RFC 6369






Internet Engineering Task Force (IETF)                     E. Haleplidis
Request for Comments: 6369                                O. Koufopavlou
Category: Informational                                       S. Denazis
ISSN: 2070-1721                                     University of Patras
                                                          September 2011


           Forwarding and Control Element Separation (ForCES)
                       Implementation Experience

Abstract



   The Forwarding and Control Element Separation (ForCES) protocol
   defines a standard communication and control mechanism through which
   a Control Element (CE) can control the behavior of a Forwarding
   Element (FE).  This document captures the experience of implementing
   the ForCES protocol and model.  Its aim is to help others by
   providing examples and possible strategies for implementing the
   ForCES protocol.

Status of This Memo



   This document is not an Internet Standards Track specification; it is
   published for informational purposes.

   This document is a product of the Internet Engineering Task Force
   (IETF).  It represents the consensus of the IETF community.  It has
   received public review and has been approved for publication by the
   Internet Engineering Steering Group (IESG).  Not all documents
   approved by the IESG are a candidate for any level of Internet
   Standard; see Section 2 of RFC 5741.

   Information about the current status of this document, any errata,
   and how to provide feedback on it may be obtained at
   http://www.rfc-editor.org/info/rfc6369.

Copyright Notice



   Copyright (c) 2011 IETF Trust and the persons identified as the
   document authors.  All rights reserved.

   This document is subject to BCP 78 and the IETF Trust's Legal
   Provisions Relating to IETF Documents
   (http://trustee.ietf.org/license-info) in effect on the date of
   publication of this document.  Please review these documents
   carefully, as they describe your rights and restrictions with respect
   to this document.  Code Components extracted from this document must




Haleplidis, et al.            Informational                     [Page 1]

RFC 6369            ForCES Implementation Experience      September 2011


   include Simplified BSD License text as described in Section 4.e of
   the Trust Legal Provisions and are provided without warranty as
   described in the Simplified BSD License.

Table of Contents



   1.  Introduction . . . . . . . . . . . . . . . . . . . . . . . . .  2
     1.1.  Document Goal  . . . . . . . . . . . . . . . . . . . . . .  3
   2.  Terminology and Conventions  . . . . . . . . . . . . . . . . .  3
   3.  ForCES Architecture  . . . . . . . . . . . . . . . . . . . . .  4
     3.1.  Pre-Association Setup - Initial Configuration  . . . . . .  5
     3.2.  TML  . . . . . . . . . . . . . . . . . . . . . . . . . . .  5
     3.3.  Model  . . . . . . . . . . . . . . . . . . . . . . . . . .  6
       3.3.1.  Components . . . . . . . . . . . . . . . . . . . . . .  6
       3.3.2.  LFBs . . . . . . . . . . . . . . . . . . . . . . . . .  9
     3.4.  Protocol . . . . . . . . . . . . . . . . . . . . . . . . . 10
       3.4.1.  TLVs . . . . . . . . . . . . . . . . . . . . . . . . . 10
       3.4.2.  Message Deserialization  . . . . . . . . . . . . . . . 13
       3.4.3.  Message Serialization  . . . . . . . . . . . . . . . . 15
   4.  Development Platforms  . . . . . . . . . . . . . . . . . . . . 15
   5.  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . 16
   6.  Security Considerations  . . . . . . . . . . . . . . . . . . . 16
   7.  References . . . . . . . . . . . . . . . . . . . . . . . . . . 17
     7.1.  Normative References . . . . . . . . . . . . . . . . . . . 17
     7.2.  Informative References . . . . . . . . . . . . . . . . . . 17

1.  Introduction



   Forwarding and Control Element Separation (ForCES) defines an
   architectural framework and associated protocols to standardize
   information exchange between the control plane and the forwarding
   plane in a ForCES Network Element (ForCES NE).  [RFC3654] defines the
   ForCES requirements, and [RFC3746] defines the ForCES framework.

   The ForCES protocol works in a master-slave mode in which Forwarding
   Elements (FEs) are slaves and Control Elements (CEs) are masters.
   The protocol includes commands for transport of Logical Functional
   Block (LFB) configuration information, association setup, status, and
   event notifications, etc.  The reader is encouraged to read the
   Forwarding and Control Element Separation Protocol [RFC5810] for
   further information.

