One document matched: draft-ietf-sipcore-rfc3265bis-00.xml
<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd"[
<!ENTITY rfc2119 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.2119.xml'>
<!ENTITY rfc2434 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.2434.xml'>
<!ENTITY rfc2616 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.2616.xml'>
<!ENTITY rfc2779 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.2779.xml'>
<!ENTITY rfc2848 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.2848.xml'>
<!ENTITY rfc3261 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.3261.xml'>
<!ENTITY rfc3265 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.3265.xml'>
<!ENTITY rfc3515 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.3515.xml'>
<!ENTITY rfc3840 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.3840.xml'>
<!ENTITY rfc3903 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.3903.xml'>
<!ENTITY rfc4483 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.4483.xml'>
<!ENTITY rfc4485 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.4485.xml'>
<!ENTITY rfc4660 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.4660.xml'>
<!ENTITY rfc5057 PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml/reference.RFC.5057.xml'>
<!ENTITY draft-ietf-sip-gruu PUBLIC '' 'http://xml.resource.org/public/rfc/bibxml3/reference.I-D.draft-ietf-sip-gruu-15.xml'>
]>
<?xml-stylesheet type='text/xsl' href='rfc2629.xslt' ?>
<?rfc toc="yes"?>
<?rfc compact="yes" ?>
<?rfc sortrefs="no" ?>
<?rfc symrefs="yes" ?>
<rfc ipr="pre5378Trust200902">
<front>
<title>SIP-Specific Event Notification</title>
<author fullname="Adam Roach" initials="A. B." surname="Roach">
<organization>Tekelec</organization>
<address>
<postal>
<street>17210 Campbell Rd.</street>
<street>Suite 250</street>
<city>Dallas</city>
<region>TX</region>
<code>75252</code>
<country>US</country>
</postal>
<email>adam@nostrum.com</email>
</address>
</author>
<date month="July" year="2009" />
<area>Real Time Applications and Infrastructure</area>
<abstract>
<t>This document describes an extension to the Session Initiation
Protocol (SIP). The purpose of this extension is to provide an
extensible framework by which SIP nodes can request notification from
remote nodes indicating that certain events have occurred.</t>
<t>Note that the event notification mechanisms defined herein are NOT
intended to be a general-purpose infrastructure for all classes of event
subscription and notification.</t>
</abstract>
</front>
<middle>
<section title="Introduction">
<t>The ability to request asynchronous notification of events proves
useful in many types of SIP services for which cooperation between
end-nodes is required. Examples of such services include automatic
callback services (based on terminal state events), buddy lists (based
on user presence events), message waiting indications (based on mailbox
state change events), and PSTN and Internet Internetworking (PINT) <xref
target="RFC2848"></xref> status (based on call state events).</t>
<t>The methods described in this document provide a framework by which
notification of these events can be ordered.</t>
<t>The event notification mechanisms defined herein are NOT intended to
be a general-purpose infrastructure for all classes of event
subscription and notification. Meeting requirements for the general
problem set of subscription and notification is far too complex for a
single protocol. Our goal is to provide a SIP-specific framework for
event notification which is not so complex as to be unusable for simple
features, but which is still flexible enough to provide powerful
services. Note, however, that event packages based on this framework may
define arbitrarily elaborate rules which govern the subscription and
notification for the events or classes of events they describe.</t>
<t>This document does not describe an extension which may be used directly;
it must be extended by other documents (herein referred to as "event
packages"). In object-oriented design terminology, it may be thought of
as an abstract base class which must be derived into an instantiatable
class by further extensions. Guidelines for creating these extensions
are described in <xref target="event-packages"></xref>.</t>
<section title="Overview of Operation">
<t>The general concept is that entities in the network can subscribe
to resource or call state for various resources or calls in the
network, and those entities (or entities acting on their behalf) can
send notifications when those states change.</t>
<t>A typical flow of messages would be:</t>
<figure>
<artwork><![CDATA[
Subscriber Notifier
|-----SUBSCRIBE---->| Request state subscription
|<-------200--------| Acknowledge subscription
|<------NOTIFY----- | Return current state information
|--------200------->|
|<------NOTIFY----- | Return current state information
|--------200------->|
]]></artwork>
</figure>
<t>Subscriptions are expired and must be refreshed by subsequent
SUBSCRIBE messages.</t>
</section>
<section title="Documentation Conventions">
<t>There are several paragraphs throughout this document which provide
motivational or clarifying text. Such passages are non-normative, and
are provided only to assist with reader comprehension. These passages
are set off from the remainder of the text by being indented thus:</t>
<t><list style="hanging">
<t>This is an example of non-normative explanatory text. It does not
form part of the specification, and is used only for
clarification.</t>
</list></t>
<t>The all-capital terms "MUST", "SHOULD", "MAY", "SHOULD NOT",
"MUST NOT", and "RECOMMENDED" are used as defined in
<xref target="RFC2119" />. In particular, implementors need to
take careful note of the meaning of "SHOULD" defined in RFC 2119.
To rephrase: violation of SHOULD-strength requirements requires
careful analysis and clearly enumerable reasons.
It is inappropriate to fail to comply with "SHOULD"-strength
requirements whimsically or for ease of implementation.</t>
<t>The use of quotation marks next to periods and commas follows the
convention used by the American Mathematical Society; although
contrary to traditional American English convention, this usage lends
clarity to certain passages.</t>
</section>
</section>
<section anchor="definitions" title="Definitions">
<t><vspace /></t>
<t><list style="hanging">
<t hangText="Event Package:">An event package is an additional
specification which defines a set of state information to be reported
by a notifier to a subscriber. Event packages also define further
syntax and semantics based on the framework defined by this document
required to convey such state information.</t>
</list></t>
<t><vspace /></t>
<t><list style="hanging">
<t hangText="Event Template-Package:">An event template-package is a
special kind of event package which defines a set of states which may
be applied to all possible event packages, including itself.</t>
</list></t>
<t><vspace /></t>
<t><list style="hanging">
<t hangText="Notification:">Notification is the act of a notifier
sending a NOTIFY message to a subscriber to inform the subscriber of
the state of a resource.</t>
</list></t>
<t><vspace /></t>
<t><list style="hanging">
<t hangText="Notifier:">A notifier is a user agent which generates
NOTIFY requests for the purpose of notifying subscribers of the state
of a resource. Notifiers typically also accept SUBSCRIBE requests to
create subscriptions.</t>
</list></t>
<t><vspace /></t>
<t><list style="hanging">
<t hangText="Subscriber:">A subscriber is a user agent which receives
NOTIFY requests from notifiers; these NOTIFY requests contain
information about the state of a resource in which the subscriber is
interested. Subscribers typically also generate SUBSCRIBE requests and
send them to notifiers to create subscriptions.</t>
</list></t>
<t><vspace /></t>
<t><list style="hanging">
<t hangText="Subscription:">A subscription is a set of application
state associated with a dialog. This application state includes a
pointer to the associated dialog, the event package name, and possibly
an identification token. Event packages will define additional
subscription state information. By definition, subscriptions exist in
both a subscriber and a notifier.</t>
</list></t>
<t><vspace /></t>
<t><list style="hanging">
<t hangText="Subscription Migration:">Subscription migration is the
act of moving a subscription from one notifier to another
notifier.</t>
</list></t>
</section>
<section anchor="methods" title="SIP Methods for Event Notification">
<section anchor="subscribe" title="SUBSCRIBE">
<t>The SUBSCRIBE method is used to request current state and state
updates from a remote node.
SUBSCRIBE is a target refresh request, as that term is defined in SIP
<xref target="RFC3261"/>.
</t>
<section anchor="duration" title="Subscription Duration">
<t>SUBSCRIBE requests SHOULD contain an "Expires" header field (defined in
SIP <xref target="RFC3261" />). This expires value indicates the
duration of the subscription. In order to keep subscriptions
effective beyond the duration communicated in the "Expires" header field,
subscribers need to refresh subscriptions on a periodic basis using
a new SUBSCRIBE message on the same dialog as defined in SIP <xref
target="RFC3261" />.</t>
<t>If no "Expires" header field is present in a SUBSCRIBE request, the
implied default is defined by the event package being used.</t>
<t>200-class responses to SUBSCRIBE requests also MUST contain an
"Expires" header field. The period of time in the response MAY be shorter
but MUST NOT be longer than specified in the request. The notifier
is explicitly allowed to shorten the duration to zero.
