Network Address Translation (NAT) has gained popularity as a method
of separating public and private address spaces, and alleviating
network address shortages. A NAT translates the addresses of packets
leaving a first routing realm to an address from a second routing
realm, and performs the reverse function for packets entering the
first routing realm from the second routing realm. This translation
is performed transparently to the hosts in either space, and may
include modification of TCP/UDP port numbers and IP addresses in
packets that traverse the NAT.
While a NAT does not require hosts to be aware of the translation, it
will require an application layer gateway (ALG) for any protocol that
transmits IP addresses or port numbers in packet payloads (such as
FTP). Additionally, a NAT will not work with protocols that require
IP addresses and ports to remain unmodified between the source and
destination hosts, or protocols that prevent such modifications from
occurring (such as some IPsec modes, or application-layer end-to-end
encryption).
An alternative to a NAT is an architecture that allows the hosts
within the first (e.g., private) routing realm to directly use
addresses and other routing parameters from the second (e.g., public)
routing realm. Thus, RSIP [RSIP-FRAME] has been defined as a method
for address sharing that exhibits more transparency than NAT. In
particular, RSIP requires that an RSIP gateway (a router or gateway
between the two realms) assign at least one address from the second
routing realm, and perhaps some other resources, to each RSIP host.
An RSIP host is a host in the first routing realm that needs to
establish end-to-end connectivity to a host, entity or device in the
second routing realm. Thus, the second routing realm is not directly
Borella, et al. Experimental [Page 3]
RFC 3103 RSIP Protocol Specification October 2001
accessible from the RSIP host, but this system allows packets to
maintain their integrity from the RSIP host to their destination.
ALGs are not required in the RSIP gateway.
RSIP requires that hosts be modified so that they place some number
of layer three, layer four or other values from those assigned by the
RSIP gateway in each packet bound for the second routing realm.
This document discusses a method for assigning parameters to an RSIP
host from an RSIP gateway. The requirements, scope, and
applicability of RSIP, as well as its interaction with other layer 3
protocols, are discussed in a companion framework document [RSIP-
FRAME]. Extensions to this protocol that enable end-to-end IPsec are
discussed in [RSIP-IPSEC].
The keywords "MUST", "MUST NOT", "REQUIRED", "SHOULD", "SHOULD NOT",
"SHALL", "SHALL NOT", "MAY" and "MAY NOT" that appear in this
document are to be interpreted as described in [RFC2119].
Private Realm
A routing realm that uses private IP addresses from the ranges
(10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16) specified in
[RFC1918], or addresses that are non-routable from the Internet.
Public Realm
A routing realm with unique network addresses assigned by the
Internet Assigned Number Authority (IANA) or an equivalent address
registry.
RSIP Host
A host within the private realm that acquires publicly unique
parameters from an RSIP gateway through the use of the RSIP
client/server protocol.
RSIP Gateway
A router situated on the boundary between a private realm and a
public realm and owns one or more public IP addresses. An RSIP
gateway is responsible for public parameter management and
assignment to RSIP hosts. An RSIP gateway may act as a NAT router
for hosts within the private realm that are not RSIP enabled.
Borella, et al. Experimental [Page 4]
RFC 3103 RSIP Protocol Specification October 2001
RSIP Client
An application program that performs the client portion of the
RSIP client/server protocol. An RSIP client application MUST
exist on all RSIP hosts, and MAY exist on RSIP gateways.
RSIP Server
An application program that performs the server portion of the
RSIP client/server protocol. An RSIP server application MUST
exist on all RSIP gateways.
RSA-IP: Realm Specific Address IP
An RSIP method in which each RSIP host is allocated a unique IP
address from the public realm. Discussed in detail in [RSIP-
FRAME]
RSAP-IP: Realm Specific Address and Port IP
An RSIP method in which each RSIP host is allocated an IP address
(possibly shared with other RSIP hosts) and some number of per-
address unique ports from the public realm. Discussed in detail
in [RSIP-FRAME]
Binding
An association of some combination of a local address, one or more
local ports, a remote address, and a remote port with an RSIP
host.
Resource
A general way to refer to an item that an RSIP host leases from an
RSIP gateway; e.g., an address or port.
All other terminology found in this document is consistent with that
of [RFC2663] and [RSIP-FRAME].
For simplicity, in the remainder of this document we will assume that
the RSIP hosts in the first routing realm (network) use private
(e.g., see [RFC1918]) IP addresses, and that the second routing realm
(network) uses public IP addresses. (This assumption is made without
loss of generality and the ensuing discussion applies to more general
Borella, et al. Experimental [Page 5]
RFC 3103 RSIP Protocol Specification October 2001
cases.) The RSIP gateway connects the public and private realms and
contains interfaces to both. Other NAT terminology found in this
document is defined in [RFC2663].
The diagram below describes an exemplary reference architecture for
RSIP.
RSIP Host RSIP Gateway Host
Xa Na Nb Yb
[X]------( Addr sp. A )----[N]-----( Addr sp. B )-------[Y]
( Network ) ( Network )
Hosts X and Y belong to different addressing realms A and B,
respectively, and N is an RSIP gateway (which may also perform NAT
functions). N has two interfaces: Na on address space A, and Nb on
address space B. N may have a pool of addresses in address space B
which it can assign to or lend to X and other hosts in address space
A. These addresses are not shown above, but they can be denoted as
Nb1, Nb2, Nb3 and so on.
Host X, needing to establish an end-to-end connection to a network
entity Y situated within address space B, first negotiates and
obtains assignment of the resources from the RSIP gateway. Upon
assignment of these parameters, the RSIP gateway creates a mapping,
of X's addressing information and the assigned resources. This
binding enables the RSIP gateway to correctly de-multiplex and
forward inbound traffic generated by Y for X. A lease time is
associated with each bind.
Using the public parameters assigned by the RSIP gateway, RSIP hosts
tunnel data packets across address space A to the RSIP gateway. The
RSIP gateway acts as the end point of such tunnels, stripping off the
outer headers and routing the inner packets onto the public realm.
As mentioned above, an RSIP gateway maintains a mapping of the
assigned public parameters as demultiplexing fields for uniquely
mapping them to RSIP host private addresses. When a packet from the
public realm arrives at the RSIP gateway and it matches a given set
of demultiplexing fields, then the RSIP gateway will tunnel it to the
appropriate RSIP host. The tunnel headers of outbound packets from X
to Y, given that X has been assigned Nb, are as follows:
+---------+---------+---------+
| X -> Na | Nb -> Y | payload |
+---------+---------+---------+
Borella, et al. Experimental [Page 6]
RFC 3103 RSIP Protocol Specification October 2001
There are two basic flavors of RSIP: RSA-IP and RSAP-IP. RSIP hosts
and gateways MUST support RSAP-IP and MAY support RSA-IP. Details of
RSA-IP and RSAP-IP are found in [RSIP-FRAME].
RSIP is an application layer protocol that requires the use of a
transport layer protocol for end-to-end delivery of packets.
RSIP gateways MUST support TCP, and SHOULD support UDP. Due to the
fact that RSIP may be deployed across a wide variety of network
links, RSIP hosts SHOULD support TCP, because of TCP's robustness
across said variety of links. However, RSIP hosts MAY support UDP
instead of TCP, or both UDP and TCP.
For RSIP hosts and gateways using UDP, timeout and retransmissions
MUST occur. We recommend a binary exponential backoff scheme with an
initial duration of 12.5 ms, and a maximum of six retries (seven
total attempts before failure). However, these parameters MAY be
adjusted or tuned for specific link types or scenarios.
Once a host and gateway have established a registration using either
TCP or UDP, they may not switch between the two protocols for the
duration of the registration. The decision of whether to use TCP or
UDP is made by the client, and is determined by the transport
protocol of the first packet sent by a client in a successful
registration procedure.
An RSIP host can be in exactly one of three fundamental relationships
with respect to an RSIP gateway:
Unregistered: The RSIP gateway does not know of the RSIP host's
existence, and it will not forward or deliver globally addressed
packets on behalf of the host. The only valid RSIP-related action
for an RSIP host to perform in this state is to request
registration with an RSIP gateway.
Registered: The RSIP gateway knows of the RSIP host and has assigned
it a client ID and has specified the flow policies that it
requires of the host. However, no resources, such as addresses or
ports, have been allocated to the host, and the gateway will not
forward or deliver globally addressed packets on behalf of the
host. All registrations have an associated lease time. If this
lease time expires, the RSIP host automatically reverts to the
unregistered state.
Borella, et al. Experimental [Page 7]
RFC 3103 RSIP Protocol Specification October 2001
Assigned: The RSIP gateway has granted one or more bindings of
resources to the host. The gateway will forward and deliver
globally addressed packets on behalf of the host. Each binding
has an associated lease time. If this lease time expires, the
binding is automatically revoked.