   [RFC5812] presents a formal way to define FE LFBs using XML.  LFB
   configuration components, capabilities, and associated events are
   defined when LFBs are formally created.  The LFBs within the
   Forwarding Element (FE) are accordingly controlled in a standardized
   way by the ForCES protocol.




Haleplidis, et al.            Informational                     [Page 2]

RFC 6369            ForCES Implementation Experience      September 2011


   The Transport Mapping Layer (TML) transports the protocol messages.
   The TML is where the issues of how to achieve transport-level
   reliability, congestion control, multicast, ordering, etc., are
   handled.  It is expected that more than one TML will be standardized.
   The various possible TMLs could vary their implementations based on
   the capabilities of underlying media and transport.  However, since
   each TML is standardized, interoperability is guaranteed as long as
   both endpoints support the same TML.  All ForCES protocol layer
   implementations must be portable across all TMLs.  Although more than
   one TML may be standardized for the ForCES protocol, all ForCES
   implementations must implement the Stream Control Transmission
   Protocol (SCTP) TML [RFC5811].

   The Forwarding and Control Element Separation Applicability Statement
   [RFC6041] captures the applicable areas in which ForCES can be used.

1.1.  Document Goal



   This document captures the experience of implementing the ForCES
   protocol and model, and its main goal is to provide alternatives,
   ideas, and proposals as how it can be implemented, not to tell others
   how to implement it.

   Also, this document mentions possible problems and potential choices
   that can be made, in an attempt to help implementors develop their
   own products.

   Additionally, this document assumes that the reader has become
   familiar with the three main ForCES RFCs: the Forwarding and Control
   Element Separation Protocol [RFC5810], the Forwarding and Control
   Element Separation Forwarding Element Model [RFC5812], and the SCTP-
   Based Transport Mapping Layer (TML) for the Forwarding and Control
   Element Separation Protocol [RFC5811].

2.  Terminology and Conventions



   The terminology used in this document is the same as in the
   Forwarding and Control Element Separation Protocol [RFC5810]; some of
   the definitions below are copied from that document.

   Control Element (CE): A logical entity that implements the ForCES
   protocol and uses it to instruct one or more FEs on how to process
   packets.  CEs handle functionality such as the execution of control
   and signaling protocols.







Haleplidis, et al.            Informational                     [Page 3]

RFC 6369            ForCES Implementation Experience      September 2011


   Forwarding Element (FE): A logical entity that implements the ForCES
   protocol.  FEs use the underlying hardware to provide per-packet
   processing and handling as directed/controlled by one or more CEs via
   the ForCES protocol.

   LFB (Logical Functional Block): The basic building block that is
   operated on by the ForCES protocol.  The LFB is a well-defined,
   logically separable functional block that resides in an FE and is
   controlled by the CE via the ForCES protocol.  The LFB may reside at
   the FE's data path and process packets or may be purely an FE control
   or configuration entity that is operated on by the CE.  Note that the
   LFB is a functionally accurate abstraction of the FE's processing
   capabilities but not a hardware-accurate representation of the FE
   implementation.

   LFB Class and LFB Instance: LFBs are categorized by LFB classes.  An
   LFB instance represents an LFB class (or type) existence.  There may
   be multiple instances of the same LFB class (or type) in an FE.  An
   LFB class is represented by an LFB class ID, and an LFB instance is
   represented by an LFB instance ID.  As a result, an LFB class ID
   associated with an LFB instance ID uniquely specifies an LFB
   existence.

   LFB Component: Operational parameters of the LFBs that must be
   visible to the CEs are conceptualized in the FE model as the LFB
   components.  The LFB components include, for example, flags, single
   parameter arguments, complex arguments, and tables that the CE can
   read and/or write via the ForCES protocol.

   ForCES Protocol: While there may be multiple protocols used within
   the overall ForCES architecture, the terms "ForCES protocol" and
   "protocol" refer to the Fp reference points in the ForCES framework
   [RFC3746].  This protocol does not apply to CE-to-CE communication,
   FE-to-FE communication, or communication between FE and CE Managers.
   Basically, the ForCES protocol works in a master-slave mode in which
   FEs are slaves and CEs are masters.  This document defines the
   specifications for this ForCES protocol.

3.  ForCES Architecture



   ForCES has undergone two successful interoperability tests, where
   very few issues were caught and resolved.

