[Search] [txt|pdf|bibtex] [Tracker] [WG] [Email] [Diff1] [Diff2] [Nits]

Versions: 00 01 02                                                      
Network Working Group                                             P. Lei
Internet-Draft                                             Cisco Systems
Expires: April 9, 2006                                         P. Conrad
                                                  University of Delaware
                                                         October 6, 2005

              Services Provided By Reliable Server Pooling

Status of this Memo

   By submitting this Internet-Draft, each author represents that any
   applicable patent or other IPR claims of which he or she is aware
   have been or will be disclosed, and any of which he or she becomes
   aware will be disclosed, in accordance with Section 6 of BCP 79.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-

   Internet-Drafts are draft documents valid for a maximum of six months
   and may be updated, replaced, or obsoleted by other documents at any
   time.  It is inappropriate to use Internet-Drafts as reference
   material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at

   The list of Internet-Draft Shadow Directories can be accessed at

   This Internet-Draft will expire on April 9, 2006.

Copyright Notice

   Copyright (C) The Internet Society (2005).


   The Reliable Server Pooling architecture (abbreviated "RSerPool", and
   defined in [3]), provides a set of services and protocols for
   building fault tolerant and highly available client/server
   applications.  This memo describes the semantics of the services that
   RSerPool provides to upper layer protocols.

Lei & Conrad              Expires April 9, 2006                 [Page 1]

Internet-Draft        Services Provided by RSerPool         October 2005

Table of Contents

   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  3
   2.  Conventions Used In This Document  . . . . . . . . . . . . . .  4
   3.  Example Application Scenarios  . . . . . . . . . . . . . . . .  4
     3.1.  Example Scenario for Failover Without RSerPool . . . . . .  4
     3.2.  Example Scenario Using RSerPool Basic Mode . . . . . . . .  5
     3.3.  Example Scenario Using RSerPool Enhanced Mode  . . . . . .  7
   4.  Service Primitives . . . . . . . . . . . . . . . . . . . . . .  8
     4.1.  Initialization . . . . . . . . . . . . . . . . . . . . . .  9
     4.2.  PE Registration Services . . . . . . . . . . . . . . . . .  9
     4.3.  PE Selection Services  . . . . . . . . . . . . . . . . . .  9
     4.4.  RSerPool Managed Data Channel  . . . . . . . . . . . . . . 10
     4.5.  Failover Services  . . . . . . . . . . . . . . . . . . . . 11
       4.5.1.  State Cookie Exchange  . . . . . . . . . . . . . . . . 11
       4.5.2.  Failover Callback Function . . . . . . . . . . . . . . 11
       4.5.3.  Business Card  . . . . . . . . . . . . . . . . . . . . 13
   5.  Transport Mappings . . . . . . . . . . . . . . . . . . . . . . 13
     5.1.  Defined Transport Mappings . . . . . . . . . . . . . . . . 13
     5.2.  Transport Mappings Requirements  . . . . . . . . . . . . . 14
       5.2.1.  Mappings: Mandatory Requirements . . . . . . . . . . . 14
       5.2.2.  Mappings: Optional Requirements  . . . . . . . . . . . 14
       5.2.3.  Mappings: Other Requirements . . . . . . . . . . . . . 15
   6.  Security Considerations  . . . . . . . . . . . . . . . . . . . 15
   7.  IANA Considerations  . . . . . . . . . . . . . . . . . . . . . 15
   8.  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 15
   9.  References . . . . . . . . . . . . . . . . . . . . . . . . . . 16
   Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 17
   Intellectual Property and Copyright Statements . . . . . . . . . . 18

Lei & Conrad              Expires April 9, 2006                 [Page 2]

Internet-Draft        Services Provided by RSerPool         October 2005

1.  Introduction

   The Reliable Server Pooling architecture is defined in [3].  The
   central idea of this architecture is to provide client applications
   ("pool users") with the ability to select a server (a "pool element")
   from among a group of servers providing equivalent service (a
   "pool").  The pool is accessed via an identifier called a "pool
   handle".  The RSerPool architecture supports high-availabilty and
   load balancing by enabling a pool user to identify the most
   appropriate server from the server pool at a given time.  The
   architecture also supports failover to an alternate server when

   This memo describes how an upper layer protocol or application for a
   pool user or pool element uses the RSerPool architecture and
   protocols.  Specifically, it describes how the ASAP protocol [5] and
   transport protocols (SCTP, TCP, etc.) can be utilized to realize
   highly available services between pool users and pool elements.

