|
Network Working Group Request for Comments: 3340 Category: Standards Track |
M. Rose Dover Beach Consulting, Inc. G. Klyne Clearswift Corporation D. Crocker Brandenburg InternetWorking July 2002 |
This document specifies an Internet standards track protocol for the Internet community, and requests discussion and suggestions for improvements. Please refer to the current edition of the "Internet Official Protocol Standards" (STD 1) for the standardization state and status of this protocol. Distribution of this memo is unlimited.
Copyright © The Internet Society (2002). All Rights Reserved.
This memo describes Application Exchange (APEX) Core, an extensible, asynchronous message relaying service for application layer programs.
1. Introduction
1.1 Overview
1.2 Architecture at a Glance
2. Service Principles
2.1 Modes of Operation
2.2 Naming of Entities
2.2.1 Comparing Endpoints
3. Service Provisioning
3.1 Connection Establishment
3.2 Authentication
3.3 Authorization
3.4 Confidentiality
3.5 Relaying Integrity
3.6 Traffic Analysis
4. The APEX
4.1 Use of XML and MIME
4.2 Profile Identification and Initialization
4.3 Message Syntax
4.4 Message Semantics
4.4.1 The Attach Operation
4.4.2 The Bind Operation
4.4.3 The Terminate Operation
4.4.4 The Data Operation
4.4.4.1 Relay Processing of Data
4.4.4.2 Application Processing of Data
4.5 APEX Access Policies
4.5.1 Access Policies in the Endpoint-Relay Mode
4.5.2 Access Policies in the Relay-Relay Mode
5. APEX Options
5.1 The statusRequest Option
6. APEX Services
6.1 Use of the APEX Core DTD
6.1.1 Transaction-Identifiers
6.1.2 The Reply Element
6.2 The Report Service
7. Registration Templates
7.1 APEX Option Registration Template
7.2 APEX Service Registration Template
7.3 APEX Endpoint Application Registration Template
8. Initial Registrations
8.1 Registration: The APEX Profile
8.2 Registration: The System (Well-Known) TCP port number forapex-mesh
8.3 Registration: The System (Well-Known) TCP port number forapex-edge
Network applications can be broadly distinguished by five operational characteristics:
For example:
Messaging applications vary considerably in their operational requirements. For example, some messaging applications require assurance of timeliness and reliability, whilst others do not.
These features come at a cost, in terms of both infrastructural and configuration complexity. Accordingly, the underlying service must be extensible to support different requirements in a consistent manner.
This memo defines a core messaging service that supports a range of operational characteristics. The core service supports a variety of tailored services for both user-based and programmatic exchanges.
APEX provides an extensible, asynchronous message relaying service for application layer programs.
APEX, at its core, provides a best-effort datagram service. Each datagram, simply termed "data", is originated and received by APEX "endpoints" -- applications that dynamically attach to the APEX "relaying mesh".
The data transmitted specifies:
Options are used to alter the semantics of the service, which may occur on a per-recipient or per-data basis, and may be processed by either a single or multiple relays.
Additional APEX services are provided on top of the relaying mesh; e.g., access control and presence information.
APEX is specified, in part, as a BEEP [1] "profile". Accordingly, many aspects of APEX (e.g., authentication) are provided within the BEEP core. Throughout this memo, the terms "peer", "initiator", "listener", "client", and "server" are used in the context of BEEP. In particular, Section 2.1 of the BEEP core memo discusses the roles that a BEEP peer may perform.
When reading this memo, note that the terms "endpoint" and "relay" are specific to APEX, they do not exist in the context of BEEP.
The APEX stack:
+-------------+
| APEX | an APEX process is either:
| process |
+-------------+ - an application attached as an APEX
| | endpoint; or,
| APEX |
| | - an APEX relay
+-------------+
| | APEX services are realized as applications
| BEEP | having a special relationship with the APEX
| | relays in their administrative domain
+-------------+
| TCP |
+-------------+
| ... |
+-------------+
The APEX entities:
administrative domain #1 administrative domain #2
+----------------------------+ +----------------------------+
| +------+ | | +------+ |
| | | | | | | |
| | appl | | | | appl | |
| | | | | | | |
| +......+ +------+ | | +------+ +......+ |
| | | | | | | | | | | |
| |end- | |relay | | | |relay | |end- | |
| | point| | | | | | | | point| |
| +------+ +------+ | | +------+ +------+ |
| | | | | | | | | | | |
| | APEX | | APEX | | | | APEX | | APEX | |
| | | | | | | | | | | |
| +------+ +------+ | | +------+ +------+ |
| || || || | | || || || |
| ============= ================ ============= |
+----------------------------+ +----------------------------+
| <---- APEX relaying mesh ----> |
Note: relaying between administrative domains is configured using SRV RRs. Accordingly, the actual number of relays between two endpoints is not fixed.