Architectures in which an RSIP host is simultaneously registered with
more than one RSIP gateway are possible. In such cases, an RSIP host
may be in different relationships with different RSIP gateways at the
same time.
An RSIP gateway MAY redirect an RSIP host to use a tunnel endpoint
for data traffic that is not the RSIP gateway itself, or perhaps is a
different interface on the RSIP gateway. This is done by specifying
the tunnel endpoint's address as part of an assignment. In such an
architecture, it is desirable (though not necessary) for the RSIP
gateway to have a method with which to notify the tunnel endpoint of
assignments, and the expiration status of these assignments.
Lease times for bindings and registrations are managed as follows.
All lease times are given in units of seconds from the current time,
indicating a time in the future at which the lease will expire.
These expiration times are used in the ensuing discussion.
An initial expiration time (R) is given to a registration. Under
this registration, multiple bindings may be established, each with
their own expiration times (B1, B2, ...). When each binding is
established or extended, the registration expiration time is adjusted
so that the registration will last at least as long as the longest
lease. In other words, when binding Bi is established or extended,
the following calculation is performed: R = max(R, Bi).
Under this scheme, a registration will never expire while any
binding's lease is still valid. However, a registration may expire
when the last binding's lease expires, or at some point thereafter.
Since an RSIP gateway is likely to reside on the boundary between two
or more different administrative domains, it is desirable to enable
an RSIP gateway to be able to enforce flow-based policy. In other
words, an RSIP gateway should have the ability to explicitly control
which local addresses and ports are used to communicate with remote
addresses and ports.
In the following, macro-flow policy refers to controlling flow policy
at the granularity level of IP addresses, while micro-flow policy
refers to controlling flow policy at the granularity of IP address
Borella, et al. Experimental [Page 8]
RFC 3103 RSIP Protocol Specification October 2001
and port tuples. Of course there may be no policy at all, which
indicates that the RSIP gateway does not care about the flow
parameters used by RSIP hosts. We consider two levels of local flow
policy and three levels of remote flow policy.
Local flow policy determines the granularity of control that an RSIP
gateway has over the local addressing parameters that an RSIP host
uses for particular sessions.
Since an RSIP host must use at least an IP address allocated by the
gateway, the loosest level of local flow policy is macro-flow based.
Under local macro-flow policy, an RSIP host is allocated an IP
address (RSA-IP) or an IP address and one or more ports to use with
it (RSAP-IP). However, the host may use the ports as it desires for
establishing sessions with public hosts.
Under micro-flow policy, a host is allocated exactly one port at a
time. The host may request more ports, also one at a time. This
policy gives the gateway very tight control over local port use,
although it affords the host less flexibility.
Note that only local macro-flow policy can be used with RSA-IP, while
either local macro-flow or local micro-flow policy may be used with
RSAP-IP.
Examples of how RSIP flow policy operates are given in Appendix C.
Remote flow policy determines the granularity of control that an RSIP
gateway has over the remote (public) hosts with which an RSIP host
communicates. In particular, remote flow policy dictates what level
of detail that a host must specify addressing parameters of a remote
host or application before the RSIP gateway allows the host to
communicate with that host or application.
The simplest and loosest form of flow policy is no policy at all. In
other words, the RSIP gateway allocates addressing parameters to the
host, and the host may use these parameters to communicate with any
remote host, without explicitly notifying the gateway.
Macro-flow policy requires that the host identify the remote address
of the host that it wishes to communicate with as part of its request
for local addressing parameters. If the request is granted, the host
MUST use the specified local parameters only with the remote address
specified, and MUST NOT communicate with the remote address using any
Borella, et al. Experimental [Page 9]
RFC 3103 RSIP Protocol Specification October 2001
local parameters but the ones allocated. However, the host may
contact any port number at the remote host without explicitly
notifying the gateway.
Micro-flow policy requires that the host identify the remote address
and port of the host that it wishes to communicate with as part of
its request for local addressing parameters. If the request is
granted, the host MUST use the specified local parameters only with
the remote address and port specified, and MUST NOT communicate with
the remote address and port using any local parameters but the ones
allocated.
Remote flow policy is implemented in both the ingress and egress
directions, with respect to the location of the RSIP gateway.
An RSIP gateway must maintain state for all RSIP hosts and their
assigned resources. The amount and type of state maintained depends
on the local and remote flow policy. The required RSIP gateway state
will vary based on the RSIP method, but will always include the
chosen method's demultiplexing parameters.
An RSIP gateway serving an RSIP host using the RSA-IP method MUST
maintain the following minimum state to ensure proper mapping of
incoming packets to RSIP hosts:
- Host's private address
- Host's assigned public address(es)
An RSIP gateway serving an RSIP host using the RSAP-IP method MUST
maintain the following minimum state to ensure proper mapping of
incoming packets to RSIP hosts:
- Host's private address
- Host's assigned public address(es)
- Host's assigned port(s) per address
Regardless of whether the gateway is using RSA-IP or RSAP-IP,
additional state is necessary if either micro-flow based or macro-
flow based remote policy is used.
Borella, et al. Experimental [Page 10]
RFC 3103 RSIP Protocol Specification October 2001
If the gateway is using macro-flow based remote policy, the following
state must be maintained:
- Remote host's address
If the gateway is using micro-flow based remote policy, the following
state must be maintained:
- Remote host's address
- Remote host's port
More state MAY be used by an RSIP gateway if desired. For example,
ToS/DS bytes may be recorded in order to facilitate quality of
service support.
In this section we define the formats for RSIP parameters. Each RSIP
message contains one or more parameters that encode the information
passed between the host and gateway. The general format of all
parameters is TLV (type-length-value) consisting of a 1-byte type
followed by a 2-byte length followed by a 'length' byte value as
shown below.
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length | Value |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Value ...
+-+-+-+-+-+-+-+-+-+-+-+
The value field may be divided into a number of other fields as per
the type of the parameter. Note that the length field encodes the
number of bytes in the value field, NOT the overall number of bytes
in the parameter.
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type = 1 | Length | Addrtype |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Address...
+-+-+-+-+-+-+-+-+-+-+-+
Borella, et al. Experimental [Page 11]
RFC 3103 RSIP Protocol Specification October 2001
The address parameter contains addressing information, either an IPv4
address or netmask, an IPv6 address or netmask, or a fully qualified
domain name (FQDN). The Addrtype field is 1 byte in length,
indicating the type of address.
Addrtype Length of address field (in bytes)
---- --------------------------------
0 Reserved 0
1 IPv4 4
2 IPv4 netmask 4
3 IPv6 16
4 FQDN varies
For FQDN (Fully qualified domain name), the length of the address
field will be one less than the value of the length field, and the
name will be represented as an ASCII string (no terminating
character).
In some cases, it is necessary to specify a "don't care" value for an
address. This is signified by a setting the length field to 1 and
omitting the value field.
It is not valid for a host to request an address with an FQDN type as
its local address (See specification of ASSIGN_REQUEST_RSA-IP and
ASSIGN_REQUEST_RSAP-IP, below).
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type = 2 | Length | Number |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Port number | ...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The ports parameter encodes zero or more TCP or UDP ports. When a
single port is specified, the value of the number field is 1 and
there is one port field following the number field. When more than
one port is specified, the value of the number field will indicate
the total number of ports contained, and the parameter may take one
of two forms. If there is one port field, the ports specified are
considered to be contiguous starting at the port number specified in
the port field. Alternatively, there may be a number of port fields
equal to the value of the number field. The number of port fields
can be extrapolated from the length field.
Borella, et al. Experimental [Page 12]
RFC 3103 RSIP Protocol Specification October 2001
In some cases, it is necessary to specify a don't care value for one
or more ports (e.g., when a client application is using ephemeral
source ports). This is accomplished by setting the length field to
1, setting the number field to the number of ports necessary, and
omitting all port fields. The value of the number field MUST be
greater than or equal to one.
If micro-flow based policy applies to a given ports parameter, it
MUST contain exactly one port field.
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type = 10 | Length = 2 | Value |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Value |
+-+-+-+-+-+-+-+-+
An indicator parameter is a general-purpose parameter, the use of
which is defined by the message that it appears in. An RSIP message
that uses an indicator parameter MUST define the meaning and
interpretation of all of the indicator's possible values.
Borella, et al. Experimental [Page 15]
RFC 3103 RSIP Protocol Specification October 2001
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type = 12 | Length | Vendor ID |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Vendor ID | Subtype | Value...
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The vendor specific parameter is used to encode parameters that are
defined by a particular vendor. The vendor ID field is the vendor-
specific ID assigned by IANA. Subtypes are defined and used by each
vendor as necessary. An RSIP host or gateway SHOULD silently ignore
vendor-specific messages that it does not understand.
