<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE rfc [
  <!ENTITY nbsp    "&#160;">
  <!ENTITY zwsp   "&#8203;">
  <!ENTITY nbhy   "&#8209;">
  <!ENTITY wj     "&#8288;">
]>
<rfc xmlns:xi="http://www.w3.org/2001/XInclude"
     ipr="trust200902"
     docName="draft-han-fann-use-cases-00"
     category="info"
     submissionType="IETF"
     consensus="true"
     tocInclude="true"
     sortRefs="true"
     symRefs="true"
     xml:lang="en"
     version="3">

  <front>
    <title abbrev="FANN Use Cases">Use Cases for Fast Network Notifications (FANN)</title>
    <seriesInfo name="Internet-Draft" value="draft-han-fann-use-cases-00"/>

    <author fullname="Zhengxin Han" initials="Z." surname="Han">
      <organization>China Unicom</organization>
      <address>
        <postal>
          <city>Beijing</city>
          <country>China</country>
        </postal>
        <email>hanzx21@chinaunicom.cn</email>
      </address>
    </author>

    <author fullname="Mike McBride" initials="M." surname="McBride">
      <organization>Futurewei</organization>
      <address>
        <email>mmcbride7@gmail.com</email>
      </address>
    </author>

    <author fullname="Jie Dong" initials="J." surname="Dong">
      <organization>Huawei</organization>
      <address>
        <email>jie.dong@huawei.com</email>
      </address>
    </author>

    <author fullname="Jing Gao" initials="J." surname="Gao">
      <organization>CAICT</organization>
      <address>
        <postal>
          <city>Beijing</city>
          <country>China</country>
        </postal>
        <email>gaojing1@caict.ac.cn</email>
      </address>
    </author>

    <date year="2026" month="November"/>
    <area>Routing</area>
    <workgroup>Fast Network Notifications</workgroup>
    <keyword>FANN</keyword>
    <keyword>fast notification</keyword>
    <keyword>use cases</keyword>

    <abstract>
      <t>This document reviews the use cases for Fast
      Network Notifications (FANN) that have been described in a number of
      individual drafts. Use cases are grouped by deployment
      scenario (data center network, data center interconnect and wide
      area network) aligned with the deployment scenarios of the FANN
      framework. Each use case is summarized briefly
      with a reference to the document that describes it in detail. This is intended to help identify a representative set of
      use cases from which the key FANN requirements can be derived.</t>
    </abstract>
    
  </front>

  <middle>

    <section anchor="intro">
      <name>Introduction</name>
      <t>The FANN problem statement <xref target="I-D.ietf-fann-problem-statement"/>
      describes the gap between how quickly modern forwarding hardware can
      detect an adverse network condition and how quickly remote nodes
      that could act on that condition learn about it. The FANN framework
      <xref target="I-D.song-fann-framework"/> describes the functional
      roles, information model and delivery and scoping models for
      closing that gap.</t>

      <t>A number of individual drafts describe situations in which fast
      network notifications would be useful. These drafts cover different parts 
      of the network and sometimes describe the same situation from different 
      perspectives. This document brings these use cases together so the
      working group can:</t>
      <ul>
        <li>see which deployment scenarios and notification types are
        covered, and where gaps remain;</li>
        <li>identify overlap between drafts, and</li>
        <li>select a representative set of use cases from which the key
        FANN requirements can be derived.</li>
      </ul>

      <t>This document summarizes each use case briefly. This document 
      does not specify any protocol or mechanism and the actions taken by
      notification consumers are described only to explain why the
      notification is needed.</t>
    </section>