   The purpose of this document is to describe:

   1.  the precise services provided by RSerPool to the upper layer,

   2.  the tradeoffs in choosing which services to utilize,

   3.  how applications must be designed for each of these services,

   4.  how applications written over various transports (SCTP, TCP, and
       others) can be mapped into these services.

   RSerPool services can be used in one of two modes: "Basic Mode" and
   "Enhanced Mode".  Basic Mode provides a smaller set of services than
   Enhanced Mode, but offers imposes fewer restrictions on the
   application layer protocols that can be supported.  Enhanced Mode
   provides extra capabilities, including some features that require
   applications to exchange application data messages via RSerPool
   service primitives (a restriction not present in Basic Mode).

   For Enhanced Mode, the RSerPool data exchange primitives are
   implemented by multiplexing the ASAP messages and application data
   over a single transport protocol connection or association.  This
   memo defines how to do this multiplexing over SCTP.  This memo also
   describes the requirments needed to extend support to other transport
   protocols as required.

   Note that while RSerPool services are divided into Basic and Enhanced
   Modes, both modes assume a full implementation of the ASAP protocol.
   The purpose of dividing RSerPool services into two modes is solely to

Lei & Conrad              Expires April 9, 2006                 [Page 3]

Internet-Draft        Services Provided by RSerPool         October 2005

   provide more flexibility for applications to interact with RSerPool.
   In particular, Basic Mode provides an easy migration path for legacy
   applications to take advantage of many useful RSerPool services,
   including load balancing and high availability.  Enhanced Mode
   extends the services provided by Basic Mode with enhanced failover

2.  Conventions Used In This Document

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   document are to be interpreted as described in RFC 2119 [2].

3.  Example Application Scenarios

   To illustrate the differences between Basic and Enhanced Mode, this
   section decribes example failover scenarios for:

      an application written directly over a transport layer protocol,
      not utilizing RSerPool

      an application using RSerPool Basic Mode, and

      an application using RSerPool Enhanced Mode.

3.1.  Example Scenario for Failover Without RSerPool

   Consider a typical client/server application that does not use a
   reliable server pooling framework of any kind.  Typically, the server
   is specified by a DNS name.  At some point, the application
   translates this name to an IP address (via DNS), and subsequently
   makes initial contact with the server to begin a session, via SCTP,
   TCP, UDP, or other transport protocol.  If the client loses contact
   or fails to make contact with the server (either due to server
   failure, or a failure in the network) the client must either abandon
   the session, or try to contact another server.

   In this scenario, the client must first determine that a failure took
   place.  There are several ways that a client application may
   determine that a server failed, including the following:

   1.  The client may have sent a request to the server, and may time
       out waiting for a response, or may receive a message such as "no
       route to host", "port not available", or "connection refused".

Lei & Conrad              Expires April 9, 2006                 [Page 4]

Internet-Draft        Services Provided by RSerPool         October 2005

   2.  The client may have sent a request to the server, or may have
       tried to initiate a connection or association and may have
       received a connection/association failure error.

   3.  The client may already have established a connection to server,
       but at some point receives an indication from the transport layer
       that the connection failed.

   Suppose that the client application has a feature by which the user
   can enter the hostname of a secondary server to contact in the event
   of failure.  Once the application determines that a failure took
   place on the primary server, the application can then attempt to
   resolve the hostname of the secondary server, and contact the
   secondary server to establish a session there.  This process can be
   iterated to a tertiary server, and so forth.