APEX is used in two modes:
endpoint-relay: in which the endpoint is always the BEEP initiator of the service, whilst relays are always the BEEP listeners. In this context, applications attach as endpoints, and then the transmission of data occurs.
relay-relay: in which relays typically, though not necessarily, reside in different administrative domains. In this context, applications bind as relays, and then the transmission of data occurs.
In the endpoint-relay mode, an endpoint (BEEP initiator) may:
A relay (BEEP listener), in addition to servicing requests from a BEEP initiator, may:
In the relay-relay mode, a relay (BEEP listener or initiator) may:
Endpoints are named using the following ABNF [2] syntax:
;; Domain is defined in [3], either a FQDN or a literal
entity = local "@" Domain
local = address [ "/" subaddress ]
address = token
subaddress = token
;; all non-control characters, excluding "/" and "@" delimiters
token = 1*(%x20-2E / %x30-3F / %x41-7E / UTF-8) ;; [4]
Two further conventions are applied when using this syntax:
the "apex=" convention: All endpoint identities having a local-part starting with "apex=" are reserved for use by APEX services registered with the IANA; and,
the "subaddress" convention: If the solidus character ("/", decimal code 47) occurs in the local-part, this identifies a subaddress of an endpoint identity (e.g., "fred/appl=wb@example.com" is a subaddress of the APEX endpoint "fred@example.com").
All subaddresses starting with "appl=" are reserved for use by APEX endpoint applications registered with the IANA.
Relays, although not named, serve of behalf of administrative domains, as identified by a FQDN or a domain-literal, e.g., "example.com" or "[10.0.0.1]".
In APEX, "endpoints" and "relays" are the fundamental entities. APEX is carried over BEEP, which has the "peer" as its fundamental entity. The relationship between BEEP peer entities and APEX endpoint and relay entities are defined by APEX's Access Policies (Section 4.5).
Note that since the "local" part of an entity is a string of UTF-8 [4] octets, comparison operations on the "local" part use exact matching (i.e., are case-sensitive).
Accordingly, "fred@example.com" and "Fred@example.com" refer to different endpoints. Of course, relays serving the "example.com" administrative domain may choose to treat the two endpoints identically for the purposes of routing and delivery.
Finally, note that if an APEX endpoint is represented using a transmission encoding, then, prior to comparison, the encoding is reversed. For example, if the URL encoding is used, then "apex:fred@example.com" is identical to "apex:f%72ed@example.com".
The SRV algorithm [5] is used to determine the IP/TCP addressing information assigned to the relays for an administrative domain identified by a FQDN:
service: "apex-edge" (for the endpoint-relay mode), or "apex-mesh" (for the relay-relay mode);
protocol: "tcp"; and,
domain: the administrative domain.
If the administrative domain is identified by a domain-literal, then the IP address information is taken directly from the literal and the TCP port number used is assigned by the IANA for the registration in Section 8.2.
Authentication is a matter of provisioning for each BEEP peer (c.f., Section 4.5).
An APEX relay might be provisioned to allow a BEEP peer identity to coincide with a given endpoint identity. For example, a relay in the "example.com" administrative domain may be configured to allow a BEEP peer identified as "fred@example.com" to be authorized to attach as the APEX endpoint "fred@example.com".
Authorization is a matter of provisioning for each BEEP peer (c.f., Section 4.5).
Typically, a relay requires that its BEEP peer authenticate as a prelude to authorization, but an endpoint usually does not require the same of its BEEP peer.
Confidentiality is a matter of provisioning for each BEEP peer.
Typically, any data considered sensitive by an originating endpoint will have its content encrypted for the intended recipient endpoint(s), rather than relying on hop-by-hop encryption. Similarly, an originating endpoint will sign the content if end-to- end authentication is desired.
Data are relayed according to SRV entries in the DNS. Accordingly, relaying integrity is a function of the DNS and the applications making use of the DNS. Additional assurance is provided if the BEEP initiator requires that the BEEP listener authenticate itself.
Hop-by-hop protection of data transmitted through the relaying mesh (endpoint identities and content) is afforded at the BEEP level through the use of a transport security profile. Other traffic characteristics, e.g., volume and timing of transmissions, are not protected from third-party analysis.
Section 8.1 contains the BEEP profile registration for APEX.
Each BEEP payload exchanged via APEX consists of an XML document and possibly an arbitrary MIME content.
