One document matched: draft-ietf-tsvwg-sctp-failover-10.xml


<?xml version="1.0" encoding="US-ASCII"?>
<!-- edited with XMLSPY v5 rel. 3 U (http://www.xmlspy.com)
     by Daniel M Kohn (private) -->
<!DOCTYPE rfc SYSTEM "rfc2629.dtd" [
<!ENTITY rfc2119 PUBLIC "" "http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml">
<!ENTITY rfc4690 PUBLIC "" "http://xml.resource.org/public/rfc/bibxml/reference.RFC.4960.xml">
]>
<rfc category="std" docName="draft-ietf-tsvwg-sctp-failover-10.txt"
     ipr="trust200902">
  <?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>

  <?rfc toc="yes" ?>

  <?rfc symrefs="yes" ?>

  <?rfc sortrefs="yes"?>

  <?rfc iprnotified="no" ?>

  <?rfc strict="yes" ?>

  <front>
    <title abbrev="SCTP-PF">SCTP-PF: Quick Failover Algorithm in SCTP</title>

    <author fullname="Yoshifumi Nishida" initials="Y.N" surname="Nishida">
      <organization>GE Global Research</organization>

      <address>
        <postal>
          <street>2623 Camino Ramon</street>

          <city>San Ramon</city>

          <region>CA</region>

          <code>94583</code>

          <country>USA</country>
        </postal>

        <email>nishida@wide.ad.jp</email>
      </address>
    </author>

    <author fullname="Preethi Natarajan" initials="P.N" surname="Natarajan">
      <organization>Cisco Systems</organization>

      <address>
        <postal>
          <street>510 McCarthy Blvd</street>

          <city>Milpitas</city>

          <region>CA</region>

          <code>95035</code>

          <country>USA</country>
        </postal>

        <email>prenatar@cisco.com</email>
      </address>
    </author>

    <author fullname="Armando Caro" initials="A.C" surname="Caro">
      <organization>BBN Technologies</organization>

      <address>
        <postal>
          <street>10 Moulton St.</street>

          <city>Cambridge</city>

          <region>MA</region>

          <code>02138</code>

          <country>USA</country>
        </postal>

        <email>acaro@bbn.com</email>
      </address>
    </author>

    <author fullname="Paul D. Amer" initials="P.A" surname="Amer">
      <organization>University of Delaware</organization>

      <address>
        <postal>
          <street>Computer Science Department - 434 Smith Hall</street>

          <city>Newark</city>

          <region>DE</region>

          <code>19716-2586</code>

          <country>USA</country>
        </postal>

        <email>amer@udel.edu</email>
      </address>
    </author>

    <author fullname="Karen E. E. Nielsen" initials="K.N" surname="Nielsen">
      <organization>Ericsson</organization>

      <address>
        <postal>
          <street>Kistavägen 25</street>

          <city>Stockholm</city>

          <region/>

          <code>164 80</code>

          <country>Sweden</country>
        </postal>

        <email>karen.nielsen@tieto.com</email>
      </address>
    </author>

    <date/>

    <abstract>
      <t>One of the major advantages of SCTP is the support of multi-homed
      communication. A multi-homed SCTP end-point has the ability to withstand
      network failures by migrating the traffic from an inactive network to an
      active one. However, if the failover operation as specified in RFC4960
      is followed, there can be a significant delay in the migration to the
      active destination addresses, thus severely reducing the effectiveness
      of the SCTP failover operation.</t>

      <t>This document complements RFC4960 by the introduction of a new path
      state, the Potentially Failed (PF) path state, and an associated new
      failover operation to apply during a network failure. The algorithm
      defined is called SCTP Potentially Failed Algorithm, SCTP-PF for short.
      In addition, the document complements RFC4960 by introducing alternative
      switchover operation modes for the data transfer path management after
      the recovery of a failed primary path. These modes can allow
      improvements in the performance of the operation in some network
      environments. The implementation of the additional switchover operation
      modes is an optional part of SCTP-PF.</t>

      <t>The procedures defined in the document require only minimal
      modifications to the current specification. The procedures are
      sender-side only and do not impact the SCTP receiver.</t>
    </abstract>
  </front>

  <middle>
    <section title="Introduction">
      <t>The Stream Control Transmission Protocol (SCTP) as specified in <xref
      target="RFC4960"/> supports multihoming at the transport layer -- an
      SCTP endpoint can bind to multiple IP addresses. SCTP's multihoming
      features include failure detection and failover procedures to provide
      network interface redundancy and improved end-to-end fault
      tolerance.</t>

      <t>In SCTP's current failure detection procedure, the sender must
      experience Path.Max.Retrans (PMR) number of consecutive failed
      timer-based retransmissions on a destination address before detecting a
      path failure. The sender fails over to an alternate active destination
      address only after failure detection. Until detecting the failover, the
      sender continues to transmit data on the failed path, which degrades the
      SCTP performance. Concurrent Multipath Transfer (CMT) <xref
      target="IYENGAR06"/> is an proposed extension to SCTP that allows the
      sender to transmit data on multiple paths simultaneously. Research <xref
      target="NATARAJAN09"/> shows that the current failure detection
      procedure worsens CMT performance during failover and can be
      significantly improved by employing a better failover algorithm.</t>

      <t>This document specifies an alternative failure detection and failover
      procedure, the SCTP Potentially Failed algorithm, that improves the
      performance of SCTP multi-homed operation during a failover.</t>

      <t>For multi-homed SCTP the operation after the recovery of a failed
      path equally well impacts the performance of the protocol. With the
      procedures specified in <xref target="RFC4960"/>, SCTP will, after a
      failover from the primary path, switch back to the primary path for data
      transfer as soon as this path becomes available again. From a
      performance perspective, as confirmed in research <xref
      target="CARO02"/>, such a switchback of the data transmission path is
      not optimal in general. As an optional alternative to the switchback
      operation of <xref target="RFC4960"/>, this document specifies the
      Permanent Failover procedures proposed by <xref target="CARO02"/>.</t>