   This section discusses the ForCES architecture, implementation
   challenges, and ways to overcome these challenges.






Haleplidis, et al.            Informational                     [Page 4]

RFC 6369            ForCES Implementation Experience      September 2011


3.1.  Pre-Association Setup - Initial Configuration



   The initial configuration of the FE and the CE is done by the FE
   Manager and the CE Manager, respectively.  These entities have not as
   yet been standardized.

   The simplest solution is static configuration files, which play the
   role of the Managers and are read by FEs and CEs.

   For more dynamic solutions, however, it is expected that the Managers
   will be entities that will talk to each other and exchange details
   regarding the associations.  Any developer can create any Manager,
   but they should at least be able to exchange the details below.

   From the FE Manager side:

   1.  FE Identifiers (FEIDs).

   2.  FE IP addresses, if the FEs and CEs will be communicating via
       network.

   3.  TML.  The TML that will be used.  If this is omitted, then SCTP
       must be chosen as default.

   4.  TML priority ports.  If this is omitted as well, then the CE must
       use the default values from the respective TML RFC.



   From the CE Manager side:

   1.  CE Identifiers (CEIDs).

   2.  CE IP addresses, if the FEs and CEs will be communicating via
       network.

   3.  TML.  The TML that will be used.  If this is omitted, then SCTP
       must be chosen as default.

   4.  TML priority ports.  If this is omitted as well, then the FE must
       use the default values from the respective TML RFC.

3.2.  TML



   All ForCES implementations must support the SCTP TML.  Even if
   another TML will be chosen by the developer, SCTP is mandatory and
   must be supported.






Haleplidis, et al.            Informational                     [Page 5]

RFC 6369            ForCES Implementation Experience      September 2011


   There are several issues that should concern a developer for the TML:

   1.  Security.  TML must be secure according to the respective RFC.
       For SCTP, you have to use IPsec.

   2.  Remote connection.  While ForCES is meant to be used locally,
       both interoperability tests have proven that ForCES can be
       deployed everywhere that SCTP/IP is available.  In both
       interoperability tests, there were connections between Greece and
       China, and the performance was very satisfactory.  However, in
       order for the FE and CE to work in a non-local environment, an
       implementor must ensure that the SCTP-TML ports are forwarded to
       the CE and/or FE if they are behind NATs; if there is a firewall,
       it will allow the SCTP ports through.  These were identified
       during the first ForCES interoperability test and documented in
       the Implementation Report for Forwarding and Control Element
       Separation [RFC6053].

3.3.  Model



   The ForCES model is inherently very dynamic.  Using the basic atomic
   data types that are specified in the model, new atomic (single
   valued) and/or compound (structures and arrays) datatypes can be
   built.  Thus, developers are free to create their own LFBs.  One
   other advantage that the ForCES model provides is inheritance.  New
   versions of existing LFBs can be created to suit any extra developer
   requirements.

   The difficulty for a developer is to create an architecture that is
   completely scalable so there is no need to write the same code for
   new LFBs, new components, etc.  Developers can just create code for
   the defined atomic values, and new components can then be built based
   on already written code, thus reusing it.

   The model itself provides the key, which is inheritance.

3.3.1.  Components



   First, a basic component needs to be created as the mother of all the
   components that has the basic parameters of all the components:

   o  The ID of the component.

   o  The access rights of the component.

   o  If it is an optional component.

   o  If it is of variable size.



Haleplidis, et al.            Informational                     [Page 6]

RFC 6369            ForCES Implementation Experience      September 2011


   o  Minimum data size.

   o  Maximum data size.

   If the data size of the component is not variable, then the size is
   either the minimum or the maximum size, as both should have the same
   value.

   Next, some basic functions are in order:

   o  A common constructor.

   o  A common destructor.

   o  Retrieve Component ID.

   o  Retrieve access right property.

   o  Query if it is an optional component.

   o  Get Full Data.

   o  Set Full Data.

   o  Get Sparse Data.

   o  Set Sparse Data.

   o  Del Full Data.

   o  Del Sparse Data.

   o  Get Property.

   o  Set Property.

   o  Get Value.

   o  Set Value.

   o  Del Value.

   o  Get Data.

   o  Clone component.






