>In anautonomic networkAutonomic Network, we must assume that when a device startsupup, it has no information about any peer devices, the network structure, orwhatthe specific role it must play. The ASA(s) inside the device are in the same situation. In some cases, when a new application session startsupwithin a device, the device or ASA may again lack information about relevant peers. For example, it might be necessary to set up resources on multiple other devices, coordinated and matched to each other so that there is no wasted resource. Security settings might also need updating to allow for the new device or user. The relevant peers may be different for different technical objectives. Therefore discovery needs to be repeated as often as necessary to find peers capable of acting as counterparts for each objective that a discovery initiator needs to handle. From this background we derive the next three requirements:</t><t>D2. When</li> <li>When an ASA first starts up, it may have no knowledge of the specific network to which it is attached. Therefore the discovery process must be able to support any network scenario, assuming only that the device concerned is bootstrapped from factory condition.</t> <t>D3. When</li> <li>When an ASA starts up, it must require no configured location information about any peers in order to discoverthem.</t> <t>D4. Ifthem.</li> <li>If an ASA supports multiple technical objectives, relevant peers may be different for different discovery objectives, so discovery needs to be performed separately to find counterparts for each objective. Thus, there must be a mechanism by which an ASA can separately discover peer ASAs for each of the technical objectives that it needs to manage, whenevernecessary.</t> <t>D5. Followingnecessary.</li> <li>Following discovery, an ASA will normally perform negotiation or synchronization for the corresponding objectives. The design should allow for this by conveniently linking discovery to negotiation and synchronization. It may provide an optional mechanism to combine discovery and negotiation/synchronization in a single protocolexchange.</t> <t>D6. Someexchange.</li> <li>Some objectives may only be significant on the local link, but others may be significant across the routed network and require off-link operations. Thus, the relevant peers might be immediate neighbors on the same layer 2 link, or they might be more distant and only accessible via layer 3. The mechanism must therefore provide both on-link and off-link discovery of ASAs supporting specific technicalobjectives.</t> <t>D7. Theobjectives.</li> <li> <t>The discovery process should be flexible enough to allow for special cases, such as the following:<list style="symbols"> <t>During</t> <ul spacing="normal"> <li>During initialization, a device must be able to establish mutual trust with autonomic nodes elsewhere in the network and participate in an authentication mechanism. Although this will inevitably start with a discovery action, it is a special case precisely because trust is not yet established. This topic is the subject of <xreftarget="I-D.ietf-anima-bootstrapping-keyinfra"/>.target="RFC8995" format="default"/>. We require that once trust has been established for a device, all ASAs within the device inherit the device's credentials and are also trusted. This does not preclude the device having multiplecredentials.</t> <t>credentials.</li> <li> Depending on the type of network involved, discovery of other central functions might be needed, such as the Network Operations Center (NOC) <xreftarget="I-D.ietf-anima-stable-connectivity"/>.target="RFC8368" format="default"/>. The protocol must be capable of supporting such discovery during initialization, as well as discovery during ongoingoperation.</t> </list></t> <t>D8. Theoperation.</li> </ul> </li> <li>The discovery process must not generate excessive traffic and must take account of sleepingnodes. </t> <t>D9. There must be a mechanism for handling stale discovery results.</t> </section> <section anchor="synchreq" title="Requirements for Synchronization and Negotiation Capability"> <!--<t>As background, consider the example of routing protocols, the closest approximation to autonomic networking already in widespread use. Routing protocols use a largely autonomic model based on distributed devices that communicate repeatedly with each other. The focus is reachability, so routing protocols primarily consider simple link status and metrics, and an underlying assumption is that nodes need a consistent, although partial, view of the network topology in ordernodes. </li> <li>There must be a mechanism forthe routing algorithm to converge. Also, routing is mainly based on simple information synchronization between peers, rather than on bi-directional negotiation.