      <t>Additional discussion for alternative approaches that do not require
      modifications to <xref target="RFC4960"/>, as well as discussion of path
      bouncing effects that might be caused by frequent switchover, are
      provided in the Appendices.</t>

      <t>While the Potentially Failed algorithm primarily is motivated for
      improvement of the SCTP multi-homed operation, the feature applies also
      to SCTP single-homed operation. Here the algorithm serves to provide
      increased failure detection on idle associations, whereas the failover
      or switchback aspects of the algorithm will not be activated. This is
      discussed in more detail in Appendix C.</t>
    </section>

    <section title="Conventions and Terminology">
      <t>The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
      "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
      document are to be interpreted as described in <xref
      target="RFC2119"/>.</t>
    </section>

    <section anchor="SCTP_issues" title="Issues with the SCTP Path Management">
      <t>This section describes issues in the SCTP as specified in <xref
      target="RFC4960"/> to be fixed by the approach described in this
      document.</t>

      <t>An SCTP endpoint can support multiple IP addresses. Each SCTP
      endpoint exchanges the list of its usable addresses during the initial
      negotiation with its peer. Then the endpoints select one address from
      the peer's list and use this as the primary destination address. During
      normal transmission, an SCTP endpoint sends all user data to the primary
      destination address. Also, it sends packets containing a HEARTBEAT chunk
      to all idle destination addresses at a certain interval to check the
      reachability of these destination addresses. Idle destination addresses
      normally include all non-primary destination addresses.</t>

      <t>If a sender has multiple active destination addresses, it can
      retransmit data to an non-primary destination address, if the
      transmission to the primary times out.</t>

      <t>When a sender receives an acknowledgment for DATA or HEARTBEAT chunks
      sent to one of the destination addresses, it considers that destination
      address to be active and clears the error counter for the destination
      address. If it fails to receive acknowledgments, the error count for the
      destination address is increased. If the error counter exceeds the
      tunable protocol parameter Path.Max.Retrans (PMR), the SCTP endpoint
      considers the destination address to be inactive.</t>

      <t>The failover process of SCTP is initiated when the primary path
      becomes inactive (the error counter for the primary path exceeds
      Path.Max.Retrans). If the primary path is marked inactive, SCTP chooses
      a new destination address from one of the active destinations and starts
      using this as the destination address for sending data. If the primary
      path becomes active again, SCTP reverts to using the primary destination
      address for subsequent data transmissions and stop using the non-primary
      one.</t>

      <t>One issue with this failover process defined in <xref
      target="RFC4960"/> is that it usually takes a significant amount of time
      before SCTP switches to the new destination address. Let's say the
      primary path on a multi-homed host becomes unavailable and the RTO value
      for the primary path at that time is around 1 second, it usually takes
      over 60 seconds before SCTP starts to use the non-primary path for
      initial data transmission. This is because the recommended value for
      Path.Max.Retrans in the <xref target="RFC4960"/> is 5, which requires 6
      consecutive timeouts before the failover takes place. Before SCTP
      switches to the non-primary address, SCTP keeps trying to send packets
      to the primary address and only retransmitted packets are sent to the
      non-primary address and thus can be received by the receiver. This slow
      failover process can cause significant performance degradation and is
      not acceptable in some situations.</t>

      <t>Another issue with RFC4960 failover and switchback operation is that
      once the primary path becomes active again, the traffic is
      unconditionally switched back to use this path. This is not optimal in
      some situations. This is further discussed in <xref
      target="permanent_failover"/>.</t>
    </section>

    <section anchor="SCTP_PF"
             title="SCTP with Potentially-Failed Destination State (SCTP-PF)">
      <t>To address the issues described in <xref target="SCTP_issues"/>, this
      document extends SCTP path management scheme by adding the Potentially
      Failed state and associated protocol operation. The algorithm is called
      SCTP Potentially Failed algorithm. SCTP-PF for short. The resulting SCTP
      path management operation is called SCTP Potentially Failed
      operation.</t>

      <section title="SCTP-PF Concept">
        <t>The introduction of the Potentially Failed state stems from the
        following two observations about SCTP's failure detection procedure:
        <list style="symbols">
            <t>To minimize the performance impact during failover, the sender
            should avoid transmitting data to the failed destination address
            as early as possible. In the current SCTP path management scheme,
            the sender stops transmitting data to a destination address only
            after the destination address is marked Failed (inactive). Thus, a
            smaller PMR value is better because the sender can transition a
            destination address to the Failed (inactive) state quicker.</t>

            <t>Smaller PMR values increase the chances of spurious failure
            detection where the sender incorrectly marks a destination address
            as Failed (inactive) during periods of temporary congestion. As
            <xref target="RFC4960"/> recommends for a coupling of the PMR
            value and the protocol parameter Association.Max.Retrans (AMR)
            value such spurious failure detection risks to carry over to
            spurious association failure detection and closure. Larger PMR
            values are preferable to avoid spurious failure detection.</t>
          </list></t>

        <t>From the above observations it is clear that tuning the PMR value
        involves the following trade off -- a lower value improves performance
        but increases the chances of spurious failure detection, whereas a
        higher value degrades performance and reduces spurious failure
        detection in a wide range of path conditions. Thus, tuning the
        association's PMR value is an incomplete solution to address the
        performance impact during failure.</t>