The period of time in the response is the one which defines the duration
of the subscription.</t>
<t>An "expires" parameter on the "Contact" header field has no semantics
for SUBSCRIBE and is explicitly not equivalent to an "Expires"
header field in a SUBSCRIBE request or response.</t>
<t>A natural consequence of this scheme is that a SUBSCRIBE with an
"Expires" of 0 constitutes a request to unsubscribe from an
event.</t>
<t><list style="hanging">
<t>In addition to being a request to unsubscribe, a SUBSCRIBE
message with "Expires" of 0 also causes a fetch of state; see
<xref target="polling-state" />.</t>
</list></t>
<t>Notifiers may also wish to cancel subscriptions to events; this
is useful, for example, when the resource to which a subscription
refers is no longer available. Further details on this mechanism are
discussed in <xref target="notifier-notifies" />.</t>
</section>
<section anchor="event-identification"
title="Identification of Subscribed Events and Event Classes">
<t>Identification of events is provided by three pieces of
information: Request URI, Event Type, and (optionally) message
body.</t>
<t>The Request URI of a SUBSCRIBE request, most importantly,
contains enough information to route the request to the appropriate
entity per the request routing procedures outlined in SIP <xref
target="RFC3261"></xref>. It also contains enough information to
identify the resource for which event notification is desired, but
not necessarily enough information to uniquely identify the nature
of the event (e.g., "sip:adam@example.com" would be an
appropriate URI to subscribe to for my presence state; it would also
be an appropriate URI to subscribe to the state of my voice
mailbox).</t>
<t>Subscribers MUST include exactly one "Event" header field in SUBSCRIBE
requests, indicating to which event or class of events they are
subscribing. The "Event" header field will contain a token which indicates
the type of state for which a subscription is being requested. This
token will be registered with the IANA and will correspond to an
event package which further describes the semantics of the event or
event class.</t>
<t>If the event package to which the event token corresponds defines
behavior associated with the body of its SUBSCRIBE requests, those
semantics apply.</t>
<t>Event packages may also define parameters for the Event header field;
if they do so, they must define the semantics for such
parameters.</t>
</section>
<section title="Additional SUBSCRIBE Header Values">
<t>Because SUBSCRIBE requests create a dialog as defined in SIP
<xref target="RFC3261"></xref>, they MAY contain an "Accept" header field.
This header field, if present, indicates the body formats allowed in
subsequent NOTIFY requests. Event packages MUST define the behavior
for SUBSCRIBE requests without "Accept" header fields; usually, this will
connote a single, default body type.</t>
<t>Header values not described in this document are to be
interpreted as described in SIP <xref target="RFC3261"></xref>.</t>
</section>
</section>
<section anchor="notify" title="NOTIFY">
<t>NOTIFY messages are sent to inform subscribers of changes in state
to which the subscriber has a subscription. Subscriptions are
typically put in place using the SUBSCRIBE method; however, it is
possible that other means have been used.</t>
<t>
NOTIFY is a target refresh request, as that term is defined in SIP <xref target="RFC3261"/>.
</t>
<t>A NOTIFY does not terminate its corresponding subscription; in
other words, a single SUBSCRIBE request may trigger several NOTIFY
requests.</t>
<section title="Identification of Reported Events, Event Classes, and Current State">
<t>Identification of events being reported in a notification is very
similar to that described for subscription to events (see <xref
target="event-identification"></xref>).</t>
<t>As in SUBSCRIBE requests, NOTIFY "Event" header fields will contain a
single event package name for which a notification is being
generated. The package name in the "Event" header field MUST match the
"Event" header field in the corresponding SUBSCRIBE message.</t>
<t>Event packages may define semantics associated with the body of
their NOTIFY requests; if they do so, those semantics apply. NOTIFY
bodies are expected to provide additional details about the nature
of the event which has occurred and the resultant resource
state.</t>
<t>When present, the body of the NOTIFY request MUST be formatted
into one of the body formats specified in the "Accept" header field of the
corresponding SUBSCRIBE request. This body will contain either the
state of the subscribed resource or a pointer to such state in the
form of a URI (see <xref target="uri-use" />).</t>
</section>
</section>
</section>
<section anchor="node-behavior" title="Node Behavior">
<section anchor="subscriber" title="Subscriber Behavior">
<section title="Detecting Support for SIP Events">
<t>The extension described in this document does not
make use of the use of "Require" or
"Proxy-Require" header fields; similarly, there is no token defined for
"Supported" header fields. Potential subscribers may probe for the support
of SIP Events using the OPTIONS request defined in
SIP <xref target="RFC3261" />. </t>
<t>The presence of "SUBSCRIBE" in the
"Allow" header field of any request or response indicates
support for SIP Events; further, in the absence of an "Allow"
header field, the simple presence of an "Allow-Events" header field
is sufficient to indicate that the node that sent the message
is capable of acting as a notifier (see <xref target="allow-events"/>.</t>
<t><list style="hanging">
<t>The "methods" parameter for Contact may also be used to
specifically announce support for SUBSCRIBE and NOTIFY messages
when registering. (See <xref target="RFC3840" /> for
details on the "methods" parameter).</t>
</list></t>
</section>
<section anchor="subscribing" title="Creating and Maintaining Subscriptions">
<t>
From the subscriber's perspective, a subscription proceeds
according to the following state diagram:
</t>
<figure>
<artwork><![CDATA[
+-------------+
| init |<-----------------------+
+-------------+ |
| Retry-after
Send SUBSCRIBE expires
| |
V Timer L Fires; |
+-------------+ SUBSCRIBE failure |
+------------| notify_wait |-- response; --------+ |
| +-------------+ or NOTIFY, | |
| | state=terminated | |
| | | |
++========|===================|============================|==|====++
|| | | V | ||
|| Receive NOTIFY, Receive NOTIFY, +-------------+ ||
|| state=active state=pending | terminated | ||
|| | | +-------------+ ||
|| | | A A ||
|| | V Receive NOTIFY, | | ||
|| | +-------------+ state=terminated; | | ||
|| | | pending |-- or 481 response ----+ | ||
|| | +-------------+ to SUBSCRIBE | ||
|| | | refresh | ||
|| | Receive NOTIFY, | ||
|| | state=active | ||
|| | | | ||
|| | V Receive NOTIFY, | ||
|| | +-------------+ state=terminated; | ||
|| +----------->| active |-- or 481 response -------+ ||
|| +-------------+ to SUBSCRIBE ||
|| refresh ||
|| Subscription ||
++=================================================================++
]]></artwork>
</figure>
<t>
Any transition from "notify_wait" into a "pending" or
"active" state results in a new subscription. Note
that multiple subscriptions can be generated as the
result of a single SUBSCRIBE request (see
<xref target="dialog-handling" />). Each of these
new subscriptions exists in its own independent
state machine.
</t>
<section anchor="request-sub" title="Requesting a Subscription">
<t>SUBSCRIBE is a dialog-creating method, as described in SIP
<xref target="RFC3261"></xref>.</t>
<t>When a subscriber wishes to subscribe to a particular state for
a resource, it forms a SUBSCRIBE message. If the initial SUBSCRIBE
represents a request outside of a dialog (as it typically will),
its construction follows the procedures outlined in SIP <xref
target="RFC3261"></xref> for UAC request generation outside of a
dialog.</t>
<t>This SUBSCRIBE request will be confirmed with a final response.
200-class responses indicate that the subscription has been
accepted, and that a NOTIFY will be sent immediately. A 200
response indicates that the subscription has been accepted and
that the user is authorized to subscribe to the requested
resource. A 202 response merely indicates that the subscription
has been understood, and that authorization may or may not have
been granted.</t>
<t>The "Expires" header field in a 200-class response to SUBSCRIBE
indicates the actual duration for which the subscription will
remain active (unless refreshed).</t>
<t>Non-200 class final responses indicate that no subscription or
dialog has been created, and no subsequent NOTIFY message will be
sent. All non-200 class responses (with the exception of "489",
described herein) have the same meanings and handling as described
in SIP <xref target="RFC3261"></xref>.</t>
</section>
<section title="Refreshing of Subscriptions">
<t>At any time before a subscription expires, the subscriber may
refresh the timer on such a subscription by sending another
SUBSCRIBE request on the same dialog as the existing subscription.
The handling for such a
request is the same as for the initial creation of a subscription
except as described below.</t>
<t>If a SUBSCRIBE request to refresh a subscription receives a
404, 405, 410, 416, 480-485, 489, 501, or 604 response, the
subscriber should consider the subscription terminated.
(See
<xref target="RFC5057"/> for further details and notes about the
effect of error codes on dialogs and usages within dialog,
such as subscriptions).
If the subscriber wishes to re-subscribe to
the state, he does so by composing an unrelated initial SUBSCRIBE
request with a freshly-generated Call-ID and a new, unique "From"
tag (see <xref target="request-sub" />.)</t>
<t>If a SUBSCRIBE request to refresh a subscription fails with
any error code other than those listed above, the original
subscription is still considered
valid for the duration of the most recently known "Expires" value
as negotiated by SUBSCRIBE and its response, or as communicated by
NOTIFY in the "Subscription-State" header field "expires" parameter.</t>
<t><list style="hanging">
<t>Note that many such errors indicate that there may be a
problem with the network or the notifier such that no further
NOTIFY messages will be received.</t>
</list></t>
</section>
<section anchor="unsubscribing" title="Unsubscribing">
<t>Unsubscribing is handled in the same way as refreshing of a
subscription, with the "Expires" header field set to "0". Note that a
successful unsubscription will also trigger a final NOTIFY
message.</t>
<t>
The final NOTIFY may or may not contain information about the
state of the resource; subscribers need to be prepared to
receive final NOTIFY messages both with and without state.
</t>
</section>
<section anchor="subscription-confirmation" title="Confirmation of Subscription Creation">
<t>The subscriber can expect to receive a NOTIFY message from each
node which has processed a successful subscription or subscription
refresh. To ensure that subscribers do not wait indefinitely
for a subscription to be established, a subscriber starts a
Timer L, set to 64*T1. If this Timer L expires prior to the
receipt of a NOTIFY message, the subscriber considers the
subscription failed, and cleans up any state associated with
the subscription attempt.</t>
<t>
Until Timer L expires, several NOTIFY messages may arrive
from different destinations (see <xref target="dialog-handling" />).