</t>-->handling stale discovery results.</li> </ol> </section> <section anchor="synchreq" numbered="true" toc="default"> <name>Requirements for Synchronization and Negotiation Capability</name> <t>AutonomicnetworksNetworks need to be able to manage many different types ofparameterparameters and consider many dimensions, such as latency, load, unused or limited resources, conflicting resource requests, security settings, power saving, load balancing, etc. Status information and resource metrics need to be shared between nodes for dynamic adjustment of resources and for monitoring purposes. While this might be achieved by existing protocols when they are available, the new protocol needs to be able to support parameter exchange, including mutual synchronization, even when no negotiation as such is required. In general, these parameters do not apply to all participating nodes, but only to a subset. </t><t>SN1. A<ol type="SN%d." indent="6"> <li>A basic requirement for the protocol is therefore the ability to represent, discover,synchronizesynchronize, and negotiate almost any kind of network parameter among selected subsets of participatingnodes.</t> <t>SN2. Negotiationnodes.</li> <li>Negotiation is an iterative request/response process that must be guaranteed to terminate (with success or failure). While tie-breaking rules must be defined specifically for each use case, the protocol should have some general mechanisms in support of loop and deadlock prevention, such ashop counthop-count limits ortimeouts.</t> <t>SN3. Synchronizationtimeouts.</li> <li>Synchronization must be possible for groups of nodes ranging from small to very large.</t> <t>SN4. To</li> <li>To avoid "reinventing the wheel", the protocol should be able to encapsulate the data formats used by existing configuration protocols (such asNETCONF/YANG)Network Configuration Protocol (NETCONF) and YANG) in cases where that isconvenient.</t> <t>SN5. Humanconvenient.</li> <li>Human intervention in complex situations is costly anderror-prone.error prone. Therefore, synchronization or negotiation of parameters without human intervention is desirable whenever the coordination of multiple devices can improve overall network performance. It follows that the protocol's resource requirements must be small enough to fit in any device that would otherwise need human intervention. The issue of running in constrained nodes is discussed in <xreftarget="I-D.ietf-anima-reference-model"/>.</t> <t>SN6. Humantarget="RFC8993" format="default"/>.</li> <li>Human intervention in large networks is often replaced by use of a top-down network management system (NMS). It therefore follows that the protocol, as part of the Autonomic Networking Infrastructure, should be capable of running in any device that would otherwise be managed by an NMS, and that it canco-existcoexist with anNMS,NMS and with protocols such as SNMP andNETCONF.</t> <t>SN7. SpecificNETCONF.</li> <li><t>Specific autonomic features are expected to be implemented by individual ASAs, but the protocol must be general enough to allow them. Some examples follow:<list style="symbols"> <t>Dependencies</t> <ul spacing="normal"> <li>Dependencies and conflicts: In order to decide upon a configuration for a given device, the device may need information from neighbors. This can be established through the negotiation procedure, or through synchronization if that is sufficient. However, a given item in a neighbor may depend on other information from its own neighbors, which may need another negotiation or synchronization procedure to obtain or decide. Therefore, there are potential dependencies and conflicts among negotiation or synchronization procedures. Resolving dependencies and conflicts is a matter for the individual ASAs involved. To allow this, there need to be clear boundaries and convergence mechanisms for negotiations. Also some mechanisms are needed to avoid loop dependencies or uncontrolled growth in a tree of dependencies. It is the ASA designer's responsibility to avoid or detect looping dependencies or excessive growth of dependency trees. The protocol's role is limited to bilateral signaling betweenASAs,ASAs and the avoidance of loops during bilateralsignaling.</t> <t>Recoverysignaling.</li> <li>Recovery from faults and identification of faulty devices should be as automatic as possible. The protocol's role is limited to discovery,synchronizationsynchronization, and negotiation. These processes can occur at any time, and an ASA may need to repeat any of these steps when the ASA detects an event such as a negotiation counterpartfailing.