If only an XML document is sent in the BEEP payload, then the mapping to a BEEP payload is straight-forward, e.g.,
C: MSG 1 2 . 111 39
C: Content-Type: application/beep+xml
C:
C: <terminate transID='1' />
C: END
Otherwise, if an arbitrary MIME content is present, it is indicated by a URI-reference [6] in the XML control document. The URI- reference may contain an absolute-URI (and possibly a fragment- identifier), or it may be a relative-URI consisting only of a fragment-identifier. Arbitrary MIME content is included in the BEEP payload by using a "multipart/related" [7], identified using a "cid" URL [8], and the XML control document occurs as the start of the "multipart/related", e.g.,
C: MSG 1 1 . 42 1234
C: Content-Type: multipart/related; boundary="boundary";
C: start="<1@example.com>";
C: type="application/beep+xml"
C:
C: --boundary
C: Content-Type: application/beep+xml
C: Content-ID: <1@example.com>
C:
C: <data content='cid:2@example.com'>
C: <originator identity='fred@example.com' />
C: <recipient identity='barney@example.com' />
C: </data>
C: --boundary
C: Content-Type: image/gif
C: Content-Transfer-Encoding: binary
C: Content-ID: <2@example.com>
C:
C: ...
C: --boundary--
C: END
Because BEEP provides an 8bit-wide path, a "transformative" Content- Transfer-Encoding (e.g., "base64" or "quoted-printable") should not be used. Further, note that MIME [9] requires that the value of the "Content-ID" header be globally unique.
If the arbitrary MIME content is itself an XML document, it may be contained within the control document directly as a "data-content" element, and identified using a URI-reference consisting of only a fragment-identifier, e.g.,
C: MSG 1 1 . 42 295
C: Content-Type: application/beep+xml
C:
C: <data content='#Content'>
C: <originator identity='fred@example.com' />
C: <recipient identity='barney@example.com' />
C: <data-content Name='Content'>
C: <statusResponse transID='86'>
C: <destination identity='barney@example.com'>
C: <reply code='250' />
C: </destination>
C: </statusResponse>
C: </data-content>
C: </data>
C: END
The APEX is identified as
in the BEEP "profile" element during channel creation.
No elements are required to be exchanged during channel creation; however, in the endpoint-relay mode, the BEEP initiator will typically include an "attach" element during channel creation, e.g.,
<start number='1'>
<profile uri='http://iana.org/beep/APEX'>
<![CDATA[<attach endpoint='fred@example.com'
transID='1' />]]>
</profile>
</start>
Similarly, in the relay-relay mode, the BEEP initiator will typically include an "bind" element during channel creation, e.g.,
<start number='1'>
<profile uri='http://iana.org/beep/APEX'>
<![CDATA[<bind relay='example.com'
transID='1' />]]>
</profile>
</start>
Section 9.1 defines the BEEP payloads that are used in the APEX.
When an application wants to attach to the relaying mesh as a given endpoint, it sends an "attach" element to a relay, e.g.,
+-------+ +-------+
| | -- attach -----> | |
| appl. | | relay |
| | <--------- ok -- | |
+-------+ +-------+
C: <attach endpoint='fred@example.com' transID='1' />
S: <ok />
or
+-------+ +-------+
| | -- attach -----> | |
| | | |
| | <--------- ok -- | |
| appl. | | relay |
| | -- attach -----> | |
| | | |
| | <--------- ok -- | |
+-------+ +-------+
C: <attach endpoint='fred@example.com' transID='1' />
S: <ok />
C: <attach endpoint='wilma@example.com' transID='2' />
S: <ok />
or
+-------+ +-------+
| | -- attach -----> | |
| appl. | | relay |
| | <------ error -- | |
+-------+ +-------+
C: <attach endpoint='fred@example.com' transID='1' />
S: <error code='537'>access denied</error>
The "attach" element has an "endpoint" attribute, a "transID" attribute, and contains zero or more "option" elements:
When a relay receives an "attach" element, it performs these steps:
When an application wants to identify itself as a relay, it sends a "bind" element to another relay, e.g.,
+-------+ +-------+
| | -- bind -------> | |
| relay | | relay |
| #1 | <--------- ok -- | #2 |
+-------+ +-------+
C: <bind relay='example.com' transID='1' />
S: <ok />
or
+-------+ +-------+
| | -- bind -------> | |
| | | |
| | <--------- ok -- | |
| relay | | relay |
| #1 | -- bind -------> | #2 |
| | | |
| | <--------- ok -- | |
+-------+ +-------+
C: <bind relay='example.com' transID='1' />
S: <ok />
C: <bind relay='rubble.com' transID='2' />
S: <ok />
or
+-------+ +-------+
| | -- bind -------> | |
| relay | | relay |
| #1 | <------ error -- | #2 |
+-------+ +-------+
C: <bind relay='example.com' transID='1' />
S: <error code='537'>access denied</error>
The "bind" element has a "relay" attribute, a "transID" attribute, and contains zero or more "option" elements:
When a relay receives an "bind" element, it performs these steps:
When an application or relay wants to release an attachment or binding, it sends a "terminate" element, e.g.,
+-------+ +-------+
| | -- terminate --> | |
| appl. | | relay |
| | <--------- ok -- | |
+-------+ +-------+
C: <terminate transID='1' />
S: <ok />
or
+-------+ +-------+
| | -- terminate --> | |
| appl. | | relay |
| | <------ error -- | |
+-------+ +-------+
C: <terminate transID='13' />