Each of these messages establish a new dialog and a new
subscription. After the expiration of Timer L, the subscriber
SHOULD reject any such NOTIFY messages that would otherwise
establish a new dialog with a "481" response code.
</t>
<t>Until the first NOTIFY message arrives, the subscriber
should consider the state of the subscribed resource to be in a
neutral state. Documents which define new event packages MUST
define this "neutral state" in such a way that makes sense for
their application (see <xref target="template-notifier-notifies"></xref>).</t>
<t>Due to the potential for both out-of-order messages and
forking, the subscriber MUST be prepared to receive NOTIFY
messages before the SUBSCRIBE transaction has completed.</t>
<t>Except as noted above, processing of this NOTIFY is the same as
in <xref target="subscriber-notifies"></xref>.</t>
</section>
</section>
<section anchor="subscriber-notifies"
title="Receiving and Processing State Information">
<t>
Subscribers receive information about the state of a resource
to which they have subscribed in the form of NOTIFY requests.
</t>
<t>Upon receiving a NOTIFY request, the subscriber should check that
it matches at least one of its outstanding subscriptions; if not, it
MUST return a "481 Subscription does not exist" response unless
another 400- or 500-class response is more appropriate. The rules
for matching NOTIFY requests with subscriptions that create a new
dialog are described in
<xref target="dialog-handling" />.
Notifications for
subscriptions which were created inside an existing dialog match if
they are in the same dialog and the "Event" header fields match (as
described in <xref target="event-header field" />).</t>
<t>If, for some reason, the event package designated in the "Event"
header field of the NOTIFY request is not supported, the subscriber will
respond with a "489 Bad Event" response.</t>
<t>To prevent spoofing of events, NOTIFY requests SHOULD be
authenticated, using any defined SIP authentication mechanism.</t>
<t>NOTIFY requests MUST contain "Subscription-State" header fields which
indicate the status of the subscription.</t>
<t>If the "Subscription-State" header field value is "active", it means
that the subscription has been accepted and (in general) has been
authorized. If the header field also contains an "expires" parameter, the
subscriber SHOULD take it as the authoritative subscription duration
and adjust accordingly. The "retry-after" and "reason" parameters
have no semantics for "active".</t>
<t>If the "Subscription-State" value is "pending", the subscription
has been received by the notifier, but there is insufficient policy
information to grant or deny the subscription yet. If the header field
also contains an "expires" parameter, the subscriber SHOULD take it
as the authoritative subscription duration and adjust accordingly.
No further action is necessary on the part of the subscriber. The
"retry-after" and "reason" parameters have no semantics for
"pending".</t>
<t>If the "Subscription-State" value is "terminated", the subscriber
should consider the subscription terminated. The "expires" parameter
has no semantics for "terminated" -- notifiers SHOULD NOT include
an "expires" parameter on a "Subscription-State" header field with a
value of "terminated," and subscribers MUST ignore any such parameter,
if present. If a reason code is present, the
client should behave as described below. If no reason code or an
unknown reason code is present, the client MAY attempt to
re-subscribe at any time (unless a "retry-after" parameter is
present, in which case the client SHOULD NOT attempt re-subscription
until after the number of seconds specified by the "retry-after"
parameter). The reason codes defined by this document are:</t>
<t></t>
<t><list style="hanging">
<t hangText="deactivated:">The subscription has been terminated,
but the subscriber SHOULD retry immediately with a new subscription.
One primary use of such a status code is to allow migration of
subscriptions between nodes. The "retry-after" parameter has no
semantics for "deactivated".</t>
</list></t>
<t></t>
<t><list style="hanging">
<t hangText="probation:">The subscription has been terminated, but
the client SHOULD retry at some later time. If a "retry-after"
parameter is also present, the client SHOULD wait at least the
number of seconds specified by that parameter before attempting to
re-subscribe.</t>
</list></t>
<t></t>
<t><list style="hanging">
<t hangText="rejected:">The subscription has been terminated due
to change in authorization policy. Clients SHOULD NOT attempt to
re-subscribe. The "retry-after" parameter has no semantics for
"rejected".</t>
</list></t>
<t></t>
<t><list style="hanging">
<t hangText="timeout:">The subscription has been terminated
because it was not refreshed before it expired. Clients MAY
re-subscribe immediately. The "retry-after" parameter has no
semantics for "timeout". This reason code is also associated
with polling of resource state, as detailed in <xref target="polling-state"/></t>
</list></t>
<t></t>
<t><list style="hanging">
<t hangText="giveup:">The subscription has been terminated because
the notifier could not obtain authorization in a timely fashion.
If a "retry-after" parameter is also present, the client SHOULD
wait at least the number of seconds specified by that parameter
before attempting to re-subscribe; otherwise, the client MAY retry
immediately, but will likely get put back into pending state.</t>
</list></t>
<t></t>
<t><list style="hanging">
<t hangText="noresource:">The subscription has been terminated
because the resource state which was being monitored no longer
exists. Clients SHOULD NOT attempt to re-subscribe. The
"retry-after" parameter has no semantics for "noresource".</t>
</list></t>
<t></t>
<t><list style="hanging">
<t hangText="invariant:">The subscription has been terminated
because the resource state is guaranteed not to change for
the foreseeable future. This may be the case, for example, when
subscribing to the location information of a fixed-location
land-line telephone. When using this reason code, notifiers
are advised to include a
"retry-after" parameter with a large value (for example,
31536000 -- or one year) to prevent older, RFC 3265-compliant
clients from periodically resubscribing. Clients SHOULD NOT
attempt to resubscribe after receiving a reason code of "invariant,"
regardless of the presence of or value of a "retry-after" parameter.
</t>
</list></t>
<t>
Other specifications may define new reason codes for use with
the "Subscription-State" header field.
</t>
<t>Once the notification is deemed acceptable to the subscriber, the
subscriber SHOULD return a 200 response. In general, it is not
expected that NOTIFY responses will contain bodies; however, they
MAY, if the NOTIFY request contained an "Accept" header field.</t>
<t>Other responses defined in SIP <xref target="RFC3261" /> may also
be returned, as appropriate. In no case should a NOTIFY transaction
extend for any longer than the time necessary for automated
processing. In particular, subscribers MUST NOT wait for a user
response before returning a final response to a NOTIFY request.</t>
</section>
<section title="Forking of SUBSCRIBE Messages">
<t>In accordance with the rules for proxying non-INVITE requests as
defined in SIP <xref target="RFC3261"></xref>, successful SUBSCRIBE
requests will receive only one 200-class response; however, due to
forking, the subscription may have been accepted by multiple nodes.
The subscriber MUST therefore be prepared to receive NOTIFY requests
with "From:" tags which differ from the "To:" tag received in the
SUBSCRIBE 200-class response.</t>
<t>If multiple NOTIFY messages are received in different dialogs
in response to a single
SUBSCRIBE message, each dialog represents a different destination
to which
the SUBSCRIBE request was forked. Subscriber
handling in such situations varies by event package;
see <xref target="forking"/> for details.</t>
</section>
</section>
<section anchor="notifier" title="Notifier Behavior">
<section anchor="notifier-subscribe"
title="Subscription Establishment and Maintenance">
<t>
Notifiers learn about subscription requests by receiving
SUBSCRIBE requests from interested parties.
Notifiers MUST NOT create subscriptions except upon receipt
of a SUBSCRIBE message. However, for historical reasons,
the implicit creation of subscriptions as defined in
<xref target="RFC3515"/> is still permitted.
</t>
<t><list style="hanging">
<t><xref target="RFC3265"/> allowed the creation of subscriptions
using means other than SUBSCRIBE. The only standardized use
of this mechanism is the REFER method <xref target="RFC3515"/>.