</t> <t>Sincefailing.</li> <li>Since a major goal is to minimize human intervention, it is necessary that the network can in effect "think ahead" before changing its parameters. One aspect of this is an ASA that relies on a knowledge base to predict network behavior. This is out of scope for the signaling protocol. However, another aspect is forecasting the effect of a change by a "dry run" negotiation before actually installing the change. Signaling a dry run is therefore a desirable feature of the protocol.</t> </list></t></li> </ul> <t>Note that management logging, monitoring,alertsalerts, and tools for intervention are required. However, these can only be features of individual ASAs, not of the protocol itself. Another document <xreftarget="I-D.ietf-anima-stable-connectivity"/>target="RFC8368" format="default"/> discusses how such agents may be linked into conventionalOAMOperations, Administration, and Maintenance (OAM) systems via an Autonomic Control Plane <xreftarget="I-D.ietf-anima-autonomic-control-plane"/>.target="RFC8994" format="default"/>. </t><t>SN8. The</li> <li>The protocol will be able to deal with a wide variety of technical objectives, covering any type of network parameter. Therefore the protocol will need a flexible and easily extensible format for describing objectives. At a laterstagestage, it may be desirable to adopt an explicit information model. One consideration is whether to adopt an existing information model or to design a new one.</t></li> </ol> </section> <sectiontitle="Specificnumbered="true" toc="default"> <name>Specific TechnicalRequirements"> <t>T1. ItRequirements</name> <ol type="T%d." indent="6"> <li>It should be convenient for ASA designers to define new technical objectives and for programmers to express them, without excessive impact onrun-timeruntime efficiency and footprint. In particular, it should be convenient for ASAs to be implemented independently of each other asuser spaceuser-space programs rather than as kernel code, where such a programming model is possible. The classes of device in which the protocol might run is discussed in <xreftarget="I-D.ietf-anima-reference-model"/>. </t> <t>T2. Thetarget="RFC8993" format="default"/>. </li> <li>The protocol should be easily extensible in case the initially defined discovery,synchronizationsynchronization, and negotiation mechanisms prove to be insufficient.</t> <t>T3. To</li> <li>To be a generic platform, the protocol payload format should be independent of the transport protocol or IP version. In particular, it should be able to run over IPv6 or IPv4. However, some functions, such as multicasting on a link, might need to be IP version dependent. By default, IPv6 should bepreferred.</t> <t>T4. Thepreferred.</li> <li>The protocol must be able to access off-link counterparts via routable addresses, i.e., must not be restricted to link-localoperation.</t> <t>T5. Itoperation.</li> <li>It must also be possible for an external discovery mechanism to be used, if appropriate for a given technical objective. In other words, GRASP discovery must not be a prerequisite for GRASP negotiation or synchronization.</t> <t>T6. The</li> <li>The protocol must be capable of distinguishing multiple simultaneous operations with one or more peers, especially when wait statesoccur.</t> <t>T7. Intent:occur.</li> <li>Intent: Although the distribution of Intent is out of scope for this document, the protocol must not by design exclude its use for Intent distribution.</t> <t>T8. Management</li> <li>Management monitoring,alertsalerts, and intervention: Devices should be able to report to a monitoring system. Some events must be able to generate operatoralertsalerts, and some provision for emergency intervention must be possible(e.g.(e.g., to freeze synchronization or negotiation in amis-behavingmisbehaving device). These features might not use the signaling protocol itself, but its design should not exclude suchuse.</t> <t>T9. Becauseuse.</li> <li>Because this protocol may directly cause changes to device configurations and have significant impacts on a running network, all protocol exchanges need to be fully secured against forged messages andman-in-the middleman-in-the-middle attacks, and secured as much as reasonably possible againstdenial of servicedenial-of-service attacks. There must also be an encryption mechanism to resist unwanted monitoring. However, it is not required that the protocol itself provides these security features; it may depend on an existing secure environment.</t></li> </ol> </section> </section><!