S: <error code='550'>unknown transaction-identifier</error>
or
+-------+ +-------+
| | <-- terminate -- | |
| appl. | | relay |
| | -- ok ---------> | |
+-------+ +-------+
C: <terminate transID='1' />
S: <ok />
The "terminate" element has a "transID" attribute, an optional "code" attribute, an optional "xml:lang" attribute, and may contain arbitrary textual content:
When an application or relay receives a "terminate" element, it performs these steps:
When an application or relay wants to transmit data over the relaying mesh, it sends a "data" element, e.g.,
+-------+ +-------+
| | -- data -------> | |
| appl. | | relay |
| #1 | <--------- ok -- | |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@example.com' />
</data>
S: <ok />
or
+-------+ +-------+
| | -- data -------> | |
| appl. | | relay |
| #1 | <------ error -- | |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@example.com' />
</data>
S: <error code='537'>access denied</error>
or
+-------+ +-------+
| | -- data -------> | |
| relay | | appl. |
| | <--------- ok -- | #2 |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@example.com' />
</data>
S: <ok />
The "data" element has a "content" attribute, and contains an "originator" element, one or more "recipient" elements, zero or more "option" elements, and, optionally, a "data-content" element:
The "originator" element has an "identity" attribute, and contains zero or more option elements:
Each "recipient" element has an "identity" attribute, and contains zero or more option elements:
When a relay receives a "data" element, it performs these steps:
If an APEX session is established, the new "data" is sent, and the recipient's relay returns an "ok" element, then the recipient is considered to be successfully processed.
If so, a new "data" element, containing only that "recipient" element, is sent to the corresponding application. If the recipient's endpoint returns an "ok" element, then the recipient is considered to be successfully processed.
Providing that these semantics are preserved, a relay may choose to optimize its behavior by grouping multiple recipients in a single "data" element that is subsequently transmitted.
Finally, note that a relay receiving a "data" element from an application may be configured to add administrative-specific options.
Regardless, all relays are expressly forbidden from modifying the content of the "data" element at any time.
When an application receives a "data" element, it performs these steps:
Access to APEX is provided by the juxtaposition of:
Each of these activities occurs according to the policies of the relevant administrative domain:
Typically, the identity function is used, e.g., if an application
authenticates itself as the BEEP peer named as "fred@example.com",
it is allowed to attach as the APEX endpoint named as
"fred@example.com".
However, using the "subaddress" convention of Section 2.2, an application authorized to attach as a given APEX endpoint is also authorized to attach as any subaddress of that APEX endpoint, e.g., an application authorized to attach as the APEX endpoint "fred@example.com" is also authorized to attach as the APEX endpoint "fred/appl=wb@example.com".
Typically, the identity function is used, e.g., if an application attaches as the APEX endpoint named as "fred@example.com", it is allowed to send data originating from the same APEX endpoint. However, other policies are permissible, for example, the administrative domain may allow the application attached as the APEX endpoint named as "wilma@example.com" to send data originating as either "wilma@example.com" or "fred@example.com".
If so, then typically the identity function is used. e.g., if an application authenticates itself as the BEEP peer named as "example.com", it is allowed to bind as a relay on behalf of the administrative domain "example.com".
Typically, the identity function is used, e.g., if a relay authenticates itself as being from the same administrative domain as the originator of the data, then the data is accepted.
In addition, some relays may also be configured as "trusted" intermediaries, so that if a BEEP peer authenticates itself as being from such a relay, then the data is accepted.
APEX, at its core, provides a best-effort datagram service. Options are used to alter the semantics of the core service.
The semantics of the APEX "option" element are context-specific. Accordingly, the specification of an APEX option must define:
An option registration template (Section 7.1) organizes this information.
An "option" element is contained within either a "data",
"originator", "recipient", or an "attach" element, all of which are
termed the "containing" element. The "option" element has several
attributes and contains arbitrary content:
Note that if the containing element is an "attach", then the values of the "targetHop" and "transID" attributes are ignored.
The value of the "internal" attribute is the IANA-registered name for the option. If the "internal" attribute is not present, then the value of the "external" attribute is a URI or URI with a fragment- identifier. Note that a relative-URI value is not allowed.