   In this scenario, the identification of these alternate servers is an
   additional burden placedd on the end user.  Furthermore, there is no
   capability in this model to dynamically update the identity of the
   alternate servers based on current availablity or reachability.  DNS
   has some capabilities that can be used to help, but there are
   significant limitations to these capabilities (See [4] for a
   discussion of this point).

3.2.  Example Scenario Using RSerPool Basic Mode

   Now consider the same client/server application mentioned in
   Section 3.1.  First we describe what the application programmer must
   do to modify the code to use RSerPool Basic Mode.  We then describe
   the benefits that these modifications provide.

   For pool user ("client") applications, if an ASAP implementation is
   available on the client system, there are typically only three
   modifications required to the application source code:

   1.  Instead of specifying the hostnames of primary, secondary,
       tertiary servers, etc., the application user specifies a pool

   2.  Instead of using a DNS based service (e.g. the Unix library
       function gethostbyname()) to translate from a hostname to an IP
       address, the application will invoke an RSerPool service
       primitive GETPRIMARYSERVER that takes as input a pool handle, and
       returns the IP address of the primary server.  The application
       then uses that IP address just as it would have used the IP
       address returned by the DNS in the previous scenario.

Lei & Conrad              Expires April 9, 2006                 [Page 5]

Internet-Draft        Services Provided by RSerPool         October 2005

   3.  Without the use of additional RSerPool services, failure
       detection is application specific just as in the previous
       scenario.  However, when failure is detected on the primary
       server, instead of invoking DNS translation again on the hostname
       of a secondary server, the application invokes the service
       primitive GETNEXTSERVER, which performs two functions in a single

       1.  First it indicates to the RSerPool layer the failure of the
           server returned by a previous GETPRIMARYSERVER or
           GETNEXTSERVER call.

       2.  Second, it provides the IP address of the next server that
           should be contacted, according to the best information
           available to the RSerPool layer at the present time (e.g. set
           of available pool elements, pool element policy in effect for
           the pool, etc.).

   For pool element ("server") applications where an ASAP implementation
   is available, two changes are required to the application source

   1.  The server should invoke the REGISTER service primitive upon
       startup to add itself into the server pool using an appropriate
       pool handle.  This also includes the address(es) protocol or
       mapping id, port (if required by the mapping), and pooling

   2.  The server should invoke the DEREGISTER service primitive to
       remove itself from the server pool when shutting down.

   When using these RSerPool services, RSerPool provides benefits that
   are limited (as compared to utilizing all services, described in
   Section 3.3), but nevertheless quite useful as compared to not using
   RSerPool at all (as in Section 3.1).  First, the client user need
   only supply a single string, i.e. the pool handle, rather than a list
   of servers.  Second, the decision as to which server is to be used
   can be determined dynamically by the server selection mechanism (i.e.
   a "pool policy" performed by ASAP; see [3]).  Finally, when failures
   occur, these are reported to the pool via signaling present in ASAP
   [5]) and ENRP [6], other clients will eventually know (once this
   failure is confirmed by other elements of the RSerPool architecture)
   that this server has failed.

   Utilizing this subset of services is useful for:

      applications built over connectionless protocols such as UDP that
      cannot easily be adapted to the transport layer requirements

Lei & Conrad              Expires April 9, 2006                 [Page 6]

Internet-Draft        Services Provided by RSerPool         October 2005

      required for enhanced services (see section Section 5)

      applications running on systems which do not provide an
      appropriate mapping layer for the desired transport protcol

      an expedient way to provide some of the benefits of RSerPool to
      legacy applications (regardless of the transport protocol used)

   However, to take full advantage of the RSerPool framework,
   utilization of the complete set of Enhanced Mode services as
   described in the next section is recommended.

3.3.  Example Scenario Using RSerPool Enhanced Mode

   Finally, consider the same client/server application as in
   Section 3.1, but this time, modified to take advantage of RSerPool
   Enhanced Mode services.  As in the Section 3.1, we first describe the
   modifications needed, then we describe the benefits provided.