        <t>SCTP-PF defined in this document introduces the new Potentially
        Failed (PF) destination address state in SCTP's path management
        procedure. The new Potentially Failed (PF) destination address state
        applies to SCTP single-homed operation as well as to SCTP multi-homed
        operation. The PF state was originally proposed to improve CMT
        performance <xref target="NATARAJAN09"/>. The PF state is an
        intermediate state between the Active and Failed states. SCTP's
        failure detection procedure is modified to include the PF state. The
        new failure detection algorithm assumes that loss detected by a
        timeout implies either severe congestion or failure en-route. After a
        number of consecutive timeouts on a path, the sender is unsure, and
        marks the corresponding destination address as in the PF state. A PF
        destination address is not used for data transmission except when it
        is the only destination address available (e.g., for single-homed
        SCTP) or in other special cases (discussed below). The new failure
        detection algorithm requires only sender-side changes.</t>
      </section>

      <section title="Specification of the SCTP-PF Algorithm">
        <t>The SCTP-PF operation is specified as follows: <list
            style="numbers">
            <t>The sender maintains a new tunable parameter called
            PotentiallyFailed.Max.Retrans (PFMR). The RECOMMENDED value of
            PFMR is 0 when SCTP-PF is used. The PFMR defines a new
            intermediate PF threshold on the destination address error counter
            at exceed of which the destination address is classified as PF and
            related PF state actions are to be taken. By standard RFC4960
            semantics a destination address is classified as Inactive once the
            error counter exceeds PMR. Setting PFMR larger to or equal to PMR
            does not result in definition of a PF threshold for the
            destination address. I.e., PFMR set larger to or equal to PMR
            means that the destination address never will be classified as
            PF.</t>

            <t>The error counter of an active destination address is
            incremented as specified in <xref target="RFC4960"/>. This means
            that the error counter of the destination address will be
            incremented each time the T3-rtx timer expires, or each time a
            HEARTBEAT chunk is sent when idle and not acknowledged within an
            RTO. When the value in the destination address error counter
            exceeds PFMR, the endpoint MUST mark the destination address as in
            the PF state.</t>

            <t>The PFMR threshold defines the point the destination address no
            longer is considered a good candidate for data transmission and a
            SCTP-PF sender SHOULD NOT send data to destination addresses in PF
            state when alternative destination addresses in active state are
            available. Specifically this means that: <list style="hanging">
                <t hangText="i">When there is outbound data to send and the
                destination address presently used for data transmission is in
                PF state, the sender SHOULD choose a destination address in
                active state, if one exists, and failover to deploy this
                destination address for data transmission.</t>

                <t hangText="ii">When retransmitting data that has timed out
                and the sender thus by <xref target="RFC4960"/>, section
                6.4.1, should attempt to pick a new destination address for
                data retransmission, the sender SHOULD choose an alternate
                destination transport address in active state if one
                exists.</t>

                <t hangText="iii">When there is outbound data to send and the
                SCTP user explicitly requests to send data to a destination
                address in PF state, the sender SHOULD send the data to an
                alternate destination address in active state if one
                exists.</t>
              </list>When choosing among multiple destination address in
            active state the following considerations are given: <list
                style="letters">
                <t>An SCTP sender should comply with [RFC4960], section 6.4.1,
                principles of choosing most divergent source-destination pairs
                compared with, for i.: the destination address in PF state
                that it performs a failover from, and for ii.: the destination
                address towards which the data timed out. Rules for picking
                the most divergent source-destination pair are an
                implementation decision and are not specified within this
                document.</t>

                <t>A SCTP-PF sender MAY choose to send data to a destination
                address in PF state, even if destination addresses in active
                state exist, have the SCTP-PF sender other means of
                information available that disqualifies the destination
                address in active state from being preferred. However, the
                discussion of such mechanisms is outside of the scope of the
                SCTP_PF operation specified in this document.</t>
              </list> In all cases, the sender MUST NOT change the state of
            chosen destination address, whether this state be active or PF,
            and it MUST NOT clear the error counter of the destination address
            as a result of choosing the destination address for data
            transmission.</t>

            <t>When the destination addresses are all in PF state or some in
            PF state and some in inactive state, the sender MUST choose one
            destination address in PF state and transmit or retransmit data to
            this destination address using the following rules: <list
                style="letters">
                <t>The sender SHOULD choose the destination in PF state with
                the lowest error count (fewest consecutive timeouts) for data
                transmission and transmit or retransmit data to this
                destination.</t>

                <t>When there are multiple PF destinations with same error
                count, the sender should let the choice among the multiple PF
                destination with equal error count be based on the <xref
                target="RFC4960"/>, section 6.4.1, principles of choosing most
                divergent source-destination pairs when executing (potentially
                consecutive) retransmission. Rules for picking the most
                divergent source-destination pair are an implementation
                decision and are not specified within this document.</t>

                <t>A sender MAY choose to deploy other strategies than the
                above when choosing among multiple PF destinations have the
                SCTP-PF sender other means of information available that
                qualifies a particular destination address for being used. The
                SCTP-PF protocol operation specified in this document makes no
                assumption of the existence of such other means of information
                and specifies for the above as the default operation of an
                SCTP-PF sender.</t>
              </list> The sender MUST NOT change the state and the error
            counter of any destination address regardless of whether it has
            been chosen for transmission or not.</t>

            <t>HEARTBEAT chunks MUST be send to PF destination addresses
            regardless of whether the Path Heartbeat function (Section 8.3 of
            <xref target="RFC4960"/>) is enabled for the destination address
            or not. The HB.interval of the Path Heartbeat function of <xref
            target="RFC4960"/> MUST be ignored for destination addresses in PF
            state, instead HEARTBEAT chunks are sent to destination addresses
            in PF state once per RTO. The HEARTBEAT sending begins upon that a
            destination address reaches the PF state. When a HEARTBEAT chunk
            is not acknowledged within the RTO, the sender increments the
            error counter and exponentially back off the RTO value. If the
            error counter is less than PMR, the sender transmits another
            packet containing the HEARTBEAT chunk immediately after timeout
            expiration on the previous HEARTBEAT. When data is being
            transmitted to a destination address in the PF state, the
            transmission of a HEARTBEAT chunk MAY be omitted in case receipt
            of a SACK of or a T3-rtx timer expiration on the outstanding data
            can provide equivalent information. Likewise the timeout of a
            HEARTBEAT chunk MAY be ignored if data is outstanding towards the
            destination address.</t>