The "targetHop" attribute specifies which relay(s) should process the option:
this: the option applies to this relay, and must be removed prior to transmitting the containing element.
final: the option applies to this relay, only if the relay will transmit the containing element directly to the recipient.
all: the option applies to this relay and is retained for the next.
Note that a final relay does not remove any options as it transmits the containing element directly to the recipient.
The "mustUnderstand" attribute specifies whether the relay may ignore
the option if it is unrecognized, and is consulted only if the
"targetHop" attribute indicates that the option applies to that
relay. If the option applies, and if the value of the
"mustUnderstand" attribute is "true", and if the relay does not
"understand" the option, then an error in processing has occurred.
Section 8.4 contains the APEX option registration for the "statusRequest" option.
If this option is present, then each applicable relay sends a
"statusResponse" message to the originator. This is done by issuing
a data operation whose originator is the report service associated
with the issuing relay, whose recipient is the endpoint address of
the "statusRequest" originator, and whose content is a
"statusResponse" element.
A "statusRequest" option MUST NOT be present in any data operation containing a "statusResponse" element. In general, applications should be careful to avoid potential looping behaviors if an option is received in error.
Consider these examples:
+-------+ +-------+
| | -- data -------> | |
| appl. | | relay |
| #1 | <--------- ok -- | |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@example.com' />
<option internal='statusRequest' targetHop='final'
mustUnderstand='true' transID='86' />
</data>
S: <ok />
+-------+ +-------+
| | -- data -------> | |
| relay | | appl. |
| | <--------- ok -- | #2 |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@example.com' />
<option internal='statusRequest' targetHop='final'
mustUnderstand='true' transID='86' />
</data>
S: <ok />
+-------+ +-------+
| | <------- data -- | |
| appl. | | relay |
| #1 | -- ok ---------> | |
+-------+ +-------+
C: <data content='#Content'>
<originator identity='apex=report@example.com' />
<recipient identity='fred@example.com' />
<data-content Name='Content'>
<statusResponse transID='86'>
<destination identity='barney@example.com'>
<reply code='250' />
</destination>
</statusResponse>
</data-content>
</data>
S: <ok />
or
+-------+ +-------+
| | -- data -------> | |
| appl. | | relay |
| #1 | <--------- ok -- | |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@example.com' />
<option internal='statusRequest' targetHop='final'
mustUnderstand='true' transID='86' />
</data>
S: <ok />
+-------+ +-------+
| | <------- data -- | |
| appl. | | relay |
| #1 | -- ok ---------> | |
+-------+ +-------+
C: <data content='#Content'>
<originator identity='apex=report@example.com' />
<recipient identity='fred@example.com' />
<data-content Name='Content'>
<statusResponse transID='86'>
<destination identity='barney@example.com'>
<reply code='550'>unknown endpoint
identity</reply>
</destination>
</statusResponse>
</data-content>
</data>
S: <ok />
or
+-------+ +-------+
| | -- data -------> | |
| appl. | | relay |
| #1 | <--------- ok -- | #1 |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@rubble.com' />
<option internal='statusRequest' targetHop='final'
mustUnderstand='true' transID='86' />
</data>
S: <ok />
+-------+ +-------+
| | -- data -------> | |
| relay | | relay |
| #1 | <--------- ok -- | #2 |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@rubble.com' />
<option internal='statusRequest' targetHop='final'
mustUnderstand='true' transID='86' />
</data>
S: <ok />
+-------+ +-------+
| | -- data -------> | |
| relay | | appl. |
| #2 | <--------- ok -- | #2 |
+-------+ +-------+
C: <data content='cid:1@example.com'>
<originator identity='fred@example.com' />
<recipient identity='barney@example.com' />
<option internal='statusRequest' targetHop='final'
mustUnderstand='true' transID='86' />
</data>
S: <ok />
+-------+ +-------+
| | <------- data -- | |
| relay | | relay |
| #1 | -- ok ---------> | #2 |
+-------+ +-------+
C: <data content='#Content'>
<originator identity='apex=report@rubble.com' />
<recipient identity='fred@example.com' />
<data-content Name='Content'>
<statusResponse transID='86'>
<destination identity='barney@rubble.com'>
<reply code='250' />
</destination>
</statusResponse>
</data-content>
</data>
S: <ok />
+-------+ +-------+
| | <------- data -- | |
| appl. | | relay |
| #1 | -- ok ---------> | #1 |
+-------+ +-------+
C: <data content='#Content'>
<originator identity='apex=report@rubble.com' />
<recipient identity='fred@example.com' />
<data-content Name='Content'>
<statusResponse transID='86'>
<destination identity='barney@rubble.com'>
<reply code='250' />
</destination>
</statusResponse>
</data-content>
</data>
S: <ok />
Note that a trace of a data's passage through the relaying mesh can be achieved by setting the "targetHop" attribute to "all".