   When the full suite of RSerPool services are used, all communication
   between the pool user and the pool element is mediated by the
   RSerPool framework, including session establishment and teardown, and
   the sending and receiving of data.  Accordingly, it is necessary to
   modify the application to use the service primitives (i.e. the API)
   provided by RSerPool, rather than the transport layer primitives
   provided by TCP, SCTP, or whatever transport protocol is being used.

   As in the previous case, sessions (rather than connections or
   associations) are established, and the destination endpoint is
   specified as a pool handle rather than as a list of IP addresses with
   a port number.  However, failover from one pool element to another is
   fully automatic, and can be transparent to the application:

      The RSerPool framework control channel provides maintainance
      functions to keep pool element lists, policies, etc. current.

      Since the application data (e.g. data channel) is managed by the
      RSerPool framework, unsent data (data not yet submitted by
      RSerPool to the underlying transport protocol) is automatically
      redirected to the newly selected pool element upon failover.  If
      the underlying transport layer supports retrieval of unsent data
      (as in SCTP), retrieved unsent data can also be automatically re-
      sent to the newly selected pool element.

      An application server (pool element) can provide a state cookie
      (described in Section 4.5.1) that is automatically passed on to
      another pool element (by the ASAP layer at the pool user) in the
      event of a failover.  This state cookie can be used to assist the

Lei & Conrad              Expires April 9, 2006                 [Page 7]

Internet-Draft        Services Provided by RSerPool         October 2005

      application at the new pool element in recreating whatever state
      is needed to continue a session or transaction that was
      interrupted by a failure in the communication between a pool user
      and the original pool element.

      The application client (pool user) can provide a callback function
      (described in Section 4.5.2) that is invoked on the pool user side
      in the case of a failover.  This callback function can execute any
      application specific failover code, such as generating a special
      message (or sequence of messages) that helps the new pool element
      construct any state needed to continue an in-process session.

      Suppose in a particular peer-to-peer application, PU A is
      communicating with PE B, and it so happens that PU A is also a PE
      in pool X. PU A can pass a "business card" to PE B identifying it
      as a member of pool X. In the event of a failure at A, or a
      failure in the communication link between A and B, PE B can use
      the information in the business card to contact an equivalent PE
      to PU A from pool X.

      Additionally, if the application at PU A is aware of some
      particular PEs of pool X that would be preferred for B to contact
      in the event that A becomes unreachable from B, PU A can provide
      that list to the ASAP layer, and it will be included in A's
      business card.  (See Section 4.5.3)).

   Retrofitting an existing application for Enhanced Mode requires more
   application programmer effort than retrofitting an application for
   Basic Mode.  In particular, all use of the transport layer's
   primitives (e.g. the calls to the sockets API) must be replaced by
   the use of the RSerPool primitives (e.g. the RSerPool API).  This can
   be mitigated by making the RSerPool API as close to existing
   transport APIs as possible.  However, the benefit is that failure
   detection and failover is automated in this case.  This automatic
   failure detection takes advantage of heartbeat mechanisms that are
   provided either in the underlying transport protocol, or in a mapping
   defined on top of that protocol (see Section 4.5).

   Provided that developers of APIs for RSerPool stay close to familiar
   APIs for existing transport protocols, the effort of writing a new
   applications over RSerPool Enhanced Mode need not be significantly
   different from writing the same application directly over a supported
   transport protocol or mapping.

4.  Service Primitives

   Upper layer protocols and applications may "choose" to use these

Lei & Conrad              Expires April 9, 2006                 [Page 8]

Internet-Draft        Services Provided by RSerPool         October 2005

   primitive services as needed.  By selecting and using the appropriate
   set of service primitives, a range of failover scenarios may be
   supported.  These service primitives are described in the sub-
   sections that follow.