            <t>When the sender receives a HEARTBEAT ACK from a destination
            address in PF state, the sender MUST clear the error counter of
            the destination address and transition the destination address
            back to active state. When the sender resumes data transmission on
            the destination address, it MUST do this following the
            prescriptions of Section 7.2 of <xref target="RFC4960"/>.</t>

            <t>Additional (PMR - PFMR) consecutive timeouts on a destination
            address in PF state confirm the path failure, upon which the
            destination address transitions to the inactive state. As
            described in <xref target="RFC4960"/>, the sender (i) SHOULD
            notify the ULP about this state transition, and (ii) transmit
            HEARTBEAT chunks to the inactive destination address at a lower
            frequency as described in Section 8.3 of <xref target="RFC4960"/>
            (when this function is enabled for the destination address).</t>

            <t>Acknowledgments for chunks that have been transmitted to
            multiple destinations (i.e., a chunk which has been retransmitted
            to a different destination address than the destination address to
            which the chunk was first transmitted) MUST NOT clear the error
            count for an inactive destination address and MUST NOT transition
            a PF destination address back to active state, since a sender
            cannot disambiguate whether the ACK was for the original
            transmission or the retransmission(s). The same ambiguity concerns
            the related congestion window growth. The bytes of a newly
            acknowledged chunk which has been transmitted to multiple
            destination addresses SHOULD be considered for contribution to the
            congestion window growth towards the destination address where the
            chunk was last sent. The contribution of the ACKed bytes to the
            window growth is subject to the prescriptions described in Section
            7.2 of <xref target="RFC4960"/> is fulfilled. A SCTP sender MAY
            apply a different approach for both the error count handling and
            the congestion control growth handling based on unequivocally
            information on which destination (including multiple destination
            addresses) the chunk reached. This document makes no reference to
            what such unequivocally information could consist of, neither how
            such unequivocally information could be obtained. The design of
            such an alternative approach is left to implementations.</t>

            <t>Acknowledgments for chunks that has been transmitted to one
            destination address only MUST clear the error counter for the
            destination address and MUST transition a PF destination address
            back to Active state. This situation can happen when new data is
            sent to a destination address in the PF state. It can also happen
            in situations where the destination address is in the PF state due
            to the occurrence of a spurious T3-rtx timer and Acknowledgments
            start to arrive for data sent prior to occurrence of the spurious
            T3-rtx and data has not yet been retransmitted towards other
            destinations. This document does not specify special handling for
            detection of or reaction to spurious T3-rtx timeouts, e.g., for
            special operation vis-a-vis the congestion control handling or
            data retransmission operation towards a destination address which
            undergoes a transition from active to PF to active state due to a
            spurious T3-rtx timeout. But it is noted that this is an area
            which would benefit from additional attention, experimentation and
            specification for Single Homed SCTP as well as for Multi Homed
            SCTP protocol operation.</t>

            <t>The SCTP stack SHOULD provide the ULP with the means to expose
            the PF state of its destinations as well as the means to notify
            the state transitions from Active to PF, and vice-versa. When
            doing this, such an SCTP stack MUST provide the ULP with the means
            to suppress exposure of PF state and associated state transitions
            as well.</t>
          </list></t>

        <section anchor="dormant" title="Dormant State Operation">
          <t>In a situation with complete disruption of the communication in
          between the SCTP Endpoints, the aggressive HEARTBEAT transmissions
          of SCTP-PF on destination addresses in PF state may make the
          association enter dormant state faster than a standard <xref
          target="RFC4960"/> SCTP implementation given the same setting of
          Path.Max.Retrans (PMR) and Association.Max.Retrans (AMR). For
          example, an SCTP association with two destination addresses
          typically would reach dormant state in half the time of an <xref
          target="RFC4960"/> SCTP implementation in such situations. This is
          because a SCTP PF sender will send HEARTBEATS and data
          retransmissions in parallel with RTO intervals when there are
          multiple destinations addresses in PF state. This argument pressumes
          that RTO << HB.interval of <xref target="RFC4960"/>. One could
          use higher values of PMR, which makes the dormant state situations
          less likely to happen. The downside of increasing the PMR value is
          that destination address failure detections and notifications of
          such events to ULP is weakened.</t>

          <t>A design goal of SCTP-PF is that it should provide the same level
          of disruption tolerance as an <xref target="RFC4960"/> SCTP
          implementation with the same Path.Max.Retrans (PMR) and
          Association.Max.Retrans (AMR) setting. For this reason, SCTP-PF
          SHOULD perform the following operations during dormant state, while
          this is an implementation decision in <xref target="RFC4960"/>.
          <list style="letters">
              <t>When the destination addresses are all in inactive state, the
              sender MUST choose one destination when data is transmitted. The
              sender MUST NOT change the state and the error counter of any
              destination address regardless of whether it has been chosen for
              transmission or not.</t>

              <t>The sender SHOULD choose the destination in inactive state
              with the lowest error count (fewest consecutive timeouts) for
              data transmission. When there are multiple destinations with
              same error count in inactive state, the sender SHOULD attempt to
              pick the most divergent source - destination pair from the last
              source - destination pair where failure was observed. Rules for
              picking the most divergent source-destination pair are an
              implementation decision and are not specified within this
              document. To support differentiation of inactive destination
              addresses based on their error count SCTP will need to allow for
              increment of the destination address error counters up to some
              reasonable limit above PMR+1, thus changing the prescriptions of
              <xref target="RFC4960"/>, section 8.3, in this respect. The
              exact limit to apply is not specified in this document but it is
              considered reasonable to require for such to be an order of
              magnitude higher than the PMR value. A sender MAY choose to
              deploy other strategies that the strategy defined by here. The
              strategy to prioritize the last active destination address,i.e.,
              the destination address with the fewest error counts is optimal
              when some paths are permanently inactive, but suboptimal when a
              path instability is transient.</t>
            </list></t>