Implementation experience with REFER has shown that the implicit
creation of a subscription has a number of undesirable effects,
such as the inability to signal the success of a REFER while
signaling a problem with the subscription; and difficulty
performing one action without the other. Additionally, the
proper exchange of dialog identifiers is difficult without
dialog re-use (which has its own set of problems; see
<xref target="device-targeting"/>). </t>
</list></t>
<section anchor="notifier-initial-subscribe"
title="Initial SUBSCRIBE Transaction Processing">
<t>In no case should a SUBSCRIBE transaction extend for any longer
than the time necessary for automated processing. In particular,
notifiers MUST NOT wait for a user response before returning a
final response to a SUBSCRIBE request.</t>
<t><list style="hanging">
<t>This requirement is imposed primarily to prevent the
non-INVITE transaction timeout timer F (see <xref
target="RFC3261" />) from firing during the SUBSCRIBE transaction,
since interaction with a user would often exceed 64*T1
seconds.</t>
</list></t>
<t>The notifier SHOULD check that the event package specified in
the "Event" header field is understood. If not, the notifier SHOULD
return a "489 Bad Event" response to indicate that the specified
event/event class is not understood.</t>
<t>The notifier SHOULD also perform any necessary authentication
and authorization per its local policy. See <xref
target="subscribe-auth" />.</t>
<t>The notifier MAY also check that the duration in the "Expires"
header field is not too small. If and only if the expiration interval is
greater than zero AND smaller than one hour AND less than a
notifier-configured minimum, the notifier MAY return a "423
Interval Too Brief" error which contains a "Min-Expires" header field
field. The "Min-Expires" header field is described in SIP <xref
target="RFC3261" />.</t>
<t>If the notifier is able to immediately determine that it
understands the event package, that the authenticated subscriber
is authorized to subscribe, and that there are no other barriers
to creating the subscription, it creates the subscription and a
dialog (if necessary), and returns a "200 OK" response (unless
doing so would reveal authorization policy in an undesirable
fashion; see <xref target="notifier-privacy" />).</t>
<t>If the notifier cannot immediately create the subscription
(e.g., it needs to wait for user input for authorization, or is
acting for another node which is not currently reachable), or
wishes to mask authorization policy, it will return a "202
Accepted" response. This response indicates that the request has
been received and understood, but does not necessarily imply that
the subscription has been authorized yet.</t>
<t>When a subscription is created in the notifier, it stores the
event package name as part of the subscription information.</t>
<t>The "Expires" values present in SUBSCRIBE 200-class responses
behave in the same way as they do in REGISTER responses: the
server MAY shorten the interval, but MUST NOT lengthen it.</t>
<t><list style="hanging">
<t>If the duration specified in a SUBSCRIBE message is
unacceptably short, the notifier may be able to send a 423
response, as described earlier in this section.</t>
</list></t>
<t>200-class responses to SUBSCRIBE requests will not generally
contain any useful information beyond subscription duration; their
primary purpose is to serve as a reliability mechanism. State
information will be communicated via a subsequent NOTIFY request
from the notifier.</t>
<t>The other response codes defined in SIP <xref target="RFC3261" />
may be used in response to SUBSCRIBE requests, as appropriate.</t>
</section>
<section title="Confirmation of Subscription Creation/Refreshing">
<t>Upon successfully accepting or refreshing a subscription,
notifiers MUST send a NOTIFY message immediately to communicate
the current resource state to the subscriber. This NOTIFY message
is sent on the same dialog as created by the SUBSCRIBE response.
If the resource has no meaningful state at the time that the
SUBSCRIBE message is processed, this NOTIFY message MAY contain an
empty or neutral body. See <xref target="notifier-notifies"></xref> for
further details on NOTIFY message generation.</t>
<t>Note that a NOTIFY message is always sent immediately after any
200-class response to a SUBSCRIBE request, regardless of whether
the subscription has already been authorized.</t>
</section>
<section anchor="subscribe-auth"
title="Authentication/Authorization of SUBSCRIBE requests">
<t>Privacy concerns may require that notifiers apply policy to
determine whether a particular subscriber is authorized to
subscribe to a certain set of events. Such policy may be defined
by mechanisms such as access control lists or real-time
interaction with a user. In general, authorization of subscribers
prior to authentication is not particularly useful.</t>
<t>SIP authentication mechanisms are discussed in SIP <xref
target="RFC3261" />. Note that, even if the notifier node typically
acts as a proxy, authentication for SUBSCRIBE requests will always
be performed via a "401" response, not a "407;" notifiers always
act as a user agents when accepting subscriptions and sending
notifications.</t>
<t><list style="hanging">
<t>Of course, when acting as a proxy, a node will perform normal
proxy authentication (using 407). The foregoing explanation is a
reminder that notifiers are always UAs, and as such perform UA
authentication.</t>
</list></t>
<t>If authorization fails based on an access list or some other
automated mechanism (i.e., it can be automatically authoritatively
determined that the subscriber is not authorized to subscribe),
the notifier SHOULD reply to the request with a "403 Forbidden" or
"603 Decline" response, unless doing so might reveal information
that should stay private; see <xref target="notifier-privacy" />.</t>
<t>If the notifier owner is interactively queried to determine
whether a subscription is allowed, a "202 Accept" response is
returned immediately. Note that a NOTIFY message is still formed
and sent under these circumstances, as described in the previous
section.</t>
<t>If subscription authorization was delayed and the notifier
wishes to convey that such authorization has been declined, it may
do so by sending a NOTIFY message containing a
"Subscription-State" header field with a value of "terminated" and a
reason parameter of "rejected".</t>
</section>
<section anchor="notifier-refresh" title="Refreshing of Subscriptions">
<t>When a notifier receives a subscription refresh, assuming that
the subscriber is still authorized, the notifier updates the
expiration time for subscription. As with the initial
subscription, the server MAY shorten the amount of time until
expiration, but MUST NOT increase it. The final expiration time is
placed in the "Expires" header field in the response. If the duration
specified in a SUBSCRIBE message is unacceptably short, the
notifier SHOULD respond with a "423 Interval Too Brief"
message.</t>
<t>If no refresh for a notification address is received before its
expiration time, the subscription is removed. When removing a
subscription, the notifier SHOULD send a NOTIFY message with a
"Subscription-State" value of "terminated" to inform it that the
subscription is being removed. If such a message is sent, the
"Subscription-State" header field SHOULD contain a "reason=timeout"
parameter.</t>
<t>
Clients can cause a subscription to be terminated immediately
by sending a SUBSCRIBE with an "Expires" header field
set to '0'. Notifiers largely treat this the same way
as any other subscription expiration: they send a NOTIFY message
containing a "Subscription-State" of "terminated", with a reason code
of "timeout." For consistency with state polling (see
<xref target="polling-state"/>) and subscription refreshes,
the notifier may choose to
include resource state in this final NOTIFY. However, in some
cases, including such state makes no sense. Under such circumstances,
the notifier may choose to omit state information from the
terminal NOTIFY message.
</t>
<t><list style="hanging">
<t>The sending of a NOTIFY when a subscription expires allows
the corresponding dialog to be terminated, if appropriate.</t>
</list></t>
</section>
</section>
<section anchor="notifier-notifies"
title="Sending State Information to Subscribers">
<t>
Notifiers use the NOTIFY method to send information about
the state of a resource to subscribers. The notifier's view
of a subscription is shown in the following state diagram:
</t>
<figure>
<artwork><![CDATA[
+-------------+
| init |
+-------------+
|
Receive SUBSCRIBE,
Send NOTIFY
|
V NOTIFY failure,
+-------------+ subscription expires,
+------------| resp_wait |-- or terminated ----+
| +-------------+ per local policy |
| | |
| | |
| | V
Policy grants Policy needed +-------------+
permission | | terminated |
| | +-------------+
| | A A
| V NOTIFY failure, | |
| +-------------+ subscription expires,| |
| | pending |-- or terminated -------+ |
| +-------------+ per local policy |
| | |
| Policy changed to |
| grant permission |
| | |
| V NOTIFY failure, |
| +-------------+ subscription expires, |
+----------->| active |-- or terminated ---------+
+-------------+ per local policy
]]></artwork>
</figure>
<t>When a SUBSCRIBE request is answered with a 200-class response,
the notifier MUST immediately construct and send a NOTIFY request to
the subscriber. When a change in the subscribed state occurs, the
notifier SHOULD immediately construct and send a NOTIFY request,
subject to authorization, local policy, and throttling
considerations.</t>
<t>If the NOTIFY request fails due to expiration of SIP Timer F
(transaction timeout), the notifier SHOULD remove the subscription.</t>
<t><list style="hanging">
<t>This behavior prevents unnecessary transmission of state
information for subscribers who have crashed or disappeared from
the network. Because such transmissions will be sent multiple
times, per the retransmission algorithm defined in SIP <xref
target="RFC3261" /> (instead of the typical single transmission for
functioning clients), continuing to service them when no client is
available to acknowledge them could place undue strain on a
network. Upon client restart or reestablishment of a network
connection, it is expected that clients will send SUBSCRIBE
messages to refresh potentially stale state information; such
messages will re-install subscriptions in all relevant nodes.</t>
</list></t>
<t>If the NOTIFY transaction fails due to the receipt of
a 404, 405, 410, 416, 480-485, 489, 501, or 604 response to the NOTIFY,
the notifier MUST remove the corresponding subscription.
See
<xref target="RFC5057"/> for further details and notes about the effect
of error codes on dialogs and usages within dialog (such as subscriptions).
</t>
<t><list style="hanging">
<t>A notify error response would generally indicate that something
has gone wrong with the subscriber or with some proxy on the way
to the subscriber. If the subscriber is in error, it makes the
most sense to allow the subscriber to rectify the situation (by
re-subscribing) once the error condition has been handled. If a
proxy is in error, the periodic SUBSCRIBE refreshes will
re-install subscription state once the network problem has been
resolved.</t>
</list></t>
<t>NOTIFY requests MUST contain a "Subscription-State" header field with a
value of "active", "pending", or "terminated". The "active" value
indicates that the subscription has been accepted and has been
authorized (in most cases; see <xref target="notifier-privacy" />). The
"pending" value indicates that the subscription has been received,
but that policy information is insufficient to accept or deny the
subscription at this time. The "terminated" value indicates that the
subscription is not active.</t>
<t>If the value of the "Subscription-State" header field is "active" or
"pending", the notifier MUST also include in the
"Subscription-State" header field an "expires" parameter which indicates
the time remaining on the subscription. The notifier MAY use this
mechanism to shorten a subscription; however, this mechanism MUST
NOT be used to lengthen a subscription.</t>
<t><list style="hanging">
<t>Including expiration information for active and pending
subscriptions is necessary in case the SUBSCRIBE request forks, since
the response to a forked SUBSCRIBE may not be received by the
subscriber. <xref target="RFC3265"/> allowed the notifier some discretion in
the inclusion of this parameter, so subscriber implementations
are warned to handle the lack of an "expires" parameter gracefully.