    <section anchor="terms">
      <name>Terminology</name>
      <t>This document uses the terms Notification Originator, Notification
      Consumer, Notification Relay and Notification Domain as defined in
      <xref target="I-D.song-fann-framework"/>. In addition:</t>
      <dl newline="false" spacing="normal">
        <dt>DCN:</dt>
        <dd>Data Center Network, including AI scale-out fabrics.</dd>
        <dt>DCI:</dt>
        <dd>Data Center Interconnect.</dd>
        <dt>WAN:</dt>
        <dd>Wide Area Network.</dd>
        <dt>AIDC:</dt>
        <dd>AI Data Center.</dd>
        <dt>PFC:</dt>
        <dd>Priority-based Flow Control.</dd>
        <dt>FGFC:</dt>
        <dd>Fine-Grained Flow Control, operating at flow, tenant or slice
        granularity.</dd>
        <dt>PLR / PRR:</dt>
        <dd>Point of Local Repair / Point of Remote Repair.</dd>
      </dl>
    </section>

    <section anchor="method">
      <name>Organization of Use Cases</name>
      <t>Use cases are grouped by deployment scenario: data center
      network (DCN), data center interconnect (DCI) and wide area network
      (WAN). These groupings align with the deployment scenarios in
      <xref target="I-D.song-fann-framework" section="3.7" sectionFormat="of"/>.
      Some use cases apply to more than one scenario. Each use case
      describes the event, the information the notification carries, who
      consumes it and the action taken.</t>
    </section>

    <section anchor="uc">
      <name>Use Cases</name>

      <section anchor="uc-dcn">
        <name>Data Center Network Use Cases</name>

        <section anchor="uc-dcn-frr">
          <name>Hardware-Speed Remote Protection in AI Fabrics</name>
          <t>AI training and inference fabrics are typically multi-tier
          folded Clos topologies, often running BGP only. Path diversity
          exists only in the upward direction, so a failure on a downward
          link (for example spine to leaf) cannot be bypassed by the node
          adjacent to it. Convergence targets for these fabrics are moving
          toward 100 microseconds, well below what CPU-mediated IP-FRR
          activation provides.</t>
          <t>The node that detects the failure in hardware originates a
          notification that is consumed, in the forwarding plane, by an
          upstream node that may be several hops away. That node activates
          a pre-installed backup path, avoiding the hairpinning that local
          TI-LFA repair can cause.</t>
          <t>References: <xref target="I-D.clad-rtgwg-ipfrr-aiml"/>,
          <xref target="I-D.clad-rtgwg-efficient-remote-protection"/>,
          <xref target="I-D.geng-fantel-fantel-requirements"/>.</t>
        </section>

        <section anchor="uc-dcn-lb">
          <name>Quality-Aware Load Sharing</name>
          <t>Not every event is a complete failure. A member of a link
          bundle can fail and reduce capacity, a link can degrade, or a
          link can become congested. If the notification carries a
          quantitative severity rather than up/down state, upstream nodes
          can adjust ECMP weights in proportion to remaining capacity or
          split traffic between the primary path and a repair path. This
          avoids withdrawing a usable resource entirely and avoids abrupt
          load shifts.</t>
          <t>References: <xref target="I-D.clad-rtgwg-ipfrr-aiml"/>,
          <xref target="I-D.clad-rtgwg-efficient-remote-protection"/>,
          <xref target="I-D.geng-fantel-fantel-requirements"/>.</t>
        </section>

        <section anchor="uc-dcn-plane">
          <name>Plane Evacuation in Multi-Plane Fabrics</name>
          <t>In a multi-plane AI fabric, a high-capacity NIC is split into
          several interfaces, each attached to an independent plane. A
          NIC-to-leaf link failure, a leaf-to-spine link failure or a
          degradation within a plane can make that plane unusable for some
          destinations. The node that detects the problem is a leaf or
          spine, but the node that must act is the source NIC.</t>
          <t>The notification identifies the affected plane and
          distinguishes between plane unreachable, plane degraded, and
          destination unreachable within a plane. The source NIC removes
          the plane from its candidate set for the affected destinations
          and moves traffic to the remaining planes. The notification must
          not leak across planes and its delivery path should not share
          fate with the failure.</t>
          <t>Reference: <xref target="I-D.lqg-fann-multi-plane-protection"/>.</t>
        </section>