          <t>An SCTP-PF implementation MAY keep the operation during dormant
          state an implementation decision, but it should be careful not to
          compromise the fault tolerance of the SCTP operation.</t>

          <t>The above prescriptions for SCTP-PF dormant state handling SHOULD
          NOT be coupled to the value of the PFMR, but solely to the
          activation of SCTP-PF logic in an SCTP implementation. It is further
          noted that also a standard <xref target="RFC4960"/> SCTP
          implementation can use this mode of operation to improve the fault
          tolerance (which some implementations already do).</t>
        </section>
      </section>

      <section anchor="permanent_failover" title="Permanent Failover">
        <t>This section describes an OPTIONAL switchback feature called
        Permanent Failover which is beneficiary to deploy in certain
        situations.</t>

        <section title="Background">
          <t>In <xref target="RFC4960"/>, an SCTP sender migrates the traffic
          back to the original primary destination address once this address
          becomes active again. As the CWND towards the original primary
          destination address has to be rebuilt once data transfer resumes,
          the switch back to use the original primary address is not always
          optimal. Indeed <xref target="CARO02"/> shows that the switch back
          to the original primary may degrade SCTP performance compared to
          continuing data transmission on the same path, especially, but not
          only, in scenarios where this path's characteristics are better. In
          order to mitigate this performance degradation, the Permanent
          Failover operation was proposed in <xref target="CARO02"/>. When
          SCTP changes the destination address due to failover, Permanent
          Failover operation allows SCTP sender to continue data transmission
          on the new working path even when the old primary destination
          address becomes active again. This is achieved by having SCTP
          perform a switch over of the primary path to the alternative working
          path rather than having SCTP switch back data transfer to the
          (previous) primary path.</t>

          <t>The manner of switch over operation that is most optimal in a
          given scenario depends on the relative quality of a set primary path
          versus the quality of alternative paths available as well as it
          depends on the extent to which it is desired for the mode of
          operation to enforce traffic distribution over a number of network
          paths. I.e., load distribution of traffic from multiple SCTP
          associations may be sought to be enforced by distribution of the set
          primary paths with <xref target="RFC4960"/> switchback operation.
          However as <xref target="RFC4960"/> switchback behavior is
          suboptimal in certain situations, especially in scenarios where a
          number of equally good paths are available, it is recommended for
          SCTP to support also, as alternative behavior, the Permanent
          Failover switch over modes of operation.</t>
        </section>

        <section title="Permanent Failover Algorithm">
          <t>The Permanent Failover operation requires only sender side
          changes. The details are:</t>

          <t><list style="numbers">
              <t>The sender maintains a new tunable parameter, called
              Primary.Switchover.Max.Retrans (PSMR). The PSMR MUST be set
              greater or equal to the PFMR value. Implementations MUST reject
              any other values of PSMR.</t>

              <t>When the path error counter on a set primary path exceeds
              PSMR, the SCTP implementation MUST autonomously select and set a
              new primary path.</t>

              <t>The primary path selected by the SCTP implementation MUST be
              the path which at the given time would be chosen for data
              transfer. A previously failed primary path can be used as data
              transfer path as per normal path selection when the present data
              transfer path fails.</t>

              <t>The recommended value of PSMR is PFMR when Permanent Failover
              is used. This means that no forced switchback to a previously
              failed primary path is performed. An implementation of Permanent
              Failover MUST support the setting of PSMR = PFMR. An
              implementation of Permanent Failover MAY support setting of PSMR
              > PFMR.</t>

              <t>It MUST be possible to disable the Permanent Failover and
              obtain the standard switchback operation of <xref
              target="RFC4960"/>.</t>
            </list></t>

          <t>To support optimal operation in a wider range of network
          scenarios, it it proposed for an SCTP-PF implementation to implement
          Permanent Failover operation as an optional feature. The
          implementation of the Permanent Failover feature is optional for an
          SCTP-PF implementation. For an SCTP implementation that implements
          Permanent Failover, this specification RECOMMENDS that the standard
          RFC4960 switchback operation is retained as the default
          operation.</t>
        </section>
      </section>
    </section>

    <section title="Socket API Considerations">
      <t>This section describes how the socket API defined in <xref
      target="RFC6458"/> is extended to provide a way for the application to
      control and observe the SCTP-PF behavior.</t>

      <t>Please note that this section is informational only.</t>

      <t>A socket API implementation based on <xref target="RFC6458"/> is, by
      means of the existing SCTP_PEER_ADDR_CHANGE event, extended to provide
      the event notification when a peer address enters or leaves the
      potentially failed state as well as the socket API implementation is
      extended to expose the potentially failed state of a peer address in the
      existing SCTP_GET_PEER_ADDR_INFO structure.</t>

      <t>Furthermore, two new read/write socket options for the level
      IPPROTO_SCTP and the name SCTP_PEER_ADDR_THLDS and
      SCTP_EXPOSE_POTENTIALLY_FAILED_STATE are defined as described below. The
      first socket option is used to control the values of the PFMR and PSMR
      parameters described in <xref target="SCTP_PF"/>. The second one
      controls the exposition of the potentially failed path state.</t>

      <t>Support for the SCTP_PEER_ADDR_THLDS and
      SCTP_EXPOSE_POTENTIALLY_FAILED_STATE socket options need also to be
      added to the function sctp_opt_info().</t>