Note well that this "expires" value is a parameter on
the "Subscription-State" header field, NOT an "Expires" header field.</t>
</list></t>
<t><list style="hanging"><t>
The period of time for a subscription can
be shortened to zero by the notifier. In other words,
it is perfectly valid for a SUBSCRIBE with a non-zero expires
to be answered with a NOTIFY that contains "Subscription-Status:
terminated;reason=expired". This merely means that the notifier
has shortened the subscription timeout to zero, and the subscription
has expired instantaneously. The body may contain valid state, or
it may contain a neutral state (see <xref target="template-notifier-notifies"/>).
</t></list></t>
<t>If the value of the "Subscription-State" header field is "terminated",
the notifier SHOULD also include a "reason" parameter. The notifier
MAY also include a "retry-after" parameter, where appropriate. For
details on the value and semantics of the "reason" and "retry-after"
parameters, see <xref target="subscriber-notifies" />.</t>
</section>
<section title="PINT Compatibility">
<t>The "Event" header field is considered mandatory for the purposes of
this document. However, to maintain compatibility with PINT (see
<xref target="RFC2848"></xref>), notifiers MAY interpret a SUBSCRIBE
request with no "Event" header field as requesting a subscription to PINT
events. If a notifier does not support PINT, it SHOULD return "489 Bad
Event" to any SUBSCRIBE messages without an "Event" header field.</t>
</section>
</section>
<section anchor="proxy" title="Proxy Behavior">
<t>Proxies need no additional behavior beyond that described in SIP
<xref target="RFC3261" /> to support SUBSCRIBE and NOTIFY. If a proxy wishes to
see all of the SUBSCRIBE and NOTIFY requests for a given dialog, it
MUST add a Record-Route header field to the initial SUBSCRIBE request
and all NOTIFY requests.
It MAY choose to include Record-Route in subsequent SUBSCRIBE messages;
however, these requests cannot cause the dialog's route set to
be modified.</t>
<t><list style="hanging">
<t>Note that subscribers and notifiers may elect to use S/MIME
encryption of SUBSCRIBE and NOTIFY requests; consequently, proxies
cannot rely on being able to access any information that is not
explicitly required to be proxy-readable by SIP <xref
target="RFC3261" />.</t>
</list></t>
</section>
<section title="Common Behavior">
<section anchor="dialog-handling" title="Dialog Creation and Termination">
<t>
Dialogs are created upon completion of a NOTIFY transaction for a new
subscription, unless the NOTIFY contains a "Subscription-State"
of "terminated."
</t>
<t>
NOTIFY requests are matched to such SUBSCRIBE requests if they
contain the same "Call-ID", a "To" header field "tag" parameter which
matches the "From" header field "tag" parameter of the SUBSCRIBE, and the
same "Event" header field. Rules for comparisons of the "Event"
header fields are described in <xref target="event-header field" />.
</t>
<t>A subscription is destroyed after a notifier sends a NOTIFY
request with a "Subscription-State" of "terminated." The subscriber
will generally answer such final requests with a "200 OK" response (unless
a condition warranting an alternate response has arisen). Except when
the mechanism described in <xref target="dialog-reuse"/> is used,
the destruction of a subscription results in the termination of
its associated dialog.</t>
<t><list style="hanging">
<t>A subscriber may send a SUBSCRIBE request with an "Expires"
header field of 0 in order to trigger the sending of such a NOTIFY
request; however, for the purposes of subscription and dialog
lifetime, the subscription is not considered terminated until the
NOTIFY transaction with a "Subscription-State" of "terminated"
completes.</t>
</list></t>
</section>
<section anchor="migration" title="Notifier Migration">
<t> It is often useful to allow migration of subscriptions between
notifiers. Such migration may be effected
by sending a NOTIFY message with a "Subscription-State" header field of
"terminated", and a reason parameter of "deactivated". This NOTIFY
request is otherwise normal, and is formed as described in <xref
target="notifier-notifies"/>.</t>
<t>Upon receipt of this NOTIFY message, the subscriber SHOULD
attempt to re-subscribe (as described in the preceding sections).
Note that this subscription is established on a new dialog, and does
not re-use the route set from the previous subscription dialog.</t>
<t>The actual migration is effected by making a change to the policy
(such as routing decisions) of one or more servers to which the
SUBSCRIBE request will be sent in such a way that a different node
ends up responding to the SUBSCRIBE request. This may be as simple
as a change in the local policy in the notifier from which the
subscription is migrating so that it serves as a proxy or redirect
server instead of a notifier.</t>
<t>Whether, when, and why to perform notifier migrations may be
described in individual event packages; otherwise, such decisions
are a matter of local notifier policy, and are left up to individual
implementations.</t>
</section>
<section anchor="polling-state" title="Polling Resource State">
<t>A natural consequence of the behavior described in the preceding
sections is that an immediate fetch without a persistent
subscription may be effected by sending a SUBSCRIBE with an
"Expires" of 0.</t>
<t>Of course, an immediate fetch while a subscription is active may
be effected by sending a SUBSCRIBE with an "Expires" equal to the
number of seconds remaining in the subscription.</t>
<t>Upon receipt of this SUBSCRIBE request, the notifier (or
notifiers, if the SUBSCRIBE request was forked) will send a NOTIFY
request containing resource state in the same dialog.</t>
<t>Note that the NOTIFY messages triggered by SUBSCRIBE messages
with "Expires" header fields of 0 will contain a "Subscription-State"
value of "terminated", and a "reason" parameter of "timeout".</t>
<t>Polling of event state can cause significant increases in load on
the network and notifiers; as such, it should be used only
sparingly. In particular, polling SHOULD NOT be used in
circumstances in which it will typically result in more network
messages than long-running subscriptions.</t>
<t>When polling is used, subscribers SHOULD attempt to cache
authentication credentials between polls so as to reduce the number
of messages sent.</t>
</section>
<section anchor="allow-events" title="Allow-Events header field usage">
<t>The "Allow-Events" header field, if present, includes a list of tokens
which indicates the event packages supported by a notifier.
In other words, a
user agent sending an "Allow-Events" header field is advertising that it can
process SUBSCRIBE requests and generate NOTIFY requests for all of
the event packages listed in that header field.</t>
<t>Any user agent that can act as a notifier for one or more event packages
SHOULD include
an appropriate "Allow-Events" header field indicating all supported events
in all methods which initiate dialogs and their responses (such as
INVITE) and OPTIONS responses.</t>
<t><list style="hanging">
<t>This information is very useful, for example, in allowing user
agents to render particular interface elements appropriately
according to whether the events required to implement the features
they represent are supported by the appropriate nodes.</t>
<t>
On the other hand, it doesn't necessarily make much sense
to indicate supported events inside a NOTIFY-established
dialog if the only event package supported is the
one associated with that subscription.
</t>
</list></t>
<t>Note that "Allow-Events" header fields MUST NOT be inserted by
proxies.</t>
</section>
</section>
<section anchor="device-targeting" title="Targeting Subscriptions at Devices">
<t>
<xref target="RFC3265"/> defined a mechanism by which subscriptions could share
dialogs with invite usages and with other subscriptions. The
purpose of this behavior was to allow subscribers to ensure
that a subscription arrived at the same device as an established
dialog. Unfortunately, the re-use of dialogs has proven to be
exceedingly confusing. <xref target="RFC5057"/> attempted to
clarify proper behavior in a variety of circumstances; however,
the ensuing rules remain confusing and prone to implementation
error. At the same time, the mechanism described in
<xref target="I-D.ietf-sip-gruu"/> now provides a far more elegant
and unambiguous means to achieve the same goal.
</t>
<t>
Consequently, the dialog re-use technique described in RFC 3265 is now deprecated.
</t>
<section anchor="gruu" title="Using GRUUs to Route to Devices">
<t>
Notifiers MUST implement the GRUU extension defined in
<xref target="I-D.ietf-sip-gruu"/>, and MUST use a GRUU
as their local target. This allows
subscribers to explicitly target desired devices.
</t>
<t>
If a subscriber wishes to subscribe to a resource on the same device
as an established dialog, it should check whether the remote contact
in that dialog is a GRUU (i.e., whether it contains a "gr" URI
parameter). If so, the subscriber creates a new dialog, using
the GRUU as the request URI for the new SUBSCRIBE.
</t>
<t><list style="hanging"><t>
Because GRUUs are guaranteed to route to a a specific device,
this ensures that the subscription will be routed to
the same place as the established dialog.
</t></list></t>
</section>
<section anchor="dialog-reuse" title="Sharing Dialogs">
<t>
For compatibility with older clients, subscriber and notifier
implementations may choose to allow dialog sharing. The behavior
of multiple usages within a dialog are described in
<xref target="RFC5057"/>.
</t>
<t>
Subscribers MUST NOT attempt to re-use dialogs whose remote
target is a GRUU.
</t>
<t><list style="hanging"><t>
Note that the techniques described in this section are
included for backwards compatibility purposes only.
Because subscribers cannot re-use dialogs with a GRUU
for their remote target, and because notifiers must
use GRUUs as their local target, any two implementations
that conform to this specification will automatically
use the mechanism described in <xref target="gruu"/>.
</t></list></t>
<t>
If a subscriber wishes to subscribe to a resource on the same
device as an established dialog and the remote contact is not
a GRUU, it MAY revert to dialog sharing behavior.