Haleplidis, et al.            Informational                     [Page 7]

RFC 6369            ForCES Implementation Experience      September 2011


   The Get/Set/Del Full Data, Get/Set/Del Sparse Data, and Get/Set
   Property functions handle the respective ForCES commands and return
   the respective TLV, for example, Set Full Data should return a
   RESULT-TLV.  The Get Value, Set Value, and Del Value functions are
   called from Get Full/Sparse Data, Set Full/Sparse Data, and Del Full/
   Sparse Data respectively and provide the interface to the actual
   values in the hardware, separating the forces handling logic from the
   interface to the actual values.

   The Get Data function should return the value of the data only, not
   in TLV format.

   The Clone function seems out of place.  This function must return a
   new component that has the exact same values and attributes.  This
   function is useful in array components as described further below.

   The only requirement is to implement the base atomic data types.  Any
   new atomic datatype can be built as a child of a base data type,
   which will inherit all the functions and, if necessary, override
   them.

   The struct component can then be built.  A struct component is a
   component by itself but consists of a number of atomic components.
   These atomic components create a static array within the struct.  The
   ID of each atomic component is the array's index.  For a struct
   component, the Clone function must create and return an exact copy of
   the struct component with the same static array.

   The most difficult component to be built is the array.  The
   difficulty lies in the actual benefit of the model: you have absolute
   freedom over what you build.  An array is an array of components.  In
   all rows, you have the exact same type of component, either a single
   component or a struct.  The struct can have multiple single
   components or a combination of single components, structs, arrays,
   and so on.  So, the difficulty lies in how to create a new row, a new
   component by itself.  This is where the Clone function is very
   useful.  For the array, a mother component that can spawn new
   components exactly like itself is needed.  Once a Set command is
   received, the mother component can spawn a new component if the
   targeted row does not exist and add it into the array; with the Set
   Full Data function, the value is set in the recently spawned
   component, as the spawned component knows how the data is created.
   In order to distinguish these spawned components from each other and
   their functionality, some kind of index is required that will also
   reflect how the actual data of the specific component is stored on
   the hardware.





Haleplidis, et al.            Informational                     [Page 8]

RFC 6369            ForCES Implementation Experience      September 2011


   Once the basic constructors of all possible components are created,
   then a developer only has to create LFB components or datatypes as a
   child of one of the already-created components, and the only thing
   the developer really needs to add is the three functions of Get
   Value, Set Value, and Del Value of each component, which is platform
   dependent.  The rest stays the same.

3.3.2.  LFBs



   The same architecture in the components can be used for the LFBs,
   allowing a developer to write LFB handling code only once.  The
   parent LFB has some basic attributes:

   o  The LFB Class ID.

   o  The LFB Instance ID.

   o  An Array of Components.

   o  An Array of Capabilities.

   o  An Array of Events.

   Following are some common functions:

   o  Handle Configuration Command.

   o  Handle Query Command.

   o  Get Class ID.

   o  Get Instance ID.

   Once these are created, each LFB can inherit all these from the
   parent, and the only thing it has to do is add the components that
   have already been created.

   An example can be seen in Figure 1.  The following code creates a
   part of FEProtocolLFB:












Haleplidis, et al.            Informational                     [Page 9]

RFC 6369            ForCES Implementation Experience      September 2011


   //FEID
   cui = new Component_uInt(FEPO_FEID, ACCESS_READ_ONLY, FE_id);
   Components[cui->get_ComponentId()]=cui; //Add component to array list

   //Current FEHB Policy Value
   cub = new Component_uByte(FEPO_FEHBPolicy, ACCESS_READ_WRITE, 0);
   Components[cub->get_ComponentId()]=cub; //Add component to array list

   //FEIDs for BackupCEs Array
   cui = new Component_uInt(0, ACCESS_READ_WRITE, 0);
   ca = new Component_Array(FEPO_BackupCEs, ACCESS_READ_WRITE);
   ca->AddRow(cui, 1);
   ca->AddMotherComponent(cui);
   Components[ca->get_ComponentId()]=ca; //Add component to array list

         Figure 1: Example Code for Creating Part of FEProtocolLFB

   The same concept can be applied to handling LFBs as one FE.  An FE is
   a collection of LFBs.  Thus, all LFBs can be stored in an array based
   on the LFB's class id, version, and instance.  Then, what is required
   is an LFBHandler that will handle the array of LFBs.  A specific LFB,
   for example, can be addressed using the following scheme:

   LFBs[ClassID][Version][InstanceID]

   Note: While an array can be used in components, capabilities, and
   events, a hash table or a similar concept is better suited for
   storing LFBs using the component ID as the hash key with linked lists
   for collision handling, as the created array can have large gaps if
   the values of LFB Class ID vary greatly.