RSIP messages consist of three mandatory fields, version, message
type, and overall length, followed by one or more required
parameters, followed in turn by zero or more optional parameters. In
an RSIP message, all required parameters MUST appear in the exact
order specified below. Optional parameters MAY appear in any order.
Message format is shown below:
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version | Message type | Overall length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Parameters...
+-+-+-+-+-+-+-+-+-+-+
Borella, et al. Experimental [Page 16]
RFC 3103 RSIP Protocol Specification October 2001
The version number field is a single byte and specifies the RSIP
version number that is being used. The current RSIP version number
is 1.
The message type field is a single byte and specifies the message
contained in the current packet. There may be only one message per
packet. Message types are given numerical assignments in Appendix B.
The overall length field is two bytes and contains the number of
bytes in the RSIP message, including the three mandatory fields.
Most parameters are only allowed to appear once in each message. The
exceptions are as follows:
- Multiple address parameters MUST appear in ASSIGN_REQUEST_RSA-
IP, ASSIGN_RESPONSE_RSA-IP, ASSIGN_REQUEST_RSAP-IP,
ASSIGN_RESPONSE_RSAP-IP, LISTEN_REQUEST and LISTEN_RESPONSE.
- Multiple ports parameters MUST appear in ASSIGN_REQUEST_RSAP-
IP, ASSIGN_RESPONSE_RSAP-IP, LISTEN_REQUEST and
LISTEN_RESPONSE.
- Multiple RSIP method and tunnel type parameters MAY appear in
RESISTER_RESPONSE.
- Multiple address parameters and multiple indicator parameters
MAY appear in QUERY_REQUEST and QUERY_RESPONSE.
The following message types are defined in BNF. Required parameters
are enclosed in <> and MUST appear. Optional parameters are enclosed
in [] and MAY appear. Not all message types need to be implemented
in order to be RSIP compliant. For example, an RSIP host and/or
gateway may not support LISTEN_REQUEST and LISTEN_RESPONSE, or may
only support RSAP-IP and not RSA-IP.
An ERROR_RESPONSE is used to provide error messages from an RSIP
gateway to an RSIP host. Usually, errors indicate that the RSIP
gateway cannot or will not perform an action or allocate resources on
behalf of the host. If the error is related to a particular client
ID or bind ID, these associated parameters MUST be included.
Multiple errors MAY NOT be reported in the same ERROR_RESPONSE. In
situations where more than one error has occurred, the RSIP gateway
MUST choose only one error to report.
Borella, et al. Experimental [Page 17]
RFC 3103 RSIP Protocol Specification October 2001
An ERROR_RESPONSE message MUST only be transmitted by an RSIP
gateway. An RSIP host that detects an error in a message received
from an RSIP gateway MUST silently discard the message. There are no
error conditions that can be caused by an ERROR_RESPONSE. An
ERROR_RESPONSE is typically transmitted in response to a request from
an RSIP host, but also may be transmitted asynchronously by an RSIP
gateway.
The REGISTER_REQUEST message is used by an RSIP host to establish
registration with an RSIP gateway. An RSIP host MUST register before
it requests resources or services from an RSIP gateway. Once an RSIP
host has registered with an RSIP gateway, it may not register again
until it has de-registered from that gateway.
The following message-specific error conditions exist:
- If the host is already registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
ALREADY_REGISTERED error and the RSIP host's client ID.
- If the gateway's policy will not allow the host to register,
the gateway MUST respond with an ERROR_RESPONSE containing the
REGISTRATION_DENIED error.
Borella, et al. Experimental [Page 18]
RFC 3103 RSIP Protocol Specification October 2001
The REGISTER_RESPONSE message is used by an RSIP gateway to confirm
the registration of an RSIP host, and to provide a client ID, flow
policy, and possibly a message counter and one or more RSIP methods
and/or tunnel types.
An RSIP gateway MUST assign a different client ID to each host that
is simultaneously registered with it. The RSIP gateway MAY respond
with one or more RSIP methods and tunnel types that it supports. If
an RSIP method is not specified, RSAP-IP MUST be assumed. If a
tunnel type is not specified, IP-IP MUST be assumed.
The DE-REGISTER_REQUEST message is used by an RSIP host to de-
register with an RSIP gateway. If a host de-registers from the
assigned state, all of the host's bindings are revoked. The host
SHOULD NOT de-register from the unregistered state.
The following message-specific error conditions exist:
- If the host is not registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
REGISTER_FIRST error.
- If the message contains an incorrect client ID, the gateway
MUST respond with an ERROR_RESPONSE containing the
BAD_CLIENT_ID error.
If there are no errors that result from this message, the gateway
MUST respond with an appropriate DE-REGISTER_RESPONSE. Upon de-
registering a host, an RSIP gateway must delete all binds associated
with that host and return their resources to the pool of free
resources. Once a host has de-registered, it may not use any of the
RSIP gateway's resources without registering again.
The DE-REGISTER_RESPONSE message is used by an RSIP gateway to
confirm the de-registration of an RSIP host or to force an RSIP host
to relinquish all of its bindings and terminate its relationship with
the RSIP gateway. Upon receiving a DE-REGISTER_RESPONSE message, an
RSIP host MUST stop all use of the resources that have been allocated
to it by the gateway.
An RSIP gateway MUST send a DE-REGISTER_RESPONSE in response to a
valid DE-REGISTER_REQUEST. An RSIP gateway MUST send a DE-
REGISTER_RESPONSE to an RSIP host when that host's registration lease
time times out. An RSIP gateway SHOULD send a DE-REGISTER_RESPONSE
if it detects that it will no longer be able to perform RSIP
functionality for a given host. An RSIP host MUST be ready to accept
a DE-REGISTER_RESPONSE at any moment.
Borella, et al. Experimental [Page 20]
RFC 3103 RSIP Protocol Specification October 2001
The ASSIGN_REQUEST_RSA-IP message is used by an RSIP host to request
resources to use with RSA-IP. Note that RSA-IP cannot be used in
combination with micro-flow based local policy.
The RSIP host specifies two address parameters. The RSIP host may
request a particular local address by placing that address in the
first address parameter. To indicate that it has no preference for
local address, the RSIP host may place a "don't care" value in the
address parameter.
If macro-flow based remote policy is used, the host MUST specify the
remote address that it will use this binding (if granted) to contact;
however, the remote port number MAY remain unspecified. If micro-
flow based remote policy is used, the host MUST specify the remote
address and port number that it will use this binding (if granted) to
contact. If no flow policy is used, the RSIP host may place a "don't
care" value in the value fields of the respective address and ports
parameters.
The following message-specific error conditions exist:
- If the host is not registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
REGISTER_FIRST error.
- If the message contains an incorrect client ID, the gateway
MUST respond with an ERROR_RESPONSE containing the
BAD_CLIENT_ID error.
Borella, et al. Experimental [Page 21]
RFC 3103 RSIP Protocol Specification October 2001
- If the local address parameter is a don't care value and the
RSIP gateway cannot allocate ANY addresses, the RSIP gateway
MUST respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNAVAILABLE error.
- If the local address parameter is not a don't care value there
are three possible error conditions:
o If the RSIP gateway cannot allocate ANY addresses, it MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNAVAILABLE error.
o If the RSIP gateway cannot allocate the requested address
because it is in use, the RSIP gateway MUST respond with an
ERROR_RESPONSE containing the LOCAL_ADDR_INUSE error.
o If the RSIP gateway cannot allocate the requested address
because it is not allowed by policy, the RSIP gateway MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNALLOWED error.
- If macro-flow based remote policy is used and the requested
remote address is not allowed by the RSIP gateway's policy, the
RSIP gateway MUST respond with an ERROR_RESPONSE containing the
REMOTE_ADDR_UNALLOWED error.
- If micro-flow based remote policy is used and the requested
remote address / port pair is not allowed by the RSIP gateway's
policy, the RSIP gateway MUST respond with an ERROR_RESPONSE
containing the REMOTE_ADDRPORT_UNALLOWED error.
- If an unsupported or unallowed tunnel type is specified, the
RSIP gateway MUST respond with an ERROR_RESPONSE containing the
BAD_TUNNEL_TYPE error.
- If the host has not specified local or remote address or port
information in enough detail, the RSIP gateway MUST respond
with an ERROR_RESPONSE containing the FLOW_POLICY_VIOLATION
error.
The ASSIGN_RESPONSE_RSA-IP message is used by an RSIP gateway to
deliver parameter assignments to an RSIP host using RSA-IP. A host-
wise unique bind ID, lease time, and tunnel type must be provided for
every assignment.