-- reqts --><section anchor="current"title="Capabilitynumbered="true" toc="default"> <name>Capability Analysis of CurrentProtocols">Protocols</name> <t>This appendix discusses various existing protocols with properties related to the requirements described in <xreftarget="reqts"/>.target="reqts" format="default"/>. The purpose is to evaluate whether any existing protocol, or a simple combination of existing protocols, can meet those requirements.</t> <t>Numerous protocols include some form of discovery, but these all appear to be very specific in their applicability. Service Location Protocol (SLP) <xreftarget="RFC2608"/>target="RFC2608" format="default"/> provides service discovery for managed networks, but it requires configuration of its own servers.DNS-SDDNS-Based Service Discovery (DNS-SD) <xreftarget="RFC6763"/>target="RFC6763" format="default"/> combined withmDNSMulticast DNS (mDNS) <xreftarget="RFC6762"/>target="RFC6762" format="default"/> provides service discovery for small networks with a single link layer. <xreftarget="RFC7558"/>target="RFC7558" format="default"/> aims to extend this to larger autonomousnetworksnetworks, but this is not yet standardized. However, both SLP and DNS-SD appear to target primarilyapplication layerapplication-layer services, not the layer 2 and 3 objectives relevant to basic network configuration. Both SLP and DNS-SD are text-based protocols. </t><!-- <t>Routing protocols are mainly one-way information announcements. The receiver makes independent decisions based on the received information and there is no direct feedback information to the announcing peer. This remains true even though the protocol is used in both directions between peer routers; there is state synchronization, but no negotiation, and each peer runs its route calculations independently.</t>--><t>Simple Network Management Protocol (SNMP) <xreftarget="RFC3416"/>target="RFC3416" format="default"/> uses a command/response model not well suited for peer negotiation.Network Configuration Protocol (NETCONF)NETCONF <xreftarget="RFC6241"/>target="RFC6241" format="default"/> uses an RPC model that does allow positive or negative responses from the target system, but this is still not adequate for negotiation.</t> <t>There are various existing protocols that have elementary negotiation abilities, such as Dynamic Host Configuration Protocol for IPv6 (DHCPv6) <xreftarget="RFC3315"/>,target="RFC8415" format="default"/>, Neighbor Discovery (ND) <xreftarget="RFC4861"/>,target="RFC4861" format="default"/>, Port Control Protocol (PCP) <xreftarget="RFC6887"/>,target="RFC6887" format="default"/>, Remote AuthenticationDial InDial-In User Service (RADIUS) <xreftarget="RFC2865"/>,target="RFC2865" format="default"/>, Diameter <xreftarget="RFC6733"/>,target="RFC6733" format="default"/>, etc. Most of them are configuration or management protocols. However, they either provide only a simple request/response model in a master/slave context or very limited negotiation abilities.</t> <t>There are some signaling protocols with an element of negotiation. Forexampleexample, Resource ReSerVation Protocol (RSVP) <xreftarget="RFC2205"/>target="RFC2205" format="default"/> was designed for negotiatingquality of servicequality-of-service parameters along the path of a unicast or multicast flow. RSVP is a veryspecialisedspecialized protocol aimed at end-to-end flows.<!--However, it has some flexibility, having been adapted for MPLS label distribution (RSVP-TE, <xref target="RFC3209"/>).-->A more generic design is General Internet Signalling Transport (GIST) <xreftarget="RFC5971"/>, buttarget="RFC5971" format="default"/>; however, itis complex,tries to solve many problems, making it complex, and is also aimed at per-flow signaling across many hops rather than at device-to-device signaling. However, we cannot completely exclude extended RSVP or GIST as a synchronization and negotiation protocol. They do not appear to be directlyuseableusable for peer discovery.</t> <t>RESTCONF <xreftarget="RFC8040"/>target="RFC8040" format="default"/> is a protocol intended to convey NETCONF information expressed in the YANG language via HTTP, including the ability to transit HTML intermediaries. While this is a powerful approach in the context ofcentralisedcentralized configuration of a complex network, it is not well adapted to efficient interactive negotiation between peer devices, especially simple ones that might not include YANG processing already.