Alternately, it MAY choose to
treat the remote party as incapable of servicing the subscription
(i.e., the same way it would behave if the remote party did
not support SIP events at all).
</t>
<t>
If a notifier receives a SUBSCRIBE request for a new subscription
on an existing dialog, it MAY choose to implement dialog sharing
behavior. Alternately, it may choose to fail the SUBSCRIBE
request with a 403 response. The error text of such 403
responses SHOULD indicate that dialog sharing is not supported.
</t>
<t>
To implement dialog sharing, subscribers and notifiers perform
the following additional processing:
</t>
<t><list style="symbols">
<t>
When subscriptions exist in dialogs associated with
INVITE-created application state and/or other subscriptions,
these sets of application state do not interact beyond the
behavior described for a dialog (e.g., route set handling).
In particular, multiple subscriptions within a dialog are expire
independently, and require independent SUBSCRIBE refreshes.<vspace blankLines="1"/></t>
<t>If a subscription's destruction leaves no other application state
associated with the dialog, the dialog terminates. The destruction
of other application state (such as that created by an INVITE) will
not terminate the dialog if a subscription is still associated with
that dialog. This means that, when dialog are re-used, then a dialog
created with an INVITE does not necessarily terminate upon receipt
of a BYE. Similarly, in the case that several subscriptions are
associated with a single dialog, the dialog does not terminate
until all the subscriptions in it are destroyed.<vspace blankLines="1"/></t>
<t>Subscribers MAY include an "id" parameter in SUBSCRIBE
request "Event" header field to allow differentiation between multiple
subscriptions in the same dialog. This "id" parameter, if present,
contains an opaque token which identifies the specific subscription
within a dialog. An "id" parameter is only valid within the scope of
a single dialog.<vspace blankLines="1"/></t>
<t> If an "id" parameter is present in the SUBSCRIBE message used
to establish a subscription, that "id" parameter MUST also be present
in all corresponding NOTIFY messages.<vspace blankLines="1"/></t>
<t>When a subscriber refreshes a the subscription timer,
the SUBSCRIBE MUST contain the same "Event" header field "id" parameter
as was present in the initial subscription. (Otherwise, the notifier
will interpret the SUBSCRIBE message as a request for a new
subscription in the same dialog).<vspace blankLines="1"/></t>
<t>When a subscription is created in the notifier, it stores the
any "Event" header field "id" parameter as part of the subscription
information (along with the event package name).<vspace blankLines="1"/></t>
<t>If an initial SUBSCRIBE is sent on a pre-existing dialog, a
matching NOTIFY request merely creates a new subscription associated
with that dialog.<vspace blankLines="1"/></t>
</list></t>
</section>
</section>
<section anchor="cancel" title="CANCEL Requests for SUBSCRIBE and NOTIFY">
<t>Neither SUBSCRIBE nor NOTIFY messages can be canceled. If a
UAS receives a CANCEL request that matches a known SUBSCRIBE
or NOTIFY transaction, it MUST respond to the CANCEL request,
but otherwise ignore it. In particular, the CANCEL request
MUST NOT affect processing of the SUBSCRIBE or NOTIFY request
in any way.</t>
<t>
UACs SHOULD NOT send CANCEL requests for SUBSCRIBE or NOTIFY
transactions.
</t>
</section>
</section>
<section anchor="event-packages" title="Event Packages">
<t>This section covers several issues which should be taken into
consideration when event packages based on SUBSCRIBE and NOTIFY are
proposed. Event package definitions contain sections
addressing each of these issues, ideally in the same order
and with the same titles as the following sections.</t>
<section title="Appropriateness of Usage">
<t>When designing an event package using the methods described in this
document for event notification, it is important to consider: is SIP an
appropriate mechanism for the problem set? Is SIP being selected
because of some unique feature provided by the protocol (e.g., user
mobility), or merely because "it can be done?" If you find yourself
defining event packages for notifications related to, for example,
network management or the temperature inside your car's engine, you
may want to reconsider your selection of protocols.</t>
<t><list style="hanging">
<t>Those interested in extending the mechanism defined in this
document are urged to follow the development of "Guidelines for
Authors of SIP Extensions" <xref target="RFC4485" /> for further
guidance regarding appropriate uses of SIP.</t>
</list></t>
<t>Further, it is expected that this mechanism is not to be used in
applications where the frequency of reportable events is excessively
rapid (e.g., more than about once per second). A SIP network is
generally going to be provisioned for a reasonable signaling volume;
sending a notification every time a user's GPS position changes by one
hundredth of a second could easily overload such a network.</t>
</section>
<section anchor="templates" title="Event Template-packages">
<t>Normal event packages define a set of state applied to a specific
type of resource, such as user presence, call state, and messaging
mailbox state.</t>
<t>Event template-packages are a special type of package which define
a set of state applied to other packages, such as statistics, access
policy, and subscriber lists. Event template-packages may even be
applied to other event template-packages.</t>
<t>To extend the object-oriented analogy made earlier, event
template-packages can be thought of as templatized C++ packages which
must be applied to other packages to be useful.</t>
<t>The name of an event template-package as applied to a package is
formed by appending a period followed by the event template-package
name to the end of the package. For example, if a template-package
called "winfo" were being applied to a package called "presence", the
event token used in "Event" and "Allow-Events" would be
"presence.winfo".</t>
<t>Event template-packages must be defined so that they can be applied
to any arbitrary package. In other words, event template-packages
cannot be specifically tied to one or a few "parent" packages in such
a way that they will not work with other packages.</t>
</section>
<section anchor="amount-of-state" title="Amount of State to be Conveyed">
<t>When designing event packages, it is important to consider the type
of information which will be conveyed during a notification.</t>
<t>A natural temptation is to convey merely the event (e.g., "a new
voice message just arrived") without accompanying state (e.g., "7 total
voice messages"). This complicates implementation of subscribing
entities (since they have to maintain complete state for the entity to
which they have subscribed), and also is particularly susceptible to
synchronization problems.</t>
<t>There are two possible solutions to this problem that event
packages may choose to implement.</t>
<section title="Complete State Information">
<t>For packages which typically convey state information that is
reasonably small (on the order of 1 KB or so), it is suggested that
event packages are designed so as to send complete state information
when an event occurs.</t>
<t>In some circumstances, conveying the current state alone may be
insufficient for a particular class of events. In these cases, the
event packages should include complete state information along with
the event that occurred. For example, conveying "no customer service
representatives available" may not be as useful as conveying "no
customer service representatives available; representative
sip:46@cs.xyz.int just logged off".</t>
</section>
<section title="State Deltas">
<t>In the case that the state information to be conveyed is large,
the event package may choose to detail a scheme by which NOTIFY
messages contain state deltas instead of complete state.</t>
<t>Such a scheme would work as follows: any NOTIFY sent in immediate
response to a SUBSCRIBE contains full state information. NOTIFY
messages sent because of a state change will contain only the state
information that has changed; the subscriber will then merge this
information into its current knowledge about the state of the
resource.</t>
<t>Any event package that supports delta changes to states MUST
include a version number that increases by exactly one for each
NOTIFY transaction in a subscription. Note that the state version number
appears in the body of the message, not in a SIP header field.</t>
<t>If a NOTIFY arrives that has a version number that is incremented
by more than one, the subscriber knows that a state delta has been
missed; it ignores the NOTIFY message containing the state delta
(except for the version number, which it retains to detect message
loss), and re-sends a SUBSCRIBE to force a NOTIFY containing a
complete state snapshot.</t>
</section>
</section>
<section anchor="packages" title="Event Package Responsibilities">
<t>Event packages are not required to reiterate any of the behavior
described in this document, although they may choose to do so for
clarity or emphasis. In general, though, such packages are expected to
describe only the behavior that extends or modifies the behavior
described in this document.</t>
<t>Note that any behavior designated with "SHOULD" or "MUST" in this
document is not allowed to be weakened by extension documents;
however, such documents may elect to strengthen "SHOULD" requirements
to "MUST" strength if required by their application.</t>
<t><list style="hanging">
<t>In addition to the normal sections expected in standards-track
RFCs and SIP extension documents, authors of event packages need to
address each of the issues detailed in the following subsections,
whenever applicable.</t>
</list></t>
<section title="Event Package Name">
<t>This section, which MUST be present, defines the token name to be
used to designate the event package. It MUST include the information
which appears in the IANA registration of the token. For information
on registering such types, see <xref target="iana-considerations" />.</t>
</section>
<section title="Event Package Parameters">
<t>If parameters are to be used on the "Event" header field to modify the
behavior of the event package, the syntax and semantics of such
header fields MUST be clearly defined.</t>
</section>
<section title="SUBSCRIBE Bodies">
<t>It is expected that most, but not all, event packages will define
syntax and semantics for SUBSCRIBE method bodies; these bodies will
typically modify, expand, filter, throttle, and/or set thresholds
for the class of events being requested. Designers of event packages
are strongly encouraged to re-use existing MIME types for message
bodies where practical.</t>
<t>This mandatory section of an event package defines what type or
types of event bodies are expected in SUBSCRIBE requests (or specify
that no event bodies are expected). It should point to detailed
definitions of syntax and semantics for all referenced body
types.</t>
</section>
<section title="Subscription Duration">
<t>It is RECOMMENDED that event packages give a suggested range of
times considered reasonable for the duration of a subscription. Such
packages MUST also define a default "Expires" value to be used if
none is specified.</t>
</section>
<section title="NOTIFY Bodies">
<t>The NOTIFY body is used to report state on the resource being
monitored. Each package MUST define what type or types of event
bodies are expected in NOTIFY requests. Such packages MUST specify
or cite detailed specifications for the syntax and semantics
associated with such event body.</t>
<t>Event packages also MUST define which MIME type is to be assumed
if none are specified in the "Accept" header field of the SUBSCRIBE
request.</t>
</section>
<section title="Notifier processing of SUBSCRIBE requests">
<t>This section describes the processing to be performed by the
notifier upon receipt of a SUBSCRIBE request. Such a section is
required.</t>
<t>Information in this section includes details of how to
authenticate subscribers and authorization issues for the package.