Borella, et al. Experimental [Page 22]
RFC 3103 RSIP Protocol Specification October 2001
If no remote flow policy is used, the RSIP gateway MUST use "don't
care" values for the remote address and ports parameters. If macro-
flow based remote policy is used, the remote address parameter MUST
contain the address specified in the associated request, and the
remote ports parameter MUST contain a "don't care" value. If micro-
flow based remote policy is used, the remote address and remote ports
parameters MUST contain the address and port information specified in
the associated request.
If the host detects an error or otherwise does not "understand" the
gateway's response, it SHOULD send a FREE_REQUEST with the bind ID
from the said ASSIGN_RESPONSE_RSA-IP. This will serve to help
synchronize the states of the host and gateway.
The address of a tunnel endpoint that is not the RSIP gateway MAY be
specified. If this parameter is not specified, the RSIP gateway MUST
be assumed to be the tunnel endpoint.
The ASSIGN_REQUEST_RSAP-IP message is used by an RSIP host to request
resources to use with RSAP-IP. The RSIP host specifies two address
and two port parameters, the first of each, respectively, refer to
the local address and port(s) that will be used, and the second of
each, respectively, refer to the remote address and port(s) that will
be contacted.
Borella, et al. Experimental [Page 23]
RFC 3103 RSIP Protocol Specification October 2001
An RSIP host may request a particular local address by placing that
address in the value field of the first address parameter. The RSIP
host may request particular local ports by placing them in the first
port parameter. To indicate that it has no preference for local
address or ports, the RSIP host may place a "don't care" value in the
respective address or ports parameters.
If macro-flow based remote policy is used, the host MUST specify the
remote address that it will use this binding (if granted) to contact;
however, the remote port number(s) MAY remain unspecified. If
micro-flow based remote policy is used, the host MUST specify the
remote address and port number(s) that it will use this binding (if
granted) to contact. If no flow policy is used, the RSIP host may
place a value of all 0's in the value fields of the respective
address or port parameters.
The following message-specific error conditions exist:
- If the host is not registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
REGISTER_FIRST error.
- If the message contains an incorrect client ID, the gateway
MUST respond with an ERROR_RESPONSE containing the
BAD_CLIENT_ID error.
- If the local address parameter is a don't care value and the
RSIP gateway cannot allocate ANY addresses, the RSIP gateway
MUST respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNAVAILABLE error.
Borella, et al. Experimental [Page 24]
RFC 3103 RSIP Protocol Specification October 2001
- If the local address parameter is not a don't care value there
are five possible error conditions:
o If the RSIP gateway cannot allocate ANY addresses, it MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNAVAILABLE error.
o If the RSIP gateway cannot allocate the requested address
because it is in use, the RSIP gateway MUST respond with an
ERROR_RESPONSE containing the LOCAL_ADDR_INUSE error.
o If the RSIP gateway cannot allocate the requested address
because it is not allowed by policy, the RSIP gateway MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNALLOWED error.
o If the RSIP gateway cannot allocate a requested address /
port tuple because it is in use, the RSIP gateway MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDRPORT_INUSE error.
o If the RSIP gateway cannot allocate a requested address /
port tuple because it is not allowed by policy, the RSIP
gateway MUST respond with an ERROR_RESPONSE containing the
LOCAL_ADDRPORT_UNALLOWED error.
- If the RSIP host requests a number of ports (greater that one),
but does not specify particular port numbers (i.e., uses "don't
care" values) the RSIP gateway cannot grant the entire request,
the RSIP gateway MUST return an ERROR_RESPONSE containing the
LOCAL_ADDRPORT_UNAVAILABLE error.
- If macro-flow based remote policy is used and the requested
remote address is not allowed by the RSIP gateway's policy, the
RSIP gateway MUST respond with an ERROR_RESPONSE containing the
REMOTE_ADDR_UNALLOWED error.
- If micro-flow based remote policy is used and the requested
remote address / port pair is not allowed by the RSIP gateway's
policy, the RSIP gateway MUST respond with an ERROR_RESPONSE
containing the REMOTE_ADDRPORT_UNALLOWED error.
- If an unsupported or unallowed tunnel type is specified, the
RSIP gateway MUST respond with an ERROR_RESPONSE containing the
BAD_TUNNEL_TYPE error.
Borella, et al. Experimental [Page 25]
RFC 3103 RSIP Protocol Specification October 2001
- If the host has not specified local or remote address or port
information in enough detail, the RSIP gateway MUST respond
with an ERROR_RESPONSE containing the FLOW_POLICY_VIOLATION
error.
The ASSIGN_RESPONSE_RSAP-IP message is used by an RSIP gateway to
deliver parameter assignments to an RSIP host. A host-wise unique
bind ID, lease time, and tunnel type must be provided for every
assignment.
Regardless of local flow policy, a local address and port(s) MUST be
assigned to the host. If macro-flow based local policy is used, the
host is assigned an address and one or more ports. If micro-flow
based local policy is used, the host is assigned an address and
exactly one port.
If no remote flow policy is used, the RSIP gateway MUST use "don't
care" values for the remote address and ports parameters. If macro-
flow based remote policy is used, the remote address parameter MUST
contain the address specified in the associated request, and the
remote ports parameter must contain a "don't care" value. If micro-
flow based remote policy is used, the remote address and remote ports
parameters MUST contain the address and port information specified in
the associated request.
Borella, et al. Experimental [Page 26]
RFC 3103 RSIP Protocol Specification October 2001
If the host detects an error or otherwise does not "understand" the
gateway's response, it SHOULD send a FREE_REQUEST with the bind ID
from the said ASSIGN_RESPONSE_RSAP-IP. This will serve to help
synchronize the states of the host and gateway.
The address of a tunnel endpoint that is not the RSIP gateway MAY be
specified. If this parameter is not specified, the RSIP gateway MUST
be assumed to be the tunnel endpoint.
The EXTEND_REQUEST message is used to request a lease extension to a
current bind. It may be used with both RSA-IP and RSAP-IP. The host
MUST specify its client ID and the bind ID in question, and it MAY
suggest a lease time to the gateway.
The following message-specific error conditions exist:
- If the host is not registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
REGISTER_FIRST error.
- If the message contains an incorrect client ID, the gateway
MUST respond with an ERROR_RESPONSE containing the
BAD_CLIENT_ID error.
- If the message contains an incorrect bind ID, the gateway MUST
respond with an ERROR_RESPONSE containing the BAD_BIND_ID
error.
Borella, et al. Experimental [Page 27]
RFC 3103 RSIP Protocol Specification October 2001
If the RSIP gateway grants an extension to the host's lease, it MUST
RESPOND with an appropriate EXTEND_RESPONSE message. If the lease is
not renewed, the RSIP gateway MAY let it implicitly expire by doing
nothing or make it explicitly expire by sending an appropriate
FREE_RESPONSE message.
The EXTEND_RESPONSE message is used by an RSIP gateway to grant a
requested lease extension. The gateway MUST specify the client ID of
the host, the bind ID in question, and the new assigned lease time.
The RSIP gateway will determine lease time as per its local policy.
The returned time is to be interpreted as the number of seconds
before the lease expires, counting from the time at which the message
is sent/received.
The FREE_REQUEST message is used by an RSIP host to free a binding.
The given bind ID identifies the bind to be freed. Resources may
only be freed using the granularity of a bind ID.
The following message-specific error conditions exist:
- If the host is not registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
REGISTER_FIRST error.
- If the message contains an incorrect client ID, the gateway
MUST respond with an ERROR_RESPONSE containing the
BAD_CLIENT_ID error.
- If the message contains an incorrect bind ID, the gateway MUST
respond with an ERROR_RESPONSE containing the BAD_BIND_ID
error.
If a host receives an error in response to a FREE_REQUEST, this may
indicate that the host and gateway's states have become
unsynchronized. Therefore, the host SHOULD make an effort to
resynchronize, such as freeing resources then re-requesting them, or
de-registering then re-registering.
The FREE_RESPONSE message is used by an RSIP gateway to acknowledge a
FREE_REQUEST sent by an RSIP host, and to asynchronously deallocate
resources granted to an RSIP host.
An RSIP host must always be ready to accept a FREE_RESPONSE, even if
its lease on the specified bind ID is not yet expired.
Borella, et al. Experimental [Page 29]
RFC 3103 RSIP Protocol Specification October 2001
A QUERY_REQUEST message is used by an RSIP host to ask an RSIP
gateway whether or not a particular address or network is local or
remote. The host uses this information to determine whether to
contact the host(s) directly (in the local case), or via RSIP (in the
remote case).