3.4.  Protocol



3.4.1.  TLVs



   The goal for protocol handling is to create a general and scalable
   architecture that handles all protocol messages instead of something
   implementation specific.  There are certain difficulties that have to
   be overcome first.

   Since the model allows a developer to define any LFB required, the
   protocol has been thus created to give the user the freedom to
   configure and query any component, whatever the underlying model.
   While this is a strong point for the protocol itself, one difficulty
   lies with the unknown underlying model and the unlimited number of
   types of messages that can be created, making creating generic code a
   daunting task.




Haleplidis, et al.            Informational                    [Page 10]

RFC 6369            ForCES Implementation Experience      September 2011


   Additionally, the protocol also allows two different path approaches
   to LFB components, and the CE or FE must handle both or even a mix of
   them, making a generic decoding of the protocol message difficult.

   Another difficulty also arises from the batching capabilities of the
   protocol.  You can have multiple Operations within a message; you can
   select more than one LFB to command and more than one component to
   manipulate.

   A possible solution is again provided by inheritance.  There are two
   basic components in a protocol message:

   1.  The common header.

   2.  The rest of the message.

   The rest of the message is divided in Type-Length-Value (TLV) units
   and, in one case, Index-Length-Value (ILV) units.

   The TLV hierarchy can be seen in Figure 2:

                      Common Header
                            |
            +---------------+---------------+---------------+
            |               |               |               |
         REDIRECT-TLV  LFBselect-TLV   ASResult-TLV   ASTreason-TLV
                            |
                            |
                        OPER-TLV
                            |
                            |
                      PATH-DATA-TLV  ---> Optional KEYINFO-TLV
                            |
              +-------------+-------------+-------------+
              |             |             |             |
          SPARSEDATA-TLV  RESULT-TLV  FULLDATA-TLV  PATH-DATA-TLV

                      Figure 2: ForCES TLV Hierarchy

   The above figure shows only the basic hierarchical level of TLVs and
   does not show batching.  Also, this figure does not show the
   recursion that can occur at the last level of the hierarchy.  The
   figure shows one kind of recursion with a PATH-DATA-TLV within a
   PATH-DATA-TLV.  A FULLDATA-TLV can be within a FULLDATA-TLV and a
   SPARSEDATA-TLV.  The possible combination of TLVs are described in
   detail in the Forwarding and Control Element Separation Protocol
   [RFC5810] as well as the data-packing rules.




Haleplidis, et al.            Informational                    [Page 11]

RFC 6369            ForCES Implementation Experience      September 2011


   A TLV's main attributes are:

   o  Type.

   o  Length.

   o  Data.

   o  An array of TLVs.

   The array of TLVs is the next hierarchical level of TLVs nested in
   this TLV.

   A TLV's common function could be:

   o  A basic constructor.

   o  A constructor using data from the wire.

   o  Add a new TLV for next level.

   o  Get the next TLV of next level.

   o  Get a specific TLV of next level.

   o  Replace a TLV of next level.

   o  Get the Data.

   o  Get the Length.

   o  Set the Data.

   o  Set the Length.

   o  Set the Type.

   o  Serialize the header.

   o  Serialize the TLV to be written on the wire.

   All TLVs inherit these functions and attributes and either override
   them or create new where it is required.








Haleplidis, et al.            Informational                    [Page 12]

RFC 6369            ForCES Implementation Experience      September 2011


3.4.2.  Message Deserialization



   Following is an algorithm for deserializing any protocol message:

   1.  Get the message header.

   2.  Read the length.

   3.  Check the message type to understand what kind of message this
       is.

   4.  If the length is larger than the message header, then there is
       data for this message.

   5.  A check can be made here regarding the message type and the
       length of the message.



   If the message is a Query or Config type, then there are LFBselect-
   TLVs for this level:

   1.  Read the next 2 shorts(type-length).  If the type is an
       LFBselect-TLV, then the message is valid.

   2.  Read the necessary length for this LFBselect-TLV, and create the
       LFBselect-TLV from the data of the wire.

   3.  Add this LFBselect-TLV to the main header array of LFBselect-
       TLVs.

   4.  Repeat all above steps until the rest of the message has
       finished.

   The next level of TLVs is OPER-TLVs.