4.1.  Initialization

   The INITIALIZE service is used to establish a service access point to
   communicate with the ASAP layer on the local host.  This is the first
   service accessed by either a PU or a PE.

4.2.  PE Registration Services

   Pool Elements ("server") must use the following services to add or
   remove themselves from server pools:

      REGISTER, to add the pool element into a server pool using {pool
      handle, mapping mode, protocol or mapping id, port, policy info}
      where mapping mode is defined in Section 5.  A response result
      code is returned.

      DEREGISTER, to remove the pool element from a server pool using
      {pool handle, mapping mode, protocol or mapping id, port, policy
      info} where mapping mode is defined in Section 5.  A response
      result code is returned.

4.3.  PE Selection Services

   When automatic failover is enabled, selection of a new pool element
   according to the pool policy in place is automatically performed by
   the RSerPool framework in case of a detected failure (e.g. provides
   automatic failover).  No application intervention is required.

   Automatic failover may be enabled by setting the appropriate send
   flag when used in conjuction with data channel services (described in
   Section 4.4) or explicitly during initialization when data channel
   services are not used.

      FAILOVER_INDICATION, delivered by callback, indicates that a
      failover has occurred and that any required application level
      state recovery should be performed.  The newly selected pool
      element handle is provided.

      Business Card services: when automatic failover is used, the
      exchange of business cards for rendezvous services is
      automatically performed by the RSerPool framework (e.g. no
      application intervention is required.

Lei & Conrad              Expires April 9, 2006                 [Page 9]

Internet-Draft        Services Provided by RSerPool         October 2005

   When automatic failover is not enabled, failover detection and
   selection of an alternate PE must be done by the upper layer/
   application.  The following primitives are provided:

      GET_PRIMARY_SERVER, takes as input a pool handle and returns the
      {IP address, transport protocol, transport protocol port} of the
      primary server.

      GET_NEXT_SERVER has a dual meaning.  First, it indicates to the
      RSerPool layer the failure of the server returned by a previous
      GET_PRIMARY_SERVER or GET_NEXT_SERVER call.  Second, it provides
      the {IP address, transport protocol, transport protocol port} of
      the next server that should be contacted, according to the best
      information available to the RSerPool layer at the present time.
      The appropriate pool policy for server selection for the pool
      should be used for selecting the next server.

4.4.  RSerPool Managed Data Channel

   The RSerPool framework provides these services to send and receive
   application layer data, which are used in place of the direct call of
   transport level system functions (e.g. send/sendto, recv/recvfrom)
   and provides additional functionality to those calls.

      DATA_SEND_REQUEST, to send data to a pool element by using a pool
      handle, specific pool element handle, or by transport address.
      When sending to a pool handle, the specific pool element handle
      chosen is returned.  In the case that data is sent to a pool
      handle, or specific pool element handle, the user can request
      automatic resending (on a best-effort basis) if the original pool
      element selected is unreachable.  (However, it is ultimately the
      application's responsibility to detect and recover from errors,
      using acknowledgements at the application layer if needed.)

      When sending to a specific transport address, this primitive is
      considered a "pass thru" to the underlying transport, and no
      failover services are performed.

      In each case, appropriate error code(s) are returned in the event
      of failure. (see [5] for more detail).

      DATA_RECEIVED_NOTIFICATION, delivered by callback, to indicate
      that data has been received from a pool element and to pass that
      data to the application layer protocol.  An application layer
      acknowledgement request can be indicated along with the data.

Lei & Conrad              Expires April 9, 2006                [Page 10]

Internet-Draft        Services Provided by RSerPool         October 2005

4.5.  Failover Services

   The charter of the RSerPool Working Group specifically states that
   transaction failover is out of scope for RSerPool, i.e. "if a server
   fails during processing of a transaction this transaction may be
   lost.  Some services may provide a way to handle the failure, but
   this is not guaranteed."  Accordingly, the RSerPool framework
   provides three "hooks" for applications to provide their own
   application-specific failover mechanism(s), one on the PE side (State
   Cookie Exchange), one on the PU side (Failover Callback), and one for
   entities that are combination of PU/PE (business card).