This message defines an indicator parameter with a 1-byte value field
and 2 defined values:
- 1 address
- 2 network
One or more address or network tuples may be specified. Each tuple
encodes a request regarding the locality (local or remote) of the
encoded address or network. If no tuple is specified, the RSIP
gateway should interpret the message as a request for all tuples that
it is willing to provide. Note that the FQDN form of the address
parameter cannot be used to specify the address of a network, and
only the netmask form of the address parameter can be used to specify
the netmask of a network.
If an RSIP gateway cannot determine whether a queried host or network
is local or remote, it SHOULD transmit a QUERY_RESPONSE with no
response specified for the said host or network.
Borella, et al. Experimental [Page 30]
RFC 3103 RSIP Protocol Specification October 2001
The following message-specific error conditions exist:
- If the host is not registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
REGISTER_FIRST error.
- If the message contains an incorrect client ID, the gateway
MUST respond with an ERROR_RESPONSE containing the
BAD_CLIENT_ID error.
A QUERY_RESPONSE message is used by an RSIP gateway to answer a
QUERY_REQUEST from an RSIP host.
This message defines an indicator parameter with a 1-byte value field
and 4 defined values:
- 1 local address
- 2 local network
- 3 remote address
- 4 remote network
An RSIP gateway has some leeway in how it responds to a
QUERY_REQUEST. It may just provide the information requested, if it
can provide such information. It may provide its complete list of
address and networks, in order to minimize the number of requests
that the host needs to perform in the future. How an RSIP gateway
responds may depend on network traffic considerations as well.
If an RSIP gateway sends a QUERY_RESPONSE that does not contain any
tuples, or a QUERY_RESPONSE that does not contain a tuple that
applies to an associated tuple in the associated QUERY_REQUEST, this
should be interpreted that the RSIP gateway does not know whether the
queried host or network is local or remote. Appropriate host
behavior upon receipt of such a message is to assume that the queried
host or network is remote.
Note that an RSIP gateway is not expected to maintain a complete list
of all remote hosts and networks. In fact, a typical RSIP gateway
will only maintain a list of the networks and hosts that it knows are
local (private with respect to the RSIP host).
A LISTEN_REQUEST message is sent by an RSIP host that wants to
register a service on a particular address and port number. The host
must include its client ID, local address parameter and ports
parameters, and remote address and ports parameters. The client MAY
suggest a lease time and one or more tunnel types.
Borella, et al. Experimental [Page 32]
RFC 3103 RSIP Protocol Specification October 2001
If the host wants to listen on a particular address or port, it may
specify these in the address and ports parameters. Otherwise it may
leave one or both of these parameters with "don't care" values.
If no remote flow policy is being used, the host MUST fill both the
remote address and ports parameters with "don't care" values. If
macro-flow based remote policy is used, the host MUST specify the
remote address, but MAY or MAY NOT specify the remote port(s). If
micro-flow based remote policy is used, the host MUST specify the
remote address and ports parameter.
Once a LISTEN_REQUEST has been granted, the RSIP gateway MUST forward
all packets destined to the address and port in question to the host,
even if the remote host address and port tuple has not been
previously contacted by the host.
LISTEN_REQUEST is not necessary for RSA-IP.
The following message-specific error conditions exist:
- If the host is not registered with the gateway, the gateway
MUST respond with an ERROR_RESPONSE containing the
REGISTER_FIRST error.
- If the message contains an incorrect client ID, the gateway
MUST respond with an ERROR_RESPONSE containing the
BAD_CLIENT_ID error.
- If the local address parameter is a don't care value and the
RSIP gateway cannot allocate ANY addresses, the RSIP gateway
MUST respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNAVAILABLE error.
Borella, et al. Experimental [Page 33]
RFC 3103 RSIP Protocol Specification October 2001
- If the local address parameter is not a don't care value there
are five possible error conditions:
o If the RSIP gateway cannot allocate ANY addresses, it MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNAVAILABLE error.
o If the RSIP gateway cannot allocate the requested address
because it is in use, the RSIP gateway MUST respond with an
ERROR_RESPONSE containing the LOCAL_ADDR_INUSE error.
o If the RSIP gateway cannot allocate the requested address
because it is not allowed by policy, the RSIP gateway MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDR_UNALLOWED error.
o If the RSIP gateway cannot allocate the requested address /
port tuple because it is in use, the RSIP gateway MUST
respond with an ERROR_RESPONSE containing the
LOCAL_ADDRPORT_INUSE error.
o If the RSIP gateway cannot allocate the requested address /
port tuple because it is not allowed by policy, the RSIP
gateway MUST respond with an ERROR_RESPONSE containing the
LOCAL_ADDRPORT_UNALLOWED error.
- If macro-flow based remote policy is used and the requested
remote address is not allowed by the RSIP gateway's policy, the
RSIP gateway MUST respond with an ERROR_RESPONSE containing the
REMOTE_ADDR_UNALLOWED error.
- If micro-flow based remote policy is used and the requested
remote address / port pair is not allowed by the RSIP gateway's
policy, the RSIP gateway MUST respond with an ERROR_RESPONSE
containing the REMOTE_ADDRPORT_UNALLOWED error.
- If an unsupported or unallowed tunnel type is specified, the
RSIP gateway MUST respond with an ERROR_RESPONSE containing the
BAD_TUNNEL_TYPE error.
- If the host has not specified local or remote address or port
information in enough detail, the RSIP gateway MUST respond
with an ERROR_RESPONSE containing the FLOW_POLICY_VIOLATION
error.
Borella, et al. Experimental [Page 34]
RFC 3103 RSIP Protocol Specification October 2001
A LISTEN_RESPONSE message is used by an RSIP gateway to respond to a
LISTEN_REQUEST message from an RSIP host. The RSIP gateway MUST
issue a bind ID, and specify the address and port which have been
granted to the host. The gateway must also specify a tunnel type and
lease time.
If no remote flow policy is being used, the gateway MUST fill both
the remote address and ports parameters with "don't care" values. If
macro-flow based remote policy is used, the gateway MUST specify the
remote address, but MAY or MAY NOT specify the remote port(s). If
micro-flow based remote policy is used, the gateway MUST specify the
remote address and ports parameter.
If no remote flow policy is being used, the gateway MUST fill both
the remote address and ports parameters with "don't care" values. If
macro-flow based remote policy is used, the gateway MUST specify the
remote address, but MAY or MAY NOT specify the remote port(s). If
micro-flow based remote policy is used, the gateway MUST specify the
remote address and ports parameter.
The address of a tunnel endpoint that is not the RSIP gateway MAY be
specified. If this parameter is not specified, the RSIP gateway MUST
be assumed to be the tunnel endpoint.
Borella, et al. Experimental [Page 35]
RFC 3103 RSIP Protocol Specification October 2001
Message counters are conceptually similar to sequence numbers. They
are necessary to facilitate reliability when UDP is the transport
protocol. Each UDP message is marked with a message counter. When
such a message is transmitted, the message is stored in a "last
message" buffer. For RSIP hosts, a timer is set to expire at the
appropriate timeout value.
General rules:
- When an RSIP host transmits a message with a message counter
value of n, the RSIP gateway's response will contain a message
counter value of n.
- An RSIP host will not increment its message counter value to
n+1 until it receives a message from the RSIP gateway with a
message counter value of n.
- An RSIP gateway begins all sessions with a message counter
value of 1.
- If the message counter value reaches the maximum possible 32-
bit value, it will wrap around to 1, not 0.
- If a message with a message counter value of n is transmitted
by an RSIP host, but a timer expires before a response to that
message is received, the copy of the message (from the "last
message" buffer) is retransmitted.
- When an RSIP gateway receives a duplicate copy of a message
with a message counter value of n, it transmits the contents of
its "last message" buffer.
- When the RSIP gateway transmits an asynchronous RSIP message
(an RSIP message for which there was no request by the RSIP
host), a message counter value of 0 MUST be used. Note that
only three RSIP messages can be transmitted asynchronously:
ERROR_RESPONSE, DE-REGISTER_RESPONSE, and FREE_RESPONSE. These
messages may also be transmitted in response to an RSIP host
request, so their message counter values MAY be non-zero.
- If a message counter is not present in a message from an RSIP
host, but is required, the RSIP gateway MUST respond with an
ERROR_RESPONSE containing the MESSAGE_COUNTER_REQUIRED error.
Borella, et al. Experimental [Page 36]
RFC 3103 RSIP Protocol Specification October 2001
When either the RSIP host or gateway suffers from an unrecoverable
failure, such as a crash, all RSIP-related state will be lost. In
this section, we describe the sequence of events that will occur in
both host and gateway failures, and how the host and gateway re-
synchronize.
After a host failure, the host will reboot and be unaware of any RSIP
state held on its behalf at the gateway.