   1.  Read the next 2 shorts(type-length).  If the type is an OPER-TLV,
       then the message is valid.

   2.  Read the necessary length for this OPER-TLV, and create the OPER-
       TLV from the data of the wire.

   3.  Add this OPER-TLV to the LFBselect-TLV array of TLVs.

   4.  Do this until the rest of the LFBselect-TLV has finished.

   The next level of TLVs is PATH-DATA-TLVs.

   1.  Read the next 2 shorts(type-length).  If the type is a PATH-DATA-
       TLV, then the message is valid.



Haleplidis, et al.            Informational                    [Page 13]

RFC 6369            ForCES Implementation Experience      September 2011


   2.  Read the necessary length for this PATH-DATA-TLV, and create the
       PATH-DATA-TLV from the data of the wire.

   3.  Add this PATH-DATA-TLV to the OPER-TLV's array of TLVs.

   4.  Do this until the rest of the OPER-TLV is finished.

   Here it gets interesting, as the next level of PATH-DATA-TLVs can be
   one of the following:

   o  PATH-DATA-TLVs.

   o  FULLDATA-TLV.

   o  SPARSEDATA-TLV.

   o  RESULT-TLV.

   The solution to this difficulty is recursion.  If the next TLV is a
   PATH-DATA-TLV, then the PATH-DATA-TLV that is created uses the same
   kind of deserialization until it reaches a FULLDATA-TLV or
   SPARSEDATA-TLV.  There can be only one FULLDATA-TLV or SPARSEDATA-TLV
   within a PATH-DATA-TLV.

   1.  Read the next 2 shorts(type-length).

   2.  If the Type is a PATH-DATA-TLV, then repeat the previous
       algorithm but add the PATH-DATA-TLV to this PATH-DATA-TLV's array
       of TLVs.

   3.  Do this until the rest of the PATH-DATA-TLV is finished.

   4.  If the Type is a FULLDATA-TLV, then create the FULLDATA-TLV from
       the message and add this to the PATH-DATA-TLV's array of TLVs.

   5.  If the Type is a SPARSEDATA-TLV, then create the SPARSEDATA-TLV
       from the message and add this to the PATH-DATA-TLV's array of
       TLVs.

   6.  If the Type is a RESULT-TLV, then create the RESULT-TLV from the
       message and add this to the PATH-DATA-TLV's array of TLVs.



   If the message is a Query, it must not have any kind of data inside
   the PATH-DATA-TLV.

   If the message is a Query Response, then it must have either a
   RESULT-TLV or a FULLDATA-TLV.




Haleplidis, et al.            Informational                    [Page 14]

RFC 6369            ForCES Implementation Experience      September 2011


   If the message is a Config, it must contain either a FULLDATA-TLV or
   a SPARSEDATA-TLV.

   If the message is a Config Response, it must contain a RESULT-TLV.

   More details regarding message validation can be read in Section 7 of
   the Forwarding and Control Element Separation Protocol [RFC5810].

   Note: When deserializing, implementors must take care to ignore
   padding of TLVs as all must be 32-bit aligned.  The length value in
   TLVs includes the Type and Length (4 bytes) but does not include
   padding.

3.4.3.  Message Serialization



   The same concept can be applied in the message creation process.
   Having the TLVs ready, a developer can go bottom up.  All that is
   required is the serialization function that will transform the TLV
   into bytes ready to be transferred on the network.

   For example, for the creation of a simple query from the CE to the
   FE, all the PATH-DATA-TLVs are created.  Then they will be serialized
   and inserted into an OPER-TLV, which in turn will be serialized and
   inserted into an LFBselect-TLV.  The LFBselect-TLV will then be
   serialized and entered into the Common Header, which will be passed
   to the TML to be transported to the FE.

   Having an array of TLVs inside a TLV that is next in the TLV
   hierarchy allows the developer to insert any number of next-level
   TLVs, thus creating any kind of message.

   Note: When the TLV is serialized to be written on the wire,
   implementors must take care to include padding to TLVs as all must be
   32-bit aligned.

4.  Development Platforms

   Any development platform that can support the SCTP TML and the TML of
   the developer's choosing is available for use.