4.5.1.  State Cookie Exchange

      SET_COOKIE: This is invoked by a PE to set the state cookie that
      is sent periodically over the control channel, when present, from
      a PE to a PU.  The most recently received cookie is cached by the
      PU; in the event of failover, it is forwarded to the new PE.

      COOKIE_INDICATION: This is invoked by callback at a PE, when that
      PE receives a cookie from a PU.  This cookie is an indication that
      the PU has failed over to the current PE from some other PE.  The
      contents of the cookie are provided to the PE prior to any
      data.indication for messages arriving from the PU that sent the
      cookie.  This provides a hook by which a PE "X" can send a "hint"
      to its successor PE "Y", in the event that one of X's PU's fails
      over from X to Y. PE "Y" can use the contents of the cookie to
      establish application layer state prior to processing resent or
      new messages from the PU.  The PU application layer is not
      involved in any way in this exchange; it is handled automatically
      by the ASAP layer.

4.5.2.  Failover Callback Function

   AUTHOR'S NOTE (PTC): the service defined in this section is not a
   part of section 4 of the current version ASAP draft.  It should
   either be added to the ASAP draft, or this service should be removed
   from the services draft, after discussion on the list.  Open
   question: does the cookie feature eliminate the need for this

   An PU that establishing a session with a PE can specify a callback
   function that is invoked whenever a failover has taken place.  This
   callback function is invoked immediately after the new transport
   layer connection/ association is established with a new server, and
   gives the application the opportunity to send one or more messages
   that may help the server to resume any transaction or session that
   was in progress when the first server failed.  In essence, this

Lei & Conrad              Expires April 9, 2006                [Page 11]

Internet-Draft        Services Provided by RSerPool         October 2005

   allows an application designed to put the reestablishment of state
   into the PU side instead of the PE side, if desired.

   This service that complements the cookie feature, in the following
   way: the cookie feature provides failover hooks on the PE side, where
   the callback is a failover hook for the PU side.  The on-the-wire
   impact is that it is important that the ASAP entity should invoke the
   failover callback (if any is registered) prior to resending any
   messages from previous DATA_SEND_REQUEST primitives.

   Note that if both a state cookie from a PU and a failover callback
   are present, the state cookie should be sent before the failover
   callback is issued.

   As a simple example of how such a callback is useful, consider a file
   transfer service built using RSerPool.  Let us assume that some FTP
   mirroring software is used to maintain mirrored sites, and that the
   actual mirroring is out of scope.  However, we would like to use
   RSerPool to select a server from among the available mirror sites,
   and to failover in the middle of a file transfer if a primary server

   For this example, assume that a simple request/response protocol is
   used, where one request message results in one or more response
   messages.  Each request message contains the filename, and the offset
   desired within the file, (default zero.)  Each response message
   contains some portion of the file, along with the offset, length of
   the portion in this message, and the length of the entire file.

   A single request is sufficient to result in a sequence of response
   messages from the requested offset to the end of the file.

   In this protocol, all that is needed for failover is for the
   application to:

      keep track of the lowest byte that it has not yet received from
      the server,

      provide a callback function that reissues the request to the new
      server, replacing the offset with this number.

   When there is no failover, only one request message is sent and the
   minimum number of response messages are returned; in the event of
   failover(s), single new request message is sent for each failover
   that occurs.

   While this is a simple example, for more complex application
   requirements, the failover callback could be used in a variety of

Lei & Conrad              Expires April 9, 2006                [Page 12]

Internet-Draft        Services Provided by RSerPool         October 2005


      The client might send security credentials for authentication by
      the server, and/or to provide a "key" by which the server could
      locate and setup state by accessing some application-specific (and
      out-of-scope) state sharing mechanism used by the servers.

      The client might keep track of various synchronization points in
      the transaction, and use the failover callback to replay message
      from a recent synchronization point.