      <section anchor="pf_support_api"
               title="Support for the Potentially Failed Path State">
        <t>As defined in <xref target="RFC6458"/>, the SCTP_PEER_ADDR_CHANGE
        event is provided if the status of a peer address changes. In addition
        to the state changes described in <xref target="RFC6458"/>, this event
        is also provided, if a peer address enters or leaves the potentially
        failed state. The notification as defined in <xref target="RFC6458"/>
        uses the following structure:</t>

        <figure>
          <artwork><![CDATA[
struct sctp_paddr_change {
  uint16_t spc_type;
  uint16_t spc_flags;
  uint32_t spc_length;
  struct sockaddr_storage spc_aaddr;
  uint32_t spc_state;
  uint32_t spc_error;
  sctp_assoc_t spc_assoc_id;
}
]]></artwork>
        </figure>

        <t><xref target="RFC6458"/> defines the constants SCTP_ADDR_AVAILABLE,
        SCTP_ADDR_UNREACHABLE, SCTP_ADDR_REMOVED, SCTP_ADDR_ADDED, and
        SCTP_ADDR_MADE_PRIM to be provided in the spc_state field. This
        document defines in addition to that the new constant
        SCTP_ADDR_POTENTIALLY_FAILED, which is reported if the affected
        address becomes potentially failed.</t>

        <t>The SCTP_GET_PEER_ADDR_INFO socket option defined in <xref
        target="RFC6458"/> can be used to query the state of a peer address.
        It uses the following structure:</t>

        <figure>
          <artwork><![CDATA[
struct sctp_paddrinfo {
  sctp_assoc_t spinfo_assoc_id;
  struct sockaddr_storage spinfo_address;
  int32_t spinfo_state;
  uint32_t spinfo_cwnd;
  uint32_t spinfo_srtt;
  uint32_t spinfo_rto;
  uint32_t spinfo_mtu;
};
]]></artwork>
        </figure>

        <t><xref target="RFC6458"/> defines the constants SCTP_UNCONFIRMED,
        SCTP_ACTIVE, and SCTP_INACTIVE to be provided in the spinfo_state
        field. This document defines in addition to that the new constant
        SCTP_POTENTIALLY_FAILED, which is reported if the peer address is
        potentially failed.</t>
      </section>

      <section title="Peer Address Thresholds (SCTP_PEER_ADDR_THLDS) Socket Option">
        <t>Applications can control the SCTP-PF behavior by getting or setting
        the number of consecutive timeouts before a peer address is considered
        potentially failed or unreachable and before the primary path is
        changed automatically. This socket option uses the level IPPROTO_SCTP
        and the name SCTP_PEER_ADDR_THLDS.</t>

        <t>The following structure is used to access and modify the
        thresholds:</t>

        <figure>
          <artwork><![CDATA[
struct sctp_paddrthlds {
  sctp_assoc_t spt_assoc_id;
  struct sockaddr_storage spt_address;
  uint16_t spt_pathmaxrxt;
  uint16_t spt_pathpfthld;
  uint16_t spt_pathcpthld;
};
]]></artwork>
        </figure>

        <t><list style="hanging">
            <t hangText="spt_assoc_id:">This parameter is ignored for
            one-to-one style sockets. For one-to-many style sockets the
            application may fill in an association identifier or
            SCTP_FUTURE_ASSOC. It is an error to use SCTP_{CURRENT|ALL}_ASSOC
            in spt_assoc_id.</t>

            <t hangText="spt_address:">This specifies which peer address is of
            interest. If a wild card address is provided, this socket option
            applies to all current and future peer addresses.</t>

            <t hangText="spt_pathmaxrxt:">Each peer address of interest is
            considered unreachable, if its path error counter exceeds
            spt_pathmaxrxt.</t>

            <t hangText="spt_pathpfthld:">Each peer address of interest is
            considered potentially failed, if its path error counter exceeds
            spt_pathpfthld.</t>

            <t hangText="spt_pathcpthld:">Each peer address of interest is not
            considered the primary remote address anymore, if its path error
            counter exceeds spt_pathcpthld. Using a value of 0xffff disables
            the selection of a new primary peer address. If an implementation
            does not support the automatically selection of a new primary
            address, it should indicate an error with errno set to EINVAL if a
            value different from 0xffff is used in spt_pathcpthld. Setting of
            spt_pathcpthld < spt_pathpfthld should be rejected with errno
            set to EINVAL. An implementation MAY support only setting of
            spt_pathcpthld = spt_pathpfthld and spt_pathcpthld = 0xffff. In
            this case it shall reject setting of other values with errno set
            to EINVAL.</t>
          </list></t>
      </section>

      <section title="Exposing the Potentially Failed Path State                       (SCTP_EXPOSE_POTENTIALLY_FAILED_STATE) Socket Option">
        <t>Applications can control the exposure of the potentially failed
        path state in the SCTP_PEER_ADDR_CHANGE event and the
        SCTP_GET_PEER_ADDR_INFO as described in <xref
        target="pf_support_api"/>. The default value is implementation
        specific.</t>

        <t>This socket option uses the level IPPROTO_SCTP and the name
        SCTP_EXPOSE_POTENTIALLY_FAILED_STATE.</t>

        <t>The following structure is used to control the exposition of the
        potentially failed path state:</t>

        <figure>
          <artwork><![CDATA[
struct sctp_assoc_value {
  sctp_assoc_t assoc_id;
  uint32_t assoc_value;
};
]]></artwork>
        </figure>

        <t><list style="hanging">
            <t hangText="assoc_id:">This parameter is ignored for one-to-one
            style sockets. For one-to-many style sockets the application may
            fill in an association identifier or SCTP_FUTURE_ASSOC. It is an
            error to use SCTP_{CURRENT|ALL}_ASSOC in assoc_id.</t>