        <section anchor="uc-dcn-runtime">
          <name>Notification to AI Runtimes and Communication Libraries</name>
          <t>Collective operations such as AllReduce and AllGather are
          sensitive to transient congestion, gradual degradation (jitter,
          intermittent loss, bandwidth fluctuation), and node or path
          failures, even without packet loss. Notifying communication
          libraries, runtimes and schedulers of these events, with
          context about the affected communication, lets them adapt
          communication patterns, coordinate recovery with network
          recovery or change workload placement before job completion
          time suffers.</t>
          <t>Reference: <xref target="I-D.sang-fann-fast-network-event-notification"/>.</t>

        </section>
      </section>

      <section anchor="uc-dci">
        <name>Data Center Interconnect Use Cases</name>

        <t>AI workloads increasingly span multiple data centers. Common
        AI DCI scenarios include coordinated training across AIDCs,
        distributed inference with model stages or KV cache in different
        DCs and AI fabrics spanning multiple domains and vendors.
        In each, interconnect traffic is sensitive to congestion and
        loss and the node that detects a problem is often far from the
        node that can act on it. The following subsections describe fast
        notification use cases for these scenarios.</t>

        <section anchor="uc-dci-pfc">
          <name>PFC and Fine-Grained Flow Control Interworking</name>
          <t>Lossless transport is extending from DCNs across WAN
          interconnects. PFC is unsuitable in the WAN because of head-of-line
          blocking, deadlock and congestion spreading, so the WAN uses
          fine-grained flow control at flow, tenant or slice granularity.
          The DCI edge nodes must translate between the two.</t>
          <t>For traffic flowing from a source DC across the WAN to a
          destination DC, backpressure propagates upstream along that same
          direction and requires interworking at both DCI edges. Between
          the destination DC and the WAN, congestion inside the
          destination DC produces a PFC pause toward the egress edge. When
          the edge buffer for that tenant reaches a threshold, the egress
          edge originates an FGFC backpressure notification carrying the
          tenant, priority, slice and pause time toward upstream WAN nodes
          or the ingress edge. Between the WAN and the source DC, the
          ingress edge receives an FGFC notification, either propagated
          from the egress edge or originated by a congested WAN node, and
          translates it into PFC toward the source DC.</t>
          <t>The draft describes both a single-hop mode, in which the edges
          are directly connected or intermediate nodes do not take part,
          and a multi-hop mode, in which intermediate WAN nodes relay
          backpressure.</t>
          <t>Reference: <xref target="I-D.han-fann-codeployment-pfc-fgfc"/>.</t>
        </section>

        <section anchor="uc-dci-coord">
          <name>Coordinated Flow Control and Congestion Control</name>
          <t>When AI workloads span multiple DCs, congestion control
          feedback loops stretch from microseconds to milliseconds.
          Receiver-based ECN feedback takes a full WAN round trip, and
          sending feedback directly from the congestion point to the
          source still takes about half a round trip and requires WAN
          nodes to understand transport state.</t>
          <t>Two cases are described. When congestion is in the WAN, the
          congested node sends a fine-grained flow control notification to
          its upstream neighbor, which relays it hop by hop toward the
          ingress DC gateway. WAN buffers absorb the burst while the
          source slows down. When congestion is inside the egress DC, PFC
          protects the fabric locally and the egress gateway converts the
          condition into a WAN flow control notification that propagates
          back to the ingress gateway, which may in turn notify the
          source. Notifications are directed along the reverse data path,
          not flooded, and are expected to be carried in IP.</t>
          <t>Reference: <xref target="I-D.dong-fann-fc-cc-coordinate"/>.</t>
        </section>