APEX, at its core, provides a best-effort datagram service. Within an administrative domain, all relays must be able to handle messages for any endpoint within that administrative domain. APEX services are logically defined as endpoints but, given their ubiquitous semantics, they do not necessarily need to be associated with a single physical endpoint. As such, they may be provisioned co- resident with each relay within an administrative domain, even though they are logically provided on top of the relaying mesh, i.e.,
+----------+ +----------+ +----------+ +---------+
| APEX | | APEX | | APEX | | |
| access | | presence | | report | | ... |
| service | | service | | service | | |
+----------+ +----------+ +----------+ +---------+
| | | |
| | | |
+----------------------------------------------------------------+
| |
| APEX core |
| |
+----------------------------------------------------------------+
That is, applications communicate with an APEX service by exchanging data with a "well-known endpoint" (WKE).
For example, APEX applications communicate with the report service by exchanging data with the well-known endpoint "apex=report" in the corresponding administrative domain, e.g., "apex=report@example.com" is the endpoint associated with the report service in the "example.com" administrative domain.
The specification of an APEX service must define:
A service registration template (Section 7.2) organizes this information.
Finally, note that within a single administrative domain, the relaying mesh makes use of the APEX access service in order to determine if an originator is allowed to transmit data to a recipient (c.f., Step 5.3 of Section 4.4.4.1).
The specification of an APEX service may use definitions found in the APEX core DTD (Section 9.1). For example, the reply operation (Section 6.1.2) is defined to provide a common format for responses.
In using APEX's transaction-identifiers, note the following:
For example, when an application issues the attach operation, the associated transaction-identifier has meaning only within the context of the BEEP channel used for the attach operation. When the BEEP connection is released, the channel no longer exists and the application is no longer attached to the relaying mesh.
For example, an application may attach as an endpoint, send data (containing an embedded transaction-identifier) to a service, and, some time later, detach from the relaying mesh. Later on, a second application may attach as the same endpoint, and send data of its own (also containing embedded transaction-identifiers). Subsequently, the second application may receive data from the service responding to the first application's request and containing the transaction-identifier used by the first application.
To minimize the likelihood of ambiguities with long-lived transaction-identifiers, the values of transaction-identifiers generated by applications should appear to be unpredictable.
Many APEX services make use of a reply operation. Although each service defines the circumstances in which a "reply" element is sent, the syntax of the "reply" element is defined in Section 9.1.
The "reply" element has a "code" attribute, a "transID" attribute, an optional "xml:lang" attribute, and may contain arbitrary textual content:
Section 8.5 contains the APEX service registration for the report service:
If a relay processes a "statusRequest" option (Section 5.1), then it sends data to the originator containing a "statusResponse" element (Section 9.2).
The "statusResponse" element has a "transID" attribute and contains one or more "destination" elements:
When an APEX option is registered, the following information is supplied:
Option Identification: specify the NMTOKEN or the URI that authoritatively identifies this option.
Present in: specify the APEX elements in which the option may appear.
Contains: specify the XML content that is contained within the "option" element.
Processing Rules: specify the processing rules associated with the option.
Contact Information: specify the postal and electronic contact information for the author of the profile.
When an APEX service is registered, the following information is supplied:
Well-Known Endpoint: specify the local-part of an endpoint identity, starting with "apex=".
Syntax of Messages Exchanged: specify the elements exchanged with the service.
Sequence of Messages Exchanged: specify the order in which data is exchanged with the service.
Access Control Tokens: specify the token(s) used to control access to the service (c.f., [10]).
Contact Information: specify the postal and electronic contact information for the author of the profile.
Note that the endpoints "apex=all" and "apex=core" may not be assigned.
When an APEX endpoint application is registered, the following information is supplied:
Endpoint Application: specify the subaddress used for an endpoint application, starting with "appl=".
Application Definition: specify the syntax and semantics of the endpoint application identified by this registration.
Contact Information: specify the postal and electronic contact information for the author of the profile.