If the host does not immediately attempt to re-establish a session,
it may receive RSIP packets on the RSIP client application port that
it was using before it rebooted. If an RSIP client application is
not active on this port, these packets will be responded to with ICMP
port unreachable messages. If TCP is the transport protocol, it is
likely that the connection will be terminated with a TCP RST. If an
RSIP client is active on this port, it will not recognize the session
that these packets belong to, and it SHOULD silently ignore them.
The RSIP host may also receive packets from a remote host with which
it was communicating before it rebooted. These packets will be
destined to the RSIP tunnel interface, which should not exist. Thus
they SHOULD be silently discarded by the RSIP host's stack, or the
RSIP host will transmit appropriate ICMP messages to the tunnel
endpoint (e.g., the RSIP gateway). The behavior of the system with
respect to sessions that were active before the reboot should be
similar to that of a publically addressable non-RSIP host that
reboots.
Upon rebooting, an RSIP host may attempt to establish a new RSIP
session with the RSIP gateway. Upon receiving the REGISTER_REQUEST
message, the RSIP gateway will be able to determine that, as far as
it is concerned, the RSIP host is already registered. Thus, it will
transmit an ERROR_RESPONSE with the ALREADY_REGISTERED message. Upon
receipt of this message, the RSIP host will know the client ID of its
old registration, and SHOULD immediately transmit a DE-
REGISTER_REQUEST using this client ID. After this is accomplished,
the states of the RSIP host and gateway have been synchronized, and a
new RSIP session may be established.
If the RSIP host does not de-register itself from the RSIP gateway,
it will eventually receive a DE-REGISTER_RESPONSE from the gateway,
when the gateway times out the host's session. Since the DE-
REGISTER_RESPONSE will refer to a client ID that has no meaning to
Borella, et al. Experimental [Page 37]
RFC 3103 RSIP Protocol Specification October 2001
the host, the host SHOULD silently ignore such a message. At this
point, the states of the RSIP host and gateway have been
synchronized, and a new RSIP session may be established.
After a gateway failure, the gateway will reboot and be unaware of
any RSIP state held by an RSIP host.
Since the gateway will not attempt to contact any of its RSIP hosts,
a problem will first be detected when either an RSIP host sends an
RSIP message to the gateway, an RSIP host sends tunneled data to the
gateway, or data from a remote host intended for an RSIP host
arrives.
In the first case, the RSIP gateway SHOULD immediately response to
all messages (except for a REGISTER_REQUEST) with an ERROR_RESPONSE
with a REGISTER_FIRST error. Upon receipt of such a message, an RSIP
host MUST interpret the message as an indication of a loss of
synchronization between itself and the RSIP gateway. The RSIP host
SHOULD immediately transmit a DE-REGISTRATION_REQUEST with its old
client ID (which will generate another error, but this error SHOULD
be ignored by the host). At this point, the states of the RSIP host
and gateway have been synchronized, and a new RSIP session may be
established.
In the second case, all data that an RSIP host sends to the tunneled
interface of an RSIP server will either (1) be discarded silently,
(2) responded to with an ICMP Destination Unreachable message, such
as "Communication Administratively Prohibited", or (3) blindly routed
to the intended destination. In all of the above cases, the RSIP
gateway will not have an explicit method to notify the RSIP host of
the problem. To prevent a long term communications outage, small
lease times of several minutes can be set by the RSIP gateway.
In the third case, the RSIP gateway SHOULD discard all incoming
packets and/or respond with ICMP Port Unreachable messages.
There is a significant amount of RSIP gateway policy that may be
implemented, but is beyond the scope of this document. We expect
that most of this policy will be site-specific or implementation-
specific and therefore do not make any recommendations. Examples of
general gateway policy include:
- How ports are allocated to RSIP hosts.
- Preferred length of lease times.
Borella, et al. Experimental [Page 38]
RFC 3103 RSIP Protocol Specification October 2001
- How flow policy is applied to which hosts.
- How an RSIP gateway with multiple public IP addresses that may
be leased by RSIP clients determines how to partition
and/or lease these addresses.
Once an RSIP host and gateway have established a relationship and the
host is assigned resources to use, error may occur due to the host's
misuse of the resources or its attempting to use unassigned
resources. The following error behavior is defined:
- If a host attempts to use a local address which it has not been
allocated, the RSIP gateway MUST drop the associated packet(s)
and send the host an ERROR_RESPONSE containing the
LOCAL_ADDR_UNALLOWED error.
- If a host attempts to use a local address / port tuple which it
has not been allocated, the RSIP gateway MUST drop the
associated packet(s) and send the host an ERROR_RESPONSE
containing the LOCAL_ADDRPORT_UNALLOWED error.
- If a host attempts to contact a remote address which has not
been properly specified or otherwise approved (e.g., via an
ASSIGN_RESPONSE_RSAP-IP and macro or micro based remote flow
policy), the RSIP gateway MUST drop the associated packet(s)
and send the host an ERROR_RESPONSE containing the
REMOTE_ADDR_UNALLOWED error.
- If a host attempts to contact a remote address / port tuple
which has not been properly specified or otherwise approved
(e.g., via an ASSIGN_RESPONSE_RSAP-IP and micro based remote
flow policy), the RSIP gateway MUST drop the associated
packet(s) and send the host an ERROR_RESPONSE containing the
REMOTE_ADDRPORT_UNALLOWED error.
- If a host attempts to establish or use an improper tunnel type,
the RSIP gateway MUST respond with an ERROR_RESPONSE containing
the BAD_TUNNEL_TYPE error.
- If the RSIP gateway's detects a local fault which prevents its
RSIP server module from continuing operation, the RSIP gateway
MUST respond with an ERROR_RESPONSE containing the
INTERNAL_SERVER_ERROR error.
Borella, et al. Experimental [Page 39]
RFC 3103 RSIP Protocol Specification October 2001
Regardless of local flow policy, an RSIP host may "suggest" that it
would like to use a particular local address and/or port number in a
particular binding. An RSIP gateway that cannot grant such a
request, because the specified resources are already in use, MUST
respond with an ERROR_RESPONSE containing the LOCAL_ADDR_INUSE or
LOCAL_ADDRPORT_INUSE values.
An RSIP host may initialize a publically accessible gateway (such as
an FTP or HTTP gateway) by transmitting a LISTEN_REQUEST message to
an RSIP gateway and receiving a LISTEN_RESPONSE. However, unless no
remote flow policy is used, the gateway will have to specify the
address or address and port of a single remote host that will be
allowed to contact it. Obviously, such as restriction is not very
useful for hosts that require their gateways to be accessible by any
remote host.
This indicates that there is a conflict between flow-based policy and
support for gateways. The main purpose of enforcing flow-based
policy for LISTEN_REQUESTs is that it allows an RSIP gateway tight
control over how an RSIP host uses ports and the associated
accounting. For example, an RSIP host, operating under remote
micro-flow based policy and using a protocol such as FTP, will have
to specify the address and port that it will receive FTP data on, as
well as the address and port that the gateway will transmit data
from, in a LISTEN_REQUEST.
In general, an RSIP gateway may not allow arbitrary hosts to start
public gateways because of the traffic and security concerns. Thus,
we recommend that if remote micro-flow based policy is used, that an
RSIP gateway only allow public gateways on RSIP hosts via
administrative override.
Currently, RSIP hosts can only be identified by their local IP
address or MAC address.
RSIP, in and of itself, does not provide security. It may provide
the illusion of security or privacy by hiding a private address
space, but security can only be ensured by the proper use of security
protocols and cryptographic techniques.
Borella, et al. Experimental [Page 40]
RFC 3103 RSIP Protocol Specification October 2001
An RSIP gateway should take all measures deemed necessary to prevent
its hosts from performing intentional or unintentional denial-of-
service attacks by request large sets of resources.
Currently, RSIP hosts can only be identified by their local IP
address or, in some cases, MAC address. It is desirable to allow
RSIP messages sent between a host and gateway to be authenticated.
Further discussion of such authentication can be found in [RSIP-
FRAME].
Discussion of RSIP support for end-to-end IPsec can be found in
[RSIP-IPSEC].
All of the designations below have been registered by the IANA.
- RSIP port number: 4555
- RSIP error codes (see Appendix A).
- RSIP message type codes (see Appendix B).
- RSIP tunnel types, methods, and flow policies.
RSIP parameter values are designated as follows:
- 0 Reserved
- 1-240 Assigned by IANA
- 241-255 Reserved for private use
New registrations for the above namespaces are recommended to be
allocated via the Specification Required method documented in
[RFC2434].
The authors would like to specifically thank Gabriel Montenegro, Pyda
Srisuresh, Brian Carpenter, Eliot Lear, Dan Nessett, Gary Jaszewski,
Naveen Rajanikantha, Sudhakar Ramakrishna, Jim March, and Rick Cobb
for their input. The IETF NAT working group as a whole has been
extremely helpful in the ongoing development of RSIP.