            <t hangText="assoc_value:">The potentially failed path state is
            exposed if and only if this parameter is non-zero.</t>
          </list></t>
      </section>
    </section>

    <section title="Security Considerations">
      <t>Security considerations for the use of SCTP and its APIs are
      discussed in <xref target="RFC4960"/> and <xref target="RFC6458"/>. The
      logic described here is for sender-side only enabled by configuration
      and does not have any impacts on protocol messages on the wire. No new
      chunk type or new field parameter is not required in this document.</t>
    </section>

    <section title="IANA Considerations">
      <t>This document does not create any new registries or modify the rules
      for any existing registries managed by IANA.</t>
    </section>

    <section title="Proposed Change of Status (to be Deleted before Publication)">
      <t>Initially this work looked to entail some changes of the Congestion
      Control (CC) operation of SCTP and for this reason the work was proposed
      as Experimental. These intended changes of the CC operation have since
      been judged to be irrelevant and are no longer part of the
      specification. As the specification entails no other potential harmful
      features, consensus exists in the WG to bring the work forward as
      PS.</t>

      <t>Initially concerns have been expressed about the possibility for the
      mechanism to introduce path bouncing with potential harmful network
      impacts. These concerns are believed to be unfounded. This issue is
      addressed in Appendix B.</t>

      <t>It is noted that the feature specified by this document is
      implemented by multiple SCTP SW implementations and furthermore that
      various variants of the solution have been deployed in Telco signaling
      environments for several years with good results.</t>
    </section>
  </middle>

  <back>
    <references title="Normative References">
      <?rfc include="reference.RFC.2119" ?>

      <?rfc include="reference.RFC.4960" ?>
    </references>

    <references title="Informative References">
      <?rfc include="reference.RFC.6458" ?>

      <reference anchor="IYENGAR06" target="">
        <front>
          <title>Concurrent Multipath Transfer using SCTP Multihoming over
          Independent End-to-end Paths.</title>

          <author fullname="" initials="J." surname="Iyengar">
            <organization/>
          </author>

          <author fullname="" initials="P." surname="Amer">
            <organization/>
          </author>

          <author fullname="" initials="R." surname="Stewart">
            <organization/>
          </author>

          <date month="10" year="2006"/>
        </front>

        <seriesInfo name="IEEE/ACM Trans on Networking" value="14(5)"/>
      </reference>

      <reference anchor="NATARAJAN09" target="">
        <front>
          <title>Concurrent Multipath Transfer during Path Failure</title>

          <author fullname="" initials="P." surname="Natarajan">
            <organization/>
          </author>

          <author fullname="" initials="N." surname="Ekiz">
            <organization/>
          </author>

          <author fullname="" initials="P." surname="Amer">
            <organization/>
          </author>

          <author fullname="" initials="R." surname="Stewart">
            <organization/>
          </author>

          <date month="5" year="2009"/>
        </front>

        <seriesInfo name="Computer Communications" value=""/>
      </reference>

      <reference anchor="JUNGMAIER02" target="">
        <front>
          <title>On the use of SCTP in failover scenarios</title>

          <author fullname="" initials="A." surname="Jungmaier">
            <organization/>
          </author>

          <author fullname="" initials="E." surname="Rathgeb">
            <organization/>
          </author>

          <author fullname="" initials="M." surname="Tuexen">
            <organization/>
          </author>

          <date month="7" year="2002"/>
        </front>

        <seriesInfo name="World Multiconference on Systemics, Cybernetics and Informatics"
                    value=""/>
      </reference>

      <reference anchor="GRINNEMO04" target="">
        <front>
          <title>Performance of SCTP-controlled failovers in M3UA-based
          SIGTRAN networks</title>

          <author fullname="" initials="K-J" surname="Grinnemo">
            <organization/>
          </author>

          <author fullname="" initials="A." surname="Brunstrom">
            <organization/>
          </author>

          <date month="4" year="2004"/>
        </front>

        <seriesInfo name="Advanced Simulation Technologies Conference"
                    value=""/>
      </reference>

      <reference anchor="FALLON08" target="">
        <front>
          <title>SCTP Switchover Performance Issues in WLAN
          Environments</title>

          <author fullname="" initials="S." surname="Fallon">
            <organization/>
          </author>

          <author fullname="" initials="P." surname="Jacob">
            <organization/>
          </author>

          <author fullname="" initials="Y." surname="Qiao">
            <organization/>
          </author>

          <author fullname="" initials="L." surname="Murphy">
            <organization/>
          </author>

          <author fullname="" initials="E." surname="Fallon">
            <organization/>
          </author>

          <author fullname="" initials="A." surname="Hanley">
            <organization/>
          </author>

          <date month="1" year="2008"/>
        </front>

        <seriesInfo name="IEEE CCNC" value="2008"/>
      </reference>

      <reference anchor="CARO04" target="">
        <front>
          <title>End-to-End Failover Thresholds for Transport Layer
          Multihoming</title>

          <author fullname="" initials="A." surname="Caro Jr.">
            <organization/>
          </author>

          <author fullname="" initials="P." surname="Amer">
            <organization/>
          </author>

          <author fullname="" initials="R." surname="Stewart">
            <organization/>
          </author>

          <date month="11" year="2004"/>
        </front>

        <seriesInfo name="MILCOM 2004" value=""/>
      </reference>

      <reference anchor="CARO05" target="">
        <front>
          <title>End-to-End Fault Tolerance using Transport Layer
          Multihoming</title>

          <author fullname="" initials="A." surname="Caro Jr.">
            <organization/>
          </author>

          <date month="1" year="2005"/>
        </front>

        <seriesInfo name="Ph.D Thesis, University of Delaware" value=""/>
      </reference>

      <reference anchor="CARO02" target="">
        <front>
          <title>A Two-level Threshold Recovery Mechanism for SCTP</title>