Such authorization issues may include, for example, whether all
SUBSCRIBE requests for this package are answered with 202 responses
(see <xref target="notifier-privacy" />).</t>
</section>
<section anchor="template-notifier-notifies" title="Notifier generation of NOTIFY requests">
<t>This section of an event package describes the process by which
the notifier generates and sends a NOTIFY request. This includes
detailed information about what events cause a NOTIFY to be sent,
how to compute the state information in the NOTIFY, how to generate
neutral or fake state information to hide authorization delays and
decisions from users, and whether state information is complete or
deltas for notifications; see <xref target="amount-of-state" />. Such a section
is required.</t>
<t>This section may optionally describe the behavior used to process
the subsequent response.</t>
</section>
<section title="Subscriber processing of NOTIFY requests">
<t>This section of an event package describes the process followed
by the subscriber upon receipt of a NOTIFY request, including any
logic required to form a coherent resource state (if
applicable).</t>
</section>
<section anchor="forking" title="Handling of forked requests">
<t>Each event package MUST specify whether forked SUBSCRIBE requests
are allowed to install multiple subscriptions.</t>
<t>If such behavior is not allowed, the first potential
dialog-establishing message will create a dialog. All subsequent
NOTIFY messages which correspond to the SUBSCRIBE message (i.e.,
match "To", "From", "From" header field "tag" parameter, "Call-ID",
"Event", and "Event" header field "id" parameter) but which do not
match the dialog would be rejected with a 481 response. Note that
the 200-class response to the SUBSCRIBE can arrive after a matching
NOTIFY has been received; such responses might not correlate to the
same dialog established by the NOTIFY. Except as required to
complete the SUBSCRIBE transaction, such non-matching 200-class
responses are ignored.</t>
<t>If installing of multiple subscriptions by way of a single forked
SUBSCRIBE is allowed, the subscriber establishes a new dialog
towards each notifier by returning a 200-class response to each
NOTIFY. Each dialog is then handled as its own entity, and is
refreshed independent of the other dialogs.</t>
<t>In the case that multiple subscriptions are allowed, the event
package MUST specify whether merging of the notifications to form a
single state is required, and how such merging is to be performed.
Note that it is possible that some event packages may be defined in
such a way that each dialog is tied to a mutually exclusive state
which is unaffected by the other dialogs; this MUST be clearly
stated if it is the case.</t>
</section>
<section title="Rate of notifications">
<t>Each event package is expected to define a requirement (SHOULD or
MUST strength) which defines an absolute maximum on the rate at
which notifications are allowed to be generated by a single
notifier.</t>
<t>Each package MAY further define a throttle mechanism which allows
subscribers to further limit the rate of notification.</t>
</section>
<section anchor="aggregation" title="State Aggregation">
<t>
Many event packages inherently work by collecting information
about a resource from a number of other sources -- either
through the use of PUBLISH <xref target="RFC3903"/>, by
subscribing to state information, or through other state
gathering mechanisms.
</t>
<t>Event packages that involve retrieval of state information
for a single resource from more than one source need to
consider how notifiers aggregate information into
a single, coherent state. Such packages MUST
specify how notifiers aggregate information and how they
provide authentication and authorization.</t>
</section>
<section title="Examples">
<t>Event packages SHOULD include several demonstrative message flow
diagrams paired with several typical, syntactically correct, and
complete messages.</t>
<t>It is RECOMMENDED that documents describing event packages
clearly indicate that such examples are informative and not
normative, with instructions that implementors refer to the main
text of the document for exact protocol details.</t>
</section>
<section title="Use of URIs to Retrieve State" anchor="uri-use">
<t>Some types of event packages may define state information which
is potentially too large to reasonably send in a SIP message. To
alleviate this problem, event packages may include the ability to
convey a URI instead of state information; this URI will then be
used to retrieve the actual state information.</t>
<t><xref target="RFC4483"/> defines a mechanism that can be used
by event packages to convey information in such a fashion.</t>
</section>
</section>
</section>
<section title="Security Considerations">
<section title="Access Control">
<t>The ability to accept subscriptions should be under the direct
control of the notifier's user, since many types of events may be
considered sensitive for the purposes of privacy. Similarly, the
notifier should have the ability to selectively reject subscriptions
based on the subscriber identity (based on access control lists),
using standard SIP authentication mechanisms. The methods for creation
and distribution of such access control lists is outside the scope of
this document.</t>
</section>
<section anchor="notifier-privacy" title="Notifier Privacy Mechanism">
<t>The mere act of returning a 200 or certain 4xx and 6xx responses to
SUBSCRIBE requests may, under certain circumstances, create privacy
concerns by revealing sensitive policy information. In these cases,
the notifier SHOULD always return a 202 response. While the subsequent
NOTIFY message may not convey true state, it MUST appear to contain a
potentially correct piece of data from the point of view of the
subscriber, indistinguishable from a valid response. Information about
whether a user is authorized to subscribe to the requested state is
never conveyed back to the original user under these
circumstances.</t>
<t>Individual packages and their related documents for which such a mode
of operation makes sense can further describe how and why to generate
such potentially correct data. For example, such a mode of operation
is mandated by <xref target="RFC2779"></xref> for user presence
information.</t>
</section>
<section title="Denial-of-Service attacks">
<t>The current model (one SUBSCRIBE request triggers a SUBSCRIBE
response and one or more NOTIFY requests) is a classic setup for an
amplifier node to be used in a smurf attack.</t>
<t>Also, the creation of state upon receipt of a SUBSCRIBE request can
be used by attackers to consume resources on a victim's machine,
rendering it unusable.</t>
<t>To reduce the chances of such an attack, implementations of
notifiers SHOULD require authentication. Authentication issues are
discussed in SIP <xref target="RFC3261"></xref>.</t>
</section>
<section title="Replay Attacks">
<t>Replaying of either SUBSCRIBE or NOTIFY can have detrimental
effects.</t>
<t>In the case of SUBSCRIBE messages, attackers may be able to install
any arbitrary subscription which it witnessed being installed at some
point in the past. Replaying of NOTIFY messages may be used to spoof
old state information (although a good versioning mechanism in the
body of the NOTIFY messages may help mitigate such an attack). Note
that the prohibition on sending NOTIFY messages to nodes which have
not subscribed to an event also aids in mitigating the effects of such
an attack.</t>
<t>To prevent such attacks, implementations SHOULD require
authentication with anti-replay protection. Authentication issues are
discussed in SIP <xref target="RFC3261"></xref>.</t>
</section>
<section title="Man-in-the middle attacks">
<t>Even with authentication, man-in-the-middle attacks using SUBSCRIBE
may be used to install arbitrary subscriptions, hijack existing
subscriptions, terminate outstanding subscriptions, or modify the
resource to which a subscription is being made. To prevent such
attacks, implementations SHOULD provide integrity protection across
"Contact", "Route", "Expires", "Event", and "To" header fields of SUBSCRIBE
messages, at a minimum. If SUBSCRIBE bodies are used to define further
information about the state of the call, they SHOULD be included in
the integrity protection scheme.</t>
<t>Man-in-the-middle attacks may also attempt to use NOTIFY messages
to spoof arbitrary state information and/or terminate outstanding
subscriptions. To prevent such attacks, implementations SHOULD provide
integrity protection across the "Call-ID", "CSeq", and
"Subscription-State" header fields and the bodies of NOTIFY messages.</t>
<t>Integrity protection of message header fields and bodies is discussed in
SIP <xref target="RFC3261"></xref>.</t>
</section>
<section title="Confidentiality">
<t>The state information contained in a NOTIFY message has the
potential to contain sensitive information. Implementations MAY
encrypt such information to ensure confidentiality.</t>
<t>While less likely, it is also possible that the information
contained in a SUBSCRIBE message contains information that users might
not want to have revealed. Implementations MAY encrypt such
information to ensure confidentiality.</t>
<t>To allow the remote party to hide information it considers
sensitive, all implementations SHOULD be able to handle encrypted
SUBSCRIBE and NOTIFY messages.</t>
<t>The mechanisms for providing confidentiality are detailed in SIP
<xref target="RFC3261"></xref>.</t>
</section>
</section>
<section anchor="iana-considerations" title="IANA Considerations">
<t>(This section is not applicable until this document is published as
an RFC.)</t>