        <section anchor="uc-dci-frr">
          <name>Remote Protection Across Scale-Across Interconnects</name>
          <t>Several scale-out fabrics may be interconnected across a
          campus, metro, or wide area, running IS-IS or BGP over full-mesh
          or irregular topologies. The remote protection and quality-aware
          load sharing described in <xref target="uc-dcn-frr"/> and
          <xref target="uc-dcn-lb"/> also apply here, but the distance
          between originator and consumer is larger.</t>
          <t>For coordinated training across AIDCs, fast failure,
          congestion and status notifications also allow upstream rate
          reduction, dynamic load balancing and fast failover of
          parameter synchronization traffic, reducing the risk of a
          training rollback.</t>
          <t>References: <xref target="I-D.clad-rtgwg-ipfrr-aiml"/>,
          <xref target="I-D.clad-rtgwg-efficient-remote-protection"/>,
          <xref target="I-D.hhz-fantel-sar-wan"/>.</t>
        </section>

        <section anchor="uc-dci-xdomain">
          <name>Cross-Domain AI Fabric Operation</name>
          <t>AI infrastructure may span multiple operational domains and
          vendors, for example distributed sites connected over a WAN to
          central compute pools. Events in one domain need to be
          represented consistently so they can be interpreted in
          another.</t>
          <t>Reference: <xref target="I-D.sang-fann-fast-network-event-notification"/>.</t>
        </section>
      </section>

      <section anchor="uc-wan">
        <name>Wide Area Network Use Cases</name>

        <section anchor="uc-wan-migration">
          <name>Sample Data Migration</name>
          <t>Large training data sets are moved from customer sites to
          AIDC storage over the WAN, often over fixed leased capacity. Fast
          network status and failure notifications let a controller
          allocate bandwidth on demand, adjust TE and load balancing
          policy and fail over quickly when a failure would otherwise
          cause loss.</t>
          <t>References: <xref target="I-D.hhz-fantel-sar-wan"/>,
          <xref target="I-D.hs-rtgwg-wan-lossless-uc"/>.</t>
        </section>

        <section anchor="uc-wan-rdma">
          <name>Remote Data Access over RDMA</name>
          <t>Training jobs may read sample data directly from remote
          storage using RDMA over distances of hundreds of kilometers.
          RDMA tolerates loss poorly, so fast congestion and failure
          notification to adjacent or affected nodes is needed to support
          flow-based congestion control and lossless transport.</t>
          <t>Reference: <xref target="I-D.hhz-fantel-sar-wan"/>.</t>
        </section>

        <section anchor="uc-wan-edge">
          <name>Edge-Cloud Collaborative Training and Inference</name>
          <t>In split learning, part of a model runs on enterprise
          premises and part in an AIDC, exchanging activations and
          gradients. In collaborative inference, prefill may run locally
          and decode in the AIDC, with KV cache transfer across the WAN.
          Both are latency sensitive and share the interconnect with other
          tenants. Notifications need tenant, task, or flow attribution so
          that only the responsible traffic is acted on, together with
          congestion and failure information for TE and load balancing.</t>
          <t>References: <xref target="I-D.hhz-fantel-sar-wan"/>, <xref target="I-D.song-fann-framework"/>.</t>
        </section>
      </section>
    </section>

    <section anchor="iana">
      <name>IANA Considerations</name>
      <t>This document has no IANA actions.</t>
    </section>

  </middle>

  <back>
    <references>
      <name>Informative References</name>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.ietf-fann-problem-statement.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.song-fann-framework.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.han-fann-codeployment-pfc-fgfc.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.dong-fann-fc-cc-coordinate.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.hhz-fantel-sar-wan.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.hs-rtgwg-wan-lossless-uc.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.lqg-fann-multi-plane-protection.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.sang-fann-fast-network-event-notification.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.geng-fantel-fantel-requirements.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.clad-rtgwg-efficient-remote-protection.xml"/>
      <xi:include href="https://bib.ietf.org/public/rfc/bibxml3/reference.I-D.clad-rtgwg-ipfrr-aiml.xml"/>
    </references>

    <section anchor="ack" numbered="false">
      <name>Acknowledgements</name>
      <t>The editors thank the authors of the drafts referenced in this
      document, whose use cases are summarized here.</t>
    </section>
  </back>
</rfc>