          <author fullname="" initials="A." surname="Caro Jr.">
            <organization/>
          </author>

          <author fullname="" initials="J." surname="Iyengar">
            <organization/>
          </author>

          <author fullname="" initials="P." surname="Amer">
            <organization/>
          </author>

          <author fullname="" initials="G." surname="Heinz">
            <organization/>
          </author>

          <author fullname="" initials="R." surname="Stewart">
            <organization/>
          </author>

          <date month="7" year="2002"/>
        </front>

        <seriesInfo name="Tech report, CIS Dept, University of Delaware"
                    value=""/>
      </reference>
    </references>

    <section anchor="alternative_approach"
             title="Discussions of Alternative Approaches">
      <t>This section lists alternative approaches for the issues described in
      this document. Although these approaches do not require to update
      RFC4960, we do not recommend them from the reasons described below.</t>

      <section title="Reduce Path.Max.Retrans (PMR)">
        <t>Smaller values for Path.Max.Retrans shorten the failover duration.
        In fact, this is recommended in some research results <xref
        target="JUNGMAIER02"/> <xref target="GRINNEMO04"/> <xref
        target="FALLON08"/>. For example, if when Path.Max.Retrans=0, SCTP
        switches to another destination address on a single timeout. This
        smaller value for Path.Max.Retrans can results in spurious failover,
        which might be a problem.</t>

        <t>Unlike SCTP-PF, the interval for heartbeat packets is governed by
        'HB.interval' even during failover process. 'HB.interval' is usually
        set in the order of seconds (recommended value is 30 seconds). When
        the primary path becomes inactive, the next HEARTBEAT can be
        transmitted only seconds later. Meanwhile, the primary path may have
        recovered. In such situations, post failover, an endpoint is forced to
        wait on the order of seconds before the endpoint can resume
        transmission on the primary path. However, using smaller value for
        'HB.interval' might help this situation, but it will be the waste of
        bandwidth in most cases.</t>

        <t>In addition, smaller Path.Max.Retrans values also affect
        'Association.Max.Retrans' values. When the SCTP association's error
        count (sum of error counts on all ACTIVE paths) exceeds
        Association.Max.Retrans threshold, the SCTP sender considers the peer
        endpoint unreachable and terminates the association. Therefore,
        Section 8.2 in <xref target="RFC4960"/> recommends that
        Association.Max.Retrans value should not be larger than the summation
        of the Path.Max.Retrans of each of the destination addresses, else the
        SCTP sender considers its peer reachable even when all destinations
        are INACTIVE. To avoid such inconsistent behavior an SCTP
        implementation SHOULD reduce Association.Max.Retrans accordingly
        whenever it reduces Path.Max.Retrans. However, smaller
        Association.Max.Retrans value increases chances of association
        termination during minor congestion events.</t>
      </section>

      <section title="Adjust RTO related parameters">
        <t>As several research results indicate, we can also shorten the
        duration of failover process by adjusting RTO related parameters <xref
        target="JUNGMAIER02"/> <xref target="FALLON08"/>. During failover
        process, RTO keeps being doubled. However, if we can choose smaller
        value for RTO.max, we can stop the exponential growth of RTO at some
        point. Also, choosing smaller values for RTO.initial or RTO.min can
        contribute to keep RTO value small.</t>

        <t>Similar to reducing Path.Max.Retrans, the advantage of this
        approach is that it requires no modification to the current
        specification, although it needs to ignore several recommendations
        described in the Section 15 of <xref target="RFC4960"/>. However, this
        approach requires to have enough knowledge about the network
        characteristics between end points. Otherwise, it can introduce
        adverse side-effects such as spurious timeouts.</t>
      </section>
    </section>

    <section anchor="path_bouncing"
             title="Discussions for Path Bouncing Effect">
      <t>The methods described in the document can accelerate the failover
      process. Hence, they might introduce the path bouncing effect where the
      sender keeps changing the data transmission path frequently. This sounds
      harmful to the data transfer, however several research results indicate
      that there is no serious problem with SCTP in terms of path bouncing
      effect <xref target="CARO04"/> <xref target="CARO05"/>.</t>

      <t>There are two main reasons for this. First, SCTP is basically
      designed for multipath communication, which means SCTP maintains all
      path related parameters (CWND, ssthresh, RTT, error count, etc) per each
      destination address. These parameters cannot be affected by path
      bouncing. In addition, when SCTP migrates the data transfer to another
      path, it starts with the minimal or the initial CWND. Hence, there is
      little chance for packet reordering or duplicating.</t>

      <t>Second, even if all communication paths between the end-nodes share
      the same bottleneck, the SCTP-PF results in a behavior already allowed
      by <xref target="RFC4960"/>.</t>
    </section>

    <section anchor="sh" title="SCTP-PF for SCTP Single-homed Operation ">
      <t>For a single-homed SCTP association the only tangible effect of the
      activation of SCTP-PF operation is enhanced failure detection in terms
      of potential notification of the PF state of the sole destination
      address as well as, for idle associations, more rapid entering, and
      notification, of inactive state of the destination address and more
      rapid end-point failure detection. It is believed that neither of these
      effects are harmful, provided adequate dormant state operation is
      implemented, and furthermore that they may be particularly useful for
      applications that deploys multiple SCTP associations for load balancing
      purposes. The early notification of the PF state may be used for
      preventive measures as the entering of the PF state can be used as a
      warning of potential congestion. Depending on the PMR value, the
      aggressive HEARTBEAT transmission in PF state may speed up the end-point
      failure detection (exceed of AMR threshold on the sole path error
      counter) on idle associations in case where relatively large HB.interval
      value compared to RTO (e.g. 30secs) is used.</t>
    </section>
  </back>
</rfc>

PAFTECH AB 2003-20262026-04-24 03:00:17