</t> <t>The Distributed Node Consensus Protocol (DNCP) <xreftarget="RFC7787"/>target="RFC7787" format="default"/> is defined as a generic form of a state synchronization protocol, with a proposed usage profile being the Home Networking Control Protocol (HNCP) <xreftarget="RFC7788"/>target="RFC7788" format="default"/> for configuring Homenet routers. A specific application of DNCP forautonomic networkingAutonomic Networking was proposed in <xreftarget="I-D.stenberg-anima-adncp"/>. </t> <t>DNCP "istarget="I-D.stenberg-anima-adncp" format="default"/>. According to <xref target="RFC7787" format="default"/>:</t> <blockquote><t>DNCP is designed to provide a way for each participating node to publish a set of TLV (Type-Length-Value)tuples,tuples (at most 64 KB) and to provide a shared and common view about the datapublished... DNCPpublished...</t> <t>DNCP is most suitable for data that changes onlyinfrequently... Ifinfrequently...</t> <t>If constant rapid state changes are needed, the preferable choice is to use an additional point-to-pointchannel..."</t>channel...</t></blockquote> <t>Specific features of DNCP include:<list style="symbols"> <t>Every</t> <ul spacing="normal"> <li>Every participating node has a unique nodeidentifier.</t> <t>DNCPidentifier.</li> <li>DNCP messages are encoded as a sequence of TLVobjects,objects and sent over unicast UDP or TCP, with or without (D)TLSsecurity.</t> <t>Multicastsecurity.</li> <li>Multicast is used only for discovery of DNCP neighbors when lower security isacceptable.</t> <t>Synchronizationacceptable.</li> <li>Synchronization of state is maintained by a flooding process using the Trickle algorithm. There is no bilateral synchronization or negotiationcapability.</t> <t>Thecapability.</li> <li>The HNCP profile of DNCP is designed to operate between directly connected neighbors on a shared link using UDP and link-local IPv6addresses.</t> </list>addresses.</li> </ul> <t> DNCP does not meet the needs of a general negotiationprotocol,protocol because it is designed specifically for flooding synchronization. Also, in its HNCPprofileprofile, it is limited to link-local messages and to IPv6. However, at theminimumminimum, it is a very interesting test case for this style of interaction between devices without needing a central authority, and it is a proven method of network-wide state synchronization by flooding.</t> <t>The Server Cache Synchronization Protocol (SCSP) <xreftarget="RFC2334"/>target="RFC2334" format="default"/> also describes a method for cache synchronization and cache replication among a group of nodes.</t> <t>A proposal was made some years ago for an IP based Generic Control Protocol (IGCP) <xreftarget="I-D.chaparadza-intarea-igcp"/>.target="I-D.chaparadza-intarea-igcp" format="default"/>. This was aimed at information exchange and negotiation but not directly at peer discovery. However, it has many points in common with the present work.</t> <t>None of the above solutions appears to completely meet the needs of generic discovery, statesynchronizationsynchronization, and negotiation in a single solution. Many of the protocols assume that they are working in a traditional top-down or north-south scenario, rather than a fluid peer-to-peer scenario. Most of them are specialized in one way or another. As a result, we have not identified a combination of existing protocols that meets the requirements in <xreftarget="reqts"/>.target="reqts" format="default"/>. Also, we have not identified a path by which one of the existing protocols could be extended to meet the requirements. </t> </section><!-- current --><section anchor="ack" numbered="false" toc="default"> <name>Acknowledgments</name> <t>A major contribution to the original draft version of this document was made by <contact fullname="Sheng Jiang"/>, and significant contributions were made by <contact fullname="Toerless Eckert"/>. Significant early review inputs were received from <contact fullname="Joel Halpern"/>, <contact fullname="Barry Leiba"/>, <contact fullname="Charles E. Perkins"/>, and <contact fullname="Michael Richardson"/>. <contact fullname="William Atwood"/> provided important assistance in debugging a prototype implementation.</t> <t>Valuable comments were received from <contact fullname="Michael Behringer"/>, <contact fullname="Jéferson Campos Nobre"/>, <contact fullname="Laurent Ciavaglia"/>, <contact fullname="Zongpeng Du"/>, <contact fullname="Yu Fu"/>, <contact fullname="Joel Jaeggli"/>, <contact fullname="Zhenbin Li"/>, <contact fullname="Dimitri Papadimitriou"/>, <contact fullname="Pierre Peloso"/>, <contact fullname="Reshad Rahman"/>, <contact fullname="Markus Stenberg"/>, <contact fullname="Martin Stiemerling"/>, <contact fullname="Rene Struik"/>, <contact fullname="Martin Thomson"/>, <contact fullname="Dacheng Zhang"/>, and participants in the Network Management Research Group, the ANIMA Working Group, and the IESG.</t> </section> </back> </rfc>