4.5.3.  Business Card

   This section TBD... describe a service primitive
   Set.BusinessCard.PE.List.  What should the parameters be?  This
   service primitive should also be defined in Section 4 of the ASAP

5.  Transport Mappings

   While SCTP is the preferred transport layer protocol for applications
   built for RSerPool failover mode (for reasons explained shortly), it
   is also possible to use other transport protocols as well (e.g.  TCP)
   if an SCTP implementation is not available on the client and/or
   server.  However, there are certain features present in SCTP that are
   required if the RSerPool framework is to function in failover mode.
   When a transport protocol other than SCTP is used, these features
   must be provided by an "adaption layer" (also called a "shim
   protocol") that sits between the base transport protocol (e.g.  TCP)
   and the RSerPool layer.  We refer to these "adaptation layers" or
   "shim protocols" as "mappings" as the idea is that the requirements
   of the RSerPool framework are "mapped" onto the capabilities of the
   underlying protocol (e.g.  SCTP or TCP).

5.1.  Defined Transport Mappings

   In order to support the RSerPool framework over a variety of
   transport protocols and configurations, several mappings are defined
   to provide RSerPool services over a given transport protocol.  Each
   mapping translates the requirements of the RSerPool framework onto
   the capabilities of the transport protocol desired (e.g.  SCTP, TCP,
   etc.).  Initially, three mappings are defined:

      NO_MAPPING (0x00): With this mapping, no RserPool control channel
      is provided and the application specific communication between a
      pool user and the pool element (e.g. data channel) is out of scope
      of RSerPool.  However, pool elements can register the application

Lei & Conrad              Expires April 9, 2006                [Page 13]

Internet-Draft        Services Provided by RSerPool         October 2005

      specific communication "protocol" and "port", and thus can be
      provided to pool users.

      SCTP (0x01): SCTP transport is used for the RSerPool control
      channel.  The data channel MAY be multiplexed onto the same SCTP
      association, if desired.  This mapping is the preferred mapping.

      TCP (0x02): TCP transport is used for the RSerPool control
      channel.  The data channel MAY be multiplexed onto the same TCP
      connection, if desired.

   A particular pool element might support any combination of these
   mappings in order to support a variety of pool users with different
   capabilities (i.e. different mapping support).  In this case, pool
   elements should register each mapping that it supports with its

5.2.  Transport Mappings Requirements

5.2.1.  Mappings: Mandatory Requirements

   These features MUST be present in any mapping of the RSerPool
   framework mode to TCP (or any other transport protocol):

   1.  Message orientation, which facilitates application re-
       synchronization during failover.  Messages must be "framed" in
       order to allow for undelivered message retrieval from the
       transport protocol.

   2.  A heartbeat mechanism to monitor the health of an association or

   3.  A mechanism to transport and differentiate between control
       channel messages (e.g.  ASAP messages) and data channel messages.
       For example in SCTP, the payload protocol identifier (PPID) may
       be used.

   4.  [NOTE: retrieval was eliminated here as a requirement, now that
       failover is best effort.]

5.2.2.  Mappings: Optional Requirements

   There are several additional features that are present in SCTP that
   are lacking in TCP.  While these features are not crucial to
   RSerPool, providing them in the mapping layer makes it easier for an
   application layer programmer to write to a single API.  This single
   API can then be mapped over both SCTP and TCP, as well as any other
   transport protocol for which a mapping is provided.  Since these

Lei & Conrad              Expires April 9, 2006                [Page 14]

Internet-Draft        Services Provided by RSerPool         October 2005

   features are not essential for RSerPool, they are optional in any
   defined mapping.  However, appropriate error messages or indications
   should be provided when these features are not available.  These
   features include:

   1.  Support for multiple streams

   2.  Support for unordered delivery of messages

5.2.3.  Mappings: Other Requirements

   There are some features of SCTP that a mapping may not be able to
   provide, because they would require access to transport layer
   internals, or modifications in the transport layer itself.  The
   services provided by the RSerPool layer to the application should
   therefore provide mechanisms for the upper layer to access these
   features when present (e.g. in SCTP), but also provide appropriate
   error messages or indications that these features are not available
   when they cannot be provided.  These features include:

   1.  Application access to the RTT and RTO estimates

   2.  Application access to the Path MTU value

   3.  Application access to set the lifetime parameter on outgoing SCTP

6.  Security Considerations

   [Open Issue TBD: Security issues are not discussed in this memo at
   this time, but will be added in a later version of this draft.]