Borella, et al. Experimental [Page 41]
RFC 3103 RSIP Protocol Specification October 2001
This section provides descriptions for the error values in the RSIP
error parameter.
All errors are grouped into the following categories:
100's: General errors.
101: UNKNOWN_ERROR. An error that cannot be identified has
occurred. This error should be used when all other error
messages are inappropriate.
102: USE_TCP. A host has attempted to use UDP on a server that
only supports TCP.
103: FLOW_POLICY_VIOLATION: A host has not specified address or
port information in enough detail for its assigned flow policy.
104: INTERNAL_SERVER_ERROR: An RSIP server application has
detected an unrecoverable error within itself or the RSIP
gateway.
105: MESSAGE_COUNTER_REQUIRED: An RSIP host did not use a message
counter parameter in a situation in which it should have.
106: UNSUPPORTED_RSIP_VERSION: An RSIP host sent a message with a
version number that is not supported by the RSIP gateway.
200's: Parameter and message errors. The gateway uses these errors
when it detects that a parameter or message is malformed, as well
as when it does not understand a parameter or message.
201: MISSING_PARAM. The request does not contain a required
parameter.
202: DUPLICATE_PARAM. The request contains an illegal duplicate
parameter.
203: EXTRA_PARAM. The request contains a parameter that it should
not.
204: ILLEGAL_PARAM. The gateway does not understand a parameter
type.
205: BAD_PARAM. A parameter is malformed.
Borella, et al. Experimental [Page 42]
RFC 3103 RSIP Protocol Specification October 2001
206: ILLEGAL_MESSAGE. The gateway does not understand the message
type. The message type is neither mandatory nor optional.
207: BAD_MESSAGE. A message is malformed and gateway parsing
failed.
208: UNSUPPORTED_MESSAGE: The host has transmitted an optional
message that the gateway does not support.
300's: Permission, resource, and policy errors. The gateway uses
these errors when a host has attempted to do something that it is
not permitted to do, or something that violated gateway policy.
301: REGISTER_FIRST. The RSIP host has attempted to request or
use resources without registering.
302: ALREADY_REGISTERED. The host has attempted to register again
without first de-registering.
303: ALREADY_UNREGISTERED. The host has attempted to de-register
but it is already in the unregistered state.
304: REGISTRATION_DENIED. The gateway will not allow the host to
register.
305: BAD_CLIENT_ID. The host has referred to itself with the
wrong client ID.
306: BAD_BIND_ID. The request refers to a bind ID that is not
valid for the host.
307: BAD_TUNNEL_TYPE. The request refers to a tunnel type that is
not valid for the host.
308: LOCAL_ADDR_UNAVAILABLE. The gateway is currently not able to
allocate ANY local address, but the host may try again later.
309: LOCAL_ADDRPORT_UNAVAILABLE. The gateway is currently not
able to allocate ANY local IP address / port tuple of the
requested magnitude (i.e., number of ports), but the host may
try again later.
310: LOCAL_ADDR_INUSE. The gateway was not able to allocate the
requested local address because it is currently used by another
entity.
Borella, et al. Experimental [Page 43]
RFC 3103 RSIP Protocol Specification October 2001
311: LOCAL_ADDRPORT_INUSE. The gateway was not able to allocate
the requested local address / port tuple because it is
currently used by another entity.
312: LOCAL_ADDR_UNALLOWED. The gateway will not let the host use
the specified local IP address due to policy.
313: LOCAL_ADDRPORT_UNALLOWED. The gateway will not let the host
use the specified local address / port pair due to policy.
314: REMOTE_ADDR_UNALLOWED. The gateway will not allow the host
to establish a session to the specified remote address.
315: REMOTE_ADDRPORT_UNALLOWED. The gateway will not allow the
host to establish a session to the specified remote address /
port tuple.
400's: IPsec errors. All errors specific to RSIP / IPsec operation.
See [RSIP-IPSEC].
This section defines the values assigned to RSIP message types. We
also indicate which RSIP entity, host or gateway, produces each
messages, and whether it is mandatory or optional. All *_REQUEST
messages are only to be implemented on hosts, while all *_RESPONSE
messages are only to be implemented on gateways. RSIP
implementations (both host and gateway) MUST support all mandatory
messages in order to be considered "RSIP compliant".
Borella, et al. Experimental [Page 44]
RFC 3103 RSIP Protocol Specification October 2001
Value Message Implementation Status
------------------------------------------------------------
1 ERROR_RESPONSE gateway mandatory
2 REGISTER_REQUEST host mandatory
3 REGISTER_RESPONSE gateway mandatory
4 DE-REGISTER_REQUEST host mandatory
5 DE-REGISTER_RESPONSE gateway mandatory
6 ASSIGN_REQUEST_RSA-IP host optional
7 ASSIGN_RESPONSE_RSA-IP gateway optional
8 ASSIGN_REQUEST_RSAP-IP host mandatory
9 ASSIGN_RESPONSE_RSAP-IP gateway mandatory
10 EXTEND_REQUEST host mandatory
11 EXTEND_RESPONSE gateway mandatory
12 FREE_REQUEST host mandatory
13 FREE_RESPONSE gateway mandatory
14 QUERY_REQUEST host optional
15 QUERY_RESPONSE gateway mandatory
16 LISTEN_REQUEST host optional
17 LISTEN_RESPONSE gateway optional
In this appendix, we present an exemplary series of annotated
transactions between an RSIP host and an RSIP gateway. All host to
gateway traffic is denote by `C --> S' and all gateway to host
traffic is denoted by `S --> C'. Parameter values are denoted inside
of parentheses. Versions, message types, and overall lengths are not
included in order to save space. "Don't care" values are indicated
by 0's.
A ports parameter is represented by the number of ports followed by
the port numbers, separated by dashes. For example, 2-1012-1013
indicates two ports, namely 1012 and 1013, while 16-10000 indicates
16 ports, namely 10000-10015, and 4-0 indicates four ports, but the
sender doesn't care where they are.
IPv4 addresses are assumed.
Policy
This example exhibits the loosest policy framework for RSAP-IP.
C --> S: REGISTER_REQUEST ()
The host attempts to register with the gateway.
Borella, et al. Experimental [Page 45]
RFC 3103 RSIP Protocol Specification October 2001
S --> C: REGISTER_RESPONSE (Client ID = 1, Local Flow Policy =
Macro, Remote Flow policy = None, Lease Time = 600)
The gateway responds, assigning a Client ID of 1, local macro-
flow based policy and no remote flow policy. No RSIP method is
indicated, so RSAP-IP is assumed. No tunnel type is indicated,
so IP-IP is assumed. A lease time of 600 seconds is assigned.
C --> S: ASSIGN_REQUEST_RSAP-IP: (Client ID = 1, Address (local) =
0, Ports (local) = 4-0, Address (remote) = 0, Ports (remote) =
0, Lease Time = 3600)
The host requests an address and four ports to use with it, but
doesn't care which address or ports are assigned. The host
does not specify the remote address or ports either. The host
suggests a lease time of 3600 seconds.
S --> C: ASSIGN_RESPONSE_RSAP-IP: (Client ID = 1, Bind ID = 1,
Address (local) = 149.112.240.156, Ports (local) = 4-1234,
Address (remote) = 0, Ports (remote) = 0, Lease Time = 1800,
Tunnel Type = IP-IP)
The gateway responds by indicating that a bind ID of 1 has been
assigned to IP address 149.112.240.156 with ports 1234-1237.
Any remote host may be communicated with, using any remote port
number. The lease time has been assigned to be 1800 seconds,
and the tunnel type is confirmed to be IP-IP.
The host is now able to communicate with any host on the public
network using these resources.
C --> S: QUERY_REQUEST: (Client ID = 1, Indicator = network,
Address (network) = 10.20.60.0, Address (netmask)
255.255.255.0)
The host asks the gateway if the network 10.20.60.0/24 is
local.
S --> C: QUERY_RESPONSE: (Client ID = 1, Indicator = network,
Address (network) = 10.20.60.0, Address (netmask) =
255.255.255.0)
The gateway responds indicating that the network in question is
local.
C --> S: ASSIGN_REQUEST_RSAP-IP: (Client ID = 1, Address (local) =
149.112.240.156, Ports (local) = 8-1238, Address (remote) = 0,
Ports (remote) = 0, Lease Time = 1800)
Borella, et al. Experimental [Page 46]
RFC 3103 RSIP Protocol Specification October 2001
The host requests eight more particular ports for use with
RSAP-IP with the same address. A lease of 1800 seconds is
requested. IP-IP tunneling is implied by default.