Profile Identification: http://iana.org/beep/APEX
Messages exchanged during Channel Creation: "attach", "bind"
Messages starting one-to-one exchanges: "attach", "bind", "terminate", or "data"
Messages in positive replies: "ok"
Messages in negative replies: "error"
Messages in one-to-many exchanges: none
Message Syntax: c.f., Section 9.1
Message Semantics: c.f., Section 4.4
Contact Information: c.f., the "Authors' Addresses" section of this memo
Protocol Number: TCP
Message Formats, Types, Opcodes, and Sequences: c.f., Section 9.1
Functions: c.f., Section 4.4
Use of Broadcast/Multicast: none
Proposed Name: APEX relay-relay service
Short name: apex-mesh
Contact Information: c.f., the "Authors' Addresses" section of this memo
Protocol Number: TCP
Message Formats, Types, Opcodes, and Sequences: c.f., Section 9.1
Functions: c.f., Section 4.4
Use of Broadcast/Multicast: none
Proposed Name: APEX endpoint-relay service
Short name: apex-edge
Contact Information: c.f., the "Authors' Addresses" section of this memo
Option Identification: statusRequest
Present in: APEX's "data" and "recipient" elements
Contains: nothing
Processing Rules: c.f., Section 5.1
Contact Information: c.f., the "Authors' Addresses" section of this memo
Well-Known Endpoint: apex=report
Syntax of Messages Exchanged: c.f., Section 9.2
Sequence of Messages Exchanged: c.f., Section 6.2
Access Control Tokens: none
Contact Information: c.f., the "Authors' Addresses" section of this memo
<!--
DTD for the APEX core, as of 2001-07-09
Refer to this DTD as:
<!ENTITY % APEXCORE PUBLIC "-//IETF//DTD APEX CORE//EN" "">
%APEXCORE;
-->
<!ENTITY % BEEP PUBLIC "-//IETF//DTD BEEP//EN" "">
%BEEP;
<!--
DTD data types:
entity syntax/reference example
====== ================ =======
APEX endpoint
ENDPOINT entity, fred@example.com
c.f., Section 2.2
domain, either a FQDN or a literal
DOMAIN c.f., [RFC-2821] example.com or [10.0.0.1]
seconds
SECONDS 0..2147483647 600
timestamp
TIMESTAMP c.f., [12] 2000-05-15T13:02:00-08:00
unique-identifier
UNIQID 1..2147483647 42
unique-identifier OR zero
UNIZID 0..2147483647 0
-->
<!ENTITY % ENDPOINT "CDATA">
<!ENTITY % DOMAIN "CDATA">
<!ENTITY % SECONDS "CDATA">
<!ENTITY % TIMESTAMP "CDATA">
<!ENTITY % UNIQID "CDATA">
<!ENTITY % UNIZID "CDATA">
<!--
APEX messages, exchanged as application/beep+xml
role MSG RPY ERR
====== === === ===
I attach ok error
I or L bind ok error
I or L terminate ok error
I or L data ok error
-->
<!ELEMENT attach (option*)>
<!ATTLIST attach
endpoint %ENDPOINT; #REQUIRED
transID %UNIQID; #REQUIRED>
<!ELEMENT bind (option*)>
<!ATTLIST bind
relay %DOMAIN; #REQUIRED
transID %UNIQID; #REQUIRED>
<!ELEMENT terminate (#PCDATA)>
<!ATTLIST terminate
code %XYZ; "250"
xml:lang %LANG; #IMPLIED
transID %UNIZID; "0">
<!ELEMENT data (originator,recipient+,option*,data-content?)>
<!ATTLIST data
content %URI; #REQUIRED>
<!ELEMENT originator (option*)>
<!ATTLIST originator
identity %ENDPOINT; #REQUIRED>
<!ELEMENT recipient (option*)>
<!ATTLIST recipient
identity %ENDPOINT; #REQUIRED>
<!ELEMENT data-content
ANY>
<!ATTLIST Name ID #REQUIRED>
<!ELEMENT ok EMPTY>
<!ELEMENT reply (#PCDATA)>
<!ATTLIST reply
code %XYZ; #REQUIRED
transID %UNIQID; #REQUIRED
xml:lang %LANG; #IMPLIED>
<!-- either the "internal" or the "external" attribute is present in
an option -->
<!ELEMENT option ANY>
<!ATTLIST option
internal NMTOKEN ""
external %URI; ""
targetHop (this|final|all) "final"
mustUnderstand
(true|false) "false"
transID %UNIQID; #REQUIRED
localize %LOCS; "i-default">
<!--
DTD for the APEX report service, as of 2000-12-12
Refer to this DTD as:
<!ENTITY % APEXREPORT PUBLIC "-//Blocks//DTD APEX REPORT//EN" "">
%APEXREPORT;
-->
<!ENTITY % APEXCORE PUBLIC "-//Blocks//DTD APEX CORE//EN" "">
%APEXCORE;
<!--
Synopsis of the APEX report service
service WKE: apex=report
message exchanges:
service initiates consumer replies
================= ================
statusResponse (nothing)
access control tokens: none
-->
<!ELEMENT statusResponse
(destination+)>
<!ATTLIST statusResponse
transID %UNIQID; #REQUIRED>
<!ELEMENT destination (reply)>
<!ATTLIST destination
identity %ENDPOINT; #REQUIRED>
code meaning
==== =======
250 transaction successful
421 service not available
450 requested action not taken
451 requested action aborted
454 temporary authentication failure
500 general syntax error (e.g., poorly-formed XML)
501 syntax error in parameters (e.g., non-valid XML)
504 parameter not implemented
530 authentication required
534 authentication mechanism insufficient
535 authentication failure
537 action not authorized for user
538 authentication mechanism requires encryption
550 requested action not taken
553 parameter invalid
554 transaction failed (e.g., policy violation)
555 transaction already in progress
Consult Section 3 and Section 4.5 for a discussion of security issues, e.g., relaying integrity.