7.  IANA Considerations

   [Open Issue TBD: Will there be an enumeration of the various
   transport layer mappings that must be registered with IANA?]

8.  Acknowledgements

   The authors wish to thank Maureen Stillman, Qiaobing Xie, Michael
   Tuexen, Randall Stewart, and many others for their invaluable

Lei & Conrad              Expires April 9, 2006                [Page 15]

Internet-Draft        Services Provided by RSerPool         October 2005

9.  References

   [1]  Bradner, S., "The Internet Standards Process -- Revision 3",
        BCP 9, RFC 2026, October 1996.

   [2]  Bradner, S., "Key words for use in RFCs to Indicate Requirement
        Levels", BCP 14, RFC 2119, March 1997.

   [3]  Tuexen, M., "Architecture for Reliable Server Pooling",
        draft-ietf-rserpool-arch-10 (work in progress), July 2005.

   [4]  Loughney, J., "Comparison of Protocols for Reliable Server
        Pooling", draft-ietf-rserpool-comp-10 (work in progress),
        July 2005.

   [5]  Stewart, R., "Aggregate Server Access Protocol (ASAP)",
        draft-ietf-rserpool-asap-12 (work in progress), July 2005.

   [6]  Stewart, R., "Endpoint Handlespace Redundancy Protocol (ENRP)",
        draft-ietf-rserpool-enrp-12 (work in progress), July 2005.

Lei & Conrad              Expires April 9, 2006                [Page 16]

Internet-Draft        Services Provided by RSerPool         October 2005

Authors' Addresses

   Peter Lei
   Cisco Systems
   8735 W Higgins Rd, Suite 300
   Chicago, IL  60631

   Phone: +1 773 695 8201
   Email: peterlei@cisco.com

   Phillip T. Conrad
   University of Delaware
   Dept. of Computer and Information Sciences
   103 Smith Hall
   Newark, DE  19716

   Phone: +1 302 831 8622
   Email: conrad@acm.org
   URI:   http://udel.edu/~pconrad

Lei & Conrad              Expires April 9, 2006                [Page 17]

Internet-Draft        Services Provided by RSerPool         October 2005

Intellectual Property Statement

   The IETF takes no position regarding the validity or scope of any
   Intellectual Property Rights or other rights that might be claimed to
   pertain to the implementation or use of the technology described in
   this document or the extent to which any license under such rights
   might or might not be available; nor does it represent that it has
   made any independent effort to identify any such rights.  Information
   on the procedures with respect to rights in RFC documents can be
   found in BCP 78 and BCP 79.

   Copies of IPR disclosures made to the IETF Secretariat and any
   assurances of licenses to be made available, or the result of an
   attempt made to obtain a general license or permission for the use of
   such proprietary rights by implementers or users of this
   specification can be obtained from the IETF on-line IPR repository at

   The IETF invites any interested party to bring to its attention any
   copyrights, patents or patent applications, or other proprietary
   rights that may cover technology that may be required to implement
   this standard.  Please address the information to the IETF at

Disclaimer of Validity

   This document and the information contained herein are provided on an

Copyright Statement

   Copyright (C) The Internet Society (2005).  This document is subject
   to the rights, licenses and restrictions contained in BCP 78, and
   except as set forth therein, the authors retain all their rights.


   Funding for the RFC Editor function is currently provided by the
   Internet Society.

Lei & Conrad              Expires April 9, 2006                [Page 18]