S --> C: ASSIGN_RESPONSE_RSAP-IP: (Client ID = 1, Bind ID = 2,
Address (local) = 149.112.240.156, Ports (local) = 8-1305,
Address (remote) = 0, Ports (remote) = 0, Lease Time = 1800)
The gateway grants the request with the same address, but with
a different set of ports. IP-IP tunneling is implied by
default.
C --> S: FREE_REQUEST (Client ID = 1, Bind ID = 1)
The host frees bind ID 1; i.e., ports 1234-1237 from IP address
149.112.240.156. Note that the address itself is still
assigned to the host because the host is still assigned ports
1305-1314.
S --> C: FREE_RESPONSE (Client ID = 1, Bind ID = 1)
The gateway acknowledges that Bind ID 1 has been freed.
C --> S: EXTEND_REQUEST (Client ID = 1, Bind ID = 2, Lease Time =
1800)
The host request that the lease on bind ID 1 be extended for
1800 seconds.
S --> C: EXTEND_RESPONSE (Client ID = 1, Bind ID = 2, Lease Time =
1800)
The gateway confirms the request.
S --> C: FREE_RESPONSE (Client ID = 1, Bind ID = 2)
The gateway forces the host to free the resources of bind ID 2.
C --> S: DE-REGISTER_REQUEST (Client ID = 1)
The host de-registers with the sever.
S --> C: DE-REGISTER_RESPONSE (Client ID = 1)
The gateway acknowledges that the host has de-registered.
Borella, et al. Experimental [Page 47]
RFC 3103 RSIP Protocol Specification October 2001
flow Based Policy
This example exhibits the strictest policy framework for RSAP-IP.
C --> S: REGISTER_REQUEST ()
The host attempts to register with the gateway.
S --> C: REGISTER_RESPONSE (Client ID = 5, Local Flow Policy =
Micro, Remote Flow policy = Micro, RSIP Method = RSAP-IP, RSIP
Method = RSA-IP, Tunnel Type = IP-IP, Tunnel Type = GRE, Lease
Time = 600)
The gateway responds, assigning a Client ID of 5, local micro-
flow based policy and remote micro-flow based policy. Both
RSAP-IP and RSA-IP are supported. Both IP-IP and GRE tunnel
types are supported. A lease time of 600 seconds is assigned.
C --> S: ASSIGN_REQUEST_RSAP-IP: (Client ID = 5, Address (local) =
0, Ports (local) = 0, Address (remote) = 38.196.73.6, Ports
(remote) = 21, Lease Time = 600, Tunnel Type = IP-IP)
The host requests a local address and a port assignment to use
with it. The host indicates that it wants to contact host
38.196.73.6 at port 21 (FTP control). The host requests a
lease time of 600 seconds and a tunnel type of IP-IP.
S --> C: ASSIGN_RESPONSE_RSAP-IP: (Client ID = 5, Bind ID = 1,
Address (local) = 149.112.240.156, Ports (local) = 2049,
Address (remote) = 38.196.73.6, Ports (remote) = 21, Lease Time
= 600, Tunnel Type = IP-IP)
The gateway responds by indicating that a bind ID of 1 has been
assigned to IP address 149.112.240.156 with port 2049. Only
host 38.196.73.6 at port 21 may be contacted. The lease time
has been assigned to be 600 seconds, and the tunnel type is
confirmed to be IP-IP.
C --> S: LISTEN_REQUEST: (Client ID = 5, Address (local) =
149.112.240.156, Ports (local) = 2050, Address (remote) =
38.196.73.6, Ports (remote) = 20)
The host requests a listen port 2050 at the same address that
it has been assigned. Only host 38.196.73.6 from ports 20 (FTP
data) will be able to contact it.
Borella, et al. Experimental [Page 48]
RFC 3103 RSIP Protocol Specification October 2001
S --> C: LISTEN_RESPONSE: (Client ID = 5, Address (local) =
149.112.240.156, Ports (local) = 2050, Address (remote) =
38.196.73.6, Ports (remote) = 20, Lease Time = 600, Tunnel Type
= IP-IP)
The gateway confirms the request and assigns a lease time of
600 seconds and a tunnel type of IP-IP.
C --> S: DE-REGISTER_REQUEST (Client ID = 5)
The host de-registers with the sever.
S --> C: DE-REGISTER_RESPONSE (Client ID = 5)
The gateway acknowledges that the host has de-registered. All
of the host's bindings have been implicitly revoked.
flow based Policy
This example exhibits a medium level of control for RSA-IP.
C --> S: REGISTER_REQUEST ()
The host attempts to register with the gateway.
S --> C: REGISTER_RESPONSE (Client ID = 3, Local Flow Policy =
Macro, Remote Flow policy = Macro, RSIP Method = RSAP-IP, RSIP
Method = RSA-IP, Tunnel Type = IP-IP, Tunnel Type = L2TP, Lease
Time = 600)
The gateway responds, assigning a Client ID of 3, local macro-
flow based policy and remote macro-flow based policy. Both
RSAP-IP and RSA-IP are supported. Both IP-IP and L2TP tunnel
types are supported. A lease time of 600 seconds is assigned.
C --> S: ASSIGN_REQUEST_RSA-IP: (Client ID = 3, Address (local) =
0, Address (remote) = www.foo.com, Ports (remote) = 0, Lease
Time = 3600, Tunnel Type = IP-IP)
The host requests a local address and indicates that it wants
to contact host www.foo.com.
S --> C: ERROR_RESPONSE: (Error = REMOTE_ADDR_UNALLOWED, Client ID
= 3)
The gateway indicates that the host is not permitted to
establish communication with www.foo.com.
Borella, et al. Experimental [Page 49]
RFC 3103 RSIP Protocol Specification October 2001
C --> S: ASSIGN_REQUEST_RSA-IP: (Client ID = 3, Address (local) =
0, Address (remote) = www.bar.com, Ports (remote) = 0, Lease
Time = 3600, Tunnel Type = IP-IP)
The host requests a local address and indicates that it wants
to contact host www.bar.com.
S --> C: ASSIGN_RESPONSE_RSA-IP: (Client ID = 3, Bind ID = 1,
Address (local) = 149.112.240.17, Address (remote) =
www.bar.com, Ports (remote) = 0, Lease Time = 3600, Tunnel Type
= IP-IP)
The gateway responds by granting local IP address
149.112.240.17 to the host, and permitting it to communicate
with www.bar.com, at any port. Requested lease time and tunnel
type are also granted.
C --> S: DE-REGISTER_REQUEST (Client ID = 3)
The host de-registers with the sever.
S --> C: DE-REGISTER_RESPONSE (Client ID = 3)
The gateway acknowledges that the host has de-registered. All
of the host's bindings have been implicitly revoked.
[RFC1918] Rekhter, Y., Moskowitz, B., Karrenberg, D., de Groot,
G.J. and E. Lear, "Address Allocation for Private
Internets", BCP 5, RFC 1918, February 1996.
[RFC2119] Bradner, S., "Key words for use in RFCs to indicate
requirement levels", BCP 14, RFC 2119, March 1997.
[RFC2434] Narten, T. and H. Alvestrand, "Guidelines for Writing an
IANA Considerations Section in RFCs", BCP 26, RFC 2434,
October 1998.
[RFC2663] Srisuresh, P. and M. Holdrege, "IP Network Address
Translator (NAT) Terminology and Considerations", RFC
2663, August 1999.
[RSIP-FRAME] Borella, M. Lo, J., Grabelsky, D. and G. Montenegro,
"Realm Specific IP: Framework", RFC 3102, October 2001.
[RSIP-IPSEC] Montenegro, G. and M. Borella, "RSIP Support for End-
to-end IPSEC", RFC 3104, October 2001.
Borella, et al. Experimental [Page 52]
RFC 3103 RSIP Protocol Specification October 2001
Michael Borella
CommWorks
3800 Golf Rd.
Rolling Meadows IL 60008
Phone: (847) 262-3083
EMail: mike_borella@commworks.com
David Grabelsky
CommWorks
3800 Golf Rd.
Rolling Meadows IL 60008
Phone: (847) 222-2483
EMail: david_grabelsky@commworks.com
Jeffrey Lo
Candlestick Networks, Inc
70 Las Colinas Lane,
San Jose, CA 95119
Phone: (408) 284 4132
EMail: yidarlo@yahoo.com
Kunihiro Taniguchi
NEC USA
C&C Research Labs.
110 Rio Robles
San Jose, CA 95134
Phone: (408) 943-3031
EMail: taniguti@ccrl.sj.nec.com
Borella, et al. Experimental [Page 53]
RFC 3103 RSIP Protocol Specification October 2001
Copyright (C) The Internet Society (2001). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
Acknowledgement
Funding for the RFC Editor function is currently provided by the
Internet Society.
Borella, et al. Experimental [Page 54]