Although service provisioning is a policy matter, at a minimum, all APEX implementations must provide the following tuning profiles:
for authentication: http://iana.org/beep/SASL/DIGEST-MD5
for confidentiality: http://iana.org/beep/TLS (using the
TLS_RSA_WITH_3DES_EDE_CBC_SHA cipher)
for both: http://iana.org/beep/TLS (using the
TLS_RSA_WITH_3DES_EDE_CBC_SHA cipher supporting client-side
certificates)
Further, APEX endpoint implementations may choose to offer MIME-based security services providing message integrity and confidentiality, such as OpenPGP [13] or S/MIME [14].
Regardless, since APEX is a profile of the BEEP, consult [1]'s Section 9 for a discussion of BEEP-specific security issues.
Finally, the statusRequest option (Section 5.1) may be used to expose private network topology. Accordingly, an administrator may wish to choose to disable this option except at the ingress/egress points for its administrative domain.
[1] Rose, M., "The Blocks Extensible Exchange Protocol Core", RFC
3080, March 2001.
[2] Crocker, D. and P. Overell, "Augmented BNF for Syntax
Specifications: ABNF", RFC 2234, November 1997.
[3] Klensin, J., "Simple Mail Transfer Protocol", RFC 2821, April
2001.
[4] Yergeau, F., "UTF-8, a transformation format of Unicode and ISO
10646", RFC 2044, October 1996.
[5] Gulbrandsen, A., Vixie, P. and L. Esibov, "A DNS RR for
specifying the location of services (DNS SRV)", RFC 2782,
February 2000.
[6] Berners-Lee, T., Fielding, R. and L. Masinter, "Uniform
Resource Identifiers (URI): Generic Syntax", RFC 2396, August
1998.
[7] Levinson, E., "The MIME Multipart/Related Content-type", RFC
2387, August 1998.
[8] Levinson, E., "Content-ID and Message-ID Uniform Resource
Locators", RFC 2392, August 1998.
[9] Freed, N. and N. Borenstein, "Multipurpose Internet Mail
Extensions (MIME) Part One: Format of Internet Message Bodies",
RFC 2045, November 1996.
[10] Rose, M., Klyne, G. and D. Crocker, "The Application Exchange (APEX) Access Service", RFC 3341, July 2002.
[11] Rose, M., Klyne, G. and D. Crocker, "The Application Exchange (APEX) Presence Service", Work in Progress.
[12] Newman, C. and G. Klyne, "Date and Time on the Internet: Timestamps", RFC 3339, July 2002.
[13] Elkins, M., Del Torto, D., Levien, R. and T. Roessler, "MIME Security with OpenPGP", RFC 3156, August 2001.
[14] Ramsdell, B., "S/MIME Version 3 Message Specification", RFC 2633, June 1999.
The authors gratefully acknowledge the contributions of: Jeffrey Altman, Harald Alvestrand, Eric Dixon, Ronan Klyne, Darren New, Chris Newman, Scott Pead, and Bob Wyman.
The IANA has registered "APEX" as a standards-track BEEP profile, as specified in Section 8.1.
The IANA has registered "apex-mesh" as a TCP port number, as specified in Section 8.2.
The IANA has registered "apex-edge" as a TCP port number, as specified in Section 8.3.
The IANA maintains a list of:
For each list, the IESG is responsible for assigning a designated expert to review the specification prior to the IANA making the assignment. As a courtesy to developers of non-standards track APEX options and services, the mailing list apexwg@invisible.net may be used to solicit commentary.
The IANA makes the registrations specified in Section 8.4 and Section 8.5.
Marshall T. Rose
Dover Beach Consulting, Inc.
POB 255268
Sacramento, CA 95865-5268
US
Phone: +1 916 483 8878
EMail: mrose@dbc.mtview.ca.us
Graham Klyne
Clearswift Corporation
1310 Waterside
Arlington Business Park
Theale, Reading RG7 4SA
UK
Phone: +44 11 8903 8903
EMail: Graham.Klyne@MIMEsweeper.com
David H. Crocker
Brandenburg InternetWorking
675 Spruce Drive
Sunnyvale, CA 94086
US
Phone: +1 408 246 8253
EMail: dcrocker@brandenburg.com
URI: http://www.brandenburg.com/
Copyright © The Internet Society (2002). 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.
Funding for the RFC Editor function is currently provided by the Internet Society.