   Figure 3 provides an initial survey of SCTP support for C/C++ and
   Java at the present time.









Haleplidis, et al.            Informational                    [Page 15]

RFC 6369            ForCES Implementation Experience      September 2011


         /-------------+-------------+-------------+-------------\
         |\ Platform   |             |             |             |
         | ----------\ |   Windows   |    Linux    |   Solaris   |
         |  Language  \|             |             |             |
         +-------------+-------------+-------------+-------------+
         |             |             |             |             |
         |    C/C++    |  Supported  |  Supported  |  Supported  |
         |             |             |             |             |
         +-------------+-------------+-------------+-------------+
         |             |   Limited   |             |             |
         |    Java     | Third Party |  Supported  |  Supported  |
         |             | Not from SUN|             |             |
         \-------------+-------------+-------------+-------------/

                Figure 3: SCTP Support on Operating Systems

   A developer should be aware of some limitations regarding Java
   implementations.

   Java inherently does not support unsigned types.  A workaround can be
   found in the creation of classes that do the translation of unsigned
   types to Java types.  The problem is that the unsigned long cannot be
   used as-is in the Java platform.  The proposed set of classes can be
   found in [JavaUnsignedTypes].

5.  Acknowledgements

   The authors would like to thank Adrian Farrel for sponsoring this
   document and Jamal Hadi Salim for discussions that made this document
   better.

6.  Security Considerations

   Developers of ForCES FEs and CEs must take the Security
   Considerations of the Forwarding and Control Element Separation
   Framework [RFC3746] and the Forwarding and Control Element Separation
   Protocol [RFC5810] into account.

   Also, as specified in the Security Considerations section of the
   SCTP-Based Transport Mapping Layer (TML) for the Forwarding and
   Control Element Separation Protocol [RFC5811], transport-level
   security has to be ensured by IPsec.









Haleplidis, et al.            Informational                    [Page 16]

RFC 6369            ForCES Implementation Experience      September 2011


7.  References



7.1.  Normative References



   [RFC5810]  Doria, A., Hadi Salim, J., Haas, R., Khosravi, H., Wang,
              W., Dong, L., Gopal, R., and J. Halpern, "Forwarding and
              Control Element Separation (ForCES) Protocol
              Specification", RFC 5810, March 2010.

   [RFC5811]  Hadi Salim, J. and K. Ogawa, "SCTP-Based Transport Mapping
              Layer (TML) for the Forwarding and Control Element
              Separation (ForCES) Protocol", RFC 5811, March 2010.

   [RFC5812]  Halpern, J. and J. Hadi Salim, "Forwarding and Control
              Element Separation (ForCES) Forwarding Element Model",
              RFC 5812, March 2010.

   [RFC6041]  Crouch, A., Khosravi, H., Doria, A., Wang, X., and K.
              Ogawa, "Forwarding and Control Element Separation (ForCES)
              Applicability Statement", RFC 6041, October 2010.

   [RFC6053]  Haleplidis, E., Ogawa, K., Wang, W., and J. Hadi Salim,
              "Implementation Report for Forwarding and Control Element
              Separation (ForCES)", RFC 6053, November 2010.

7.2.  Informative References



   [JavaUnsignedTypes]
              "Java Unsigned Types",
              <http://nam.ece.upatras.gr/index.php?q=node/44>.

   [RFC3654]  Khosravi, H. and T. Anderson, "Requirements for Separation
              of IP Control and Forwarding", RFC 3654, November 2003.

   [RFC3746]  Yang, L., Dantu, R., Anderson, T., and R. Gopal,
              "Forwarding and Control Element Separation (ForCES)
              Framework", RFC 3746, April 2004.














Haleplidis, et al.            Informational                    [Page 17]

RFC 6369            ForCES Implementation Experience      September 2011


Authors' Addresses



   Evangelos Haleplidis
   University of Patras
   Department of Electrical & Computer Engineering
   Patras  26500
   Greece

   EMail: ehalep@ece.upatras.gr


   Odysseas Koufopavlou
   University of Patras
   Department of Electrical & Computer Engineering
   Patras  26500
   Greece

   EMail: odysseas@ece.upatras.gr


   Spyros Denazis
   University of Patras
   Department of Electrical & Computer Engineering
   Patras  26500
   Greece

   EMail: sdena@upatras.gr
























Haleplidis, et al.            Informational                    [Page 18]