<!-- ====================================================================== -->
<section title="Event Packages">
<t>This document defines an event-type namespace which requires a
central coordinating body. The body chosen for this coordination is the
Internet Assigned Numbers Authority (IANA).</t>
<t>There are two different types of event-types: normal event packages,
and event template-packages; see <xref target="templates"></xref>. To avoid
confusion, template-package names and package names share the same
namespace; in other words, an event template-package MUST NOT share a
name with a package.</t>
<t>Following the policies outlined in "Guidelines for Writing an IANA
Considerations Section in RFCs" <xref target="RFC2434"></xref>, normal
event package identification tokens are allocated as First Come First
Served, and event template-package identification tokens are allocated
on a IETF Consensus basis.</t>
<t>Registrations with the IANA MUST include the token being registered
and whether the token is a package or a template-package. Further,
packages MUST include contact information for the party responsible for
the registration and/or a published document which describes the event
package. Event template-package token registrations MUST include a
pointer to the published RFC which defines the event
template-package.</t>
<t>Registered tokens to designate packages and template-packages MUST
NOT contain the character ".", which is used to separate
template-packages from packages.</t>
<section title="Registration Information">
<t>As this document specifies no package or template-package names,
the initial IANA registration for event types will be empty. The
remainder of the text in this section gives an example of the type of
information to be maintained by the IANA; it also demonstrates all
five possible permutations of package type, contact, and
reference.</t>
<t>The table below lists the event packages and template-packages
defined in "SIP-Specific Event Notification" [RFC xxxx]. Each name is
designated as a package or a template-package under "Type".</t>
<figure>
<artwork><![CDATA[
Package Name Type Contact Reference
------------ ---- ------- ---------
example1 package [Roach]
example2 package [Roach] [RFC xxxx]
example3 package [RFC xxxx]
example4 template [Roach] [RFC xxxx]
example5 template [RFC xxxx]
PEOPLE
------
[Roach] Adam Roach <adam.roach@tekelec.com>
REFERENCES
----------
[RFC xxxx] A.B. Roach, "SIP-Specific Event Notification", RFC XXXX,
Monthname 20XX
]]></artwork>
</figure>
</section>
<section title="Registration Template">
<figure>
<artwork><![CDATA[
To: ietf-sip-events@iana.org
Subject: Registration of new SIP event package
Package Name:
]]></artwork>
</figure>
<t><list style="hanging">
<t>(Package names must conform to the syntax described in <xref
target="event-header field" />.)</t>
</list></t>
<figure>
<artwork><![CDATA[
Is this registration for a Template Package:
]]></artwork>
</figure>
<t><list style="hanging">
<t>(indicate yes or no)</t>
</list></t>
<figure>
<artwork><![CDATA[
Published Specification(s):
]]></artwork>
</figure>
<t><list style="hanging">
<t>(Template packages require a published RFC. Other packages may
reference a specification when appropriate).</t>
</list></t>
<figure>
<artwork><![CDATA[
Person & email address to contact for further information:
]]></artwork>
</figure>
</section>
</section>
<!-- ====================================================================== -->
<section anchor="reason-codes" title="Reason Codes">
<t>
This document further defines "reason" codes for use in the "Subscription-State"
header field (see <xref target="subscriber-notifies"/>).
</t>
<t>
Following the policies outlined in "Guidelines for Writing an IANA
Considerations Section in RFCs" <xref target="RFC2434"></xref>, new
reason codes require a Standards Action.
</t>
<t>
Registrations with the IANA include the reason code being registered
and a reference to a published document which describes the event
package. Insertion of such values takes place as part of the RFC
publication process or as the result of inter-SDO liaison activity.
New reason codes must conform to the syntax of the ABNF "token"
element defined in SIP <xref target="RFC3261"/>.
</t>
<t>
<xref target="RFC4660"/> defined a new reason code prior to the establishment
of an IANA registry. We include its reason code ("badfilter") in the
initial list of reason codes to ensure a complete registry.
</t>
<t>
The IANA registry for reason code will be initialized with the following
values:
</t>
<figure>
<artwork><![CDATA[
Reason Code Reference
----------- ---------
deactivated [RFC xxxx]
probation [RFC xxxx]
rejected [RFC xxxx]
timeout [RFC xxxx]
giveup [RFC xxxx]
noresource [RFC xxxx]
invariant [RFC xxxx]
badfilter [RFC 4660]
REFERENCES
----------
[RFC xxxx] A.B. Roach, "SIP-Specific Event Notification", RFC XXXX,
Monthname 20XX
[RFC 4660] Khartabil, H., Leppanen, E., Lonnfors, M., and
J. Costa-Requena, "Functional Description of Event
Notification Filtering", September 2006.
]]></artwork></figure>
</section>
<!-- ====================================================================== -->
<section title="Header Field Names">
<t>This document registers three new header field names, described
elsewhere in this document. These header fields are defined by the following
information, which is to be added to the header field sub-registry under
http://www.iana.org/assignments/sip-parameters.</t>
<figure>
<artwork><![CDATA[
Header Name: Allow-Events
Compact Form: u
Header Name: Subscription-State
Compact Form: (none)
Header Name: Event
Compact Form: o
]]></artwork>
</figure>
</section>
<section title="Response Codes">
<t>This document registers two new response codes.
These response codes are defined
by the following information, which is to be added to the method and
response-code sub-registry under
http://www.iana.org/assignments/sip-parameters.</t>
<figure>
<artwork><![CDATA[
Response Code Number: 202
Default Reason Phrase: Accepted
Response Code Number: 489
Default Reason Phrase: Bad Event
]]></artwork>
</figure>
</section>
</section>
<section title="Syntax">
<t>This section describes the syntax extensions required for event
notification in SIP. Semantics are described in <xref
target="node-behavior"></xref>. Note that the formal syntax definitions described
in this document are expressed in the ABNF format used in SIP <xref
target="RFC3261"></xref>, and contain references to elements defined
therein.</t>
<section anchor="new-methods" title="New Methods">
<t>This document describes two new SIP methods: SUBSCRIBE and
NOTIFY.</t>
<t>This table expands on tables 2 and 3 in SIP <xref
target="RFC3261"></xref>.</t>
<figure>
<artwork><![CDATA[
Header Where SUB NOT
------ ----- --- ---
Accept R o o
Accept 2xx - -
Accept 415 o o
Accept-Encoding R o o
Accept-Encoding 2xx - -
Accept-Encoding 415 o o
Accept-Language R o o
Accept-Language 2xx - -
Accept-Language 415 o o
Alert-Info R - -
Alert-Info 180 - -
Allow R o o
Allow 2xx o o
Allow r o o
Allow 405 m m
Authentication-Info 2xx o o
Authorization R o o
Call-ID c m m
Contact R m m
Contact 1xx o o
Contact 2xx m o
Contact 3xx m m
Contact 485 o o
Content-Disposition o o
Content-Encoding o o
Content-Language o o
Content-Length t t
Content-Type * *
CSeq c m m
Date o o
Error-Info 300-699 o o
Expires o -
Expires 2xx m -
From c m m
In-Reply-To R - -
Max-Forwards R m m
Min-Expires 423 m -
MIME-Version o o
Organization o -
Priority R o -
Proxy-Authenticate 407 m m
Proxy-Authorization R o o
Proxy-Require R o o
RAck R - -
Record-Route R o o
Record-Route 2xx,401,484 o o
Reply-To - -
Require o o
Retry-After 404,413,480,486 o o
Retry-After 500,503 o o
Retry-After 600,603 o o
Route R c c
RSeq 1xx o o
Server r o o
Subject R - -
Supported R o o
Supported 2xx o o
Timestamp o o
To c(1) m m
Unsupported 420 o o
User-Agent o o
Via c m m
Warning R - o
Warning r o o
WWW-Authenticate 401 m m
]]></artwork>
</figure>
<section title="SUBSCRIBE method">
<t>"SUBSCRIBE" is added to the definition of the element "Method" in
the SIP message grammar.</t>
<t>Like all SIP method names, the SUBSCRIBE method name is case
sensitive. The SUBSCRIBE method is used to request asynchronous
notification of an event or set of events at a later time.</t>
</section>
<section title="NOTIFY method">
<t>"NOTIFY" is added to the definition of the element "Method" in
the SIP message grammar.</t>
<t>The NOTIFY method is used to notify a SIP node that an event
which has been requested by an earlier SUBSCRIBE method has
occurred. It may also provide further details about the event.</t>
</section>
</section>
<section title="New Header Fields">
<t>This table expands on tables 2 and 3 in SIP <xref
target="RFC3261"></xref>, as amended by the changes described in <xref
target="new-methods"></xref>.</t>
<figure>
<artwork><![CDATA[
Header field where proxy ACK BYE CAN INV OPT REG PRA SUB NOT
-----------------------------------------------------------------
Allow-Events R o o - o o o o o o
Allow-Events 2xx - o - o o o o o o
Allow-Events 489 - - - - - - - m m
Event R - - - - - - - m m
Subscription-State R - - - - - - - - m
]]></artwork>
</figure>
<section anchor="event-header field" title=""Event" Header Field">
<t>Event is added to the definition of the element "message-header field"
in the SIP message grammar.</t>
<t>For the purposes of matching NOTIFY messages with
SUBSCRIBE messages, the event-type portion of the "Event" header field is
compared byte-by-byte, and the "id" parameter token (if present) is
compared byte-by-byte. An "Event" header field containing an "id"
parameter never matches an "Event" header field without an "id" parameter.
No other parameters are considered when performing a comparison. SUBSCRIBE
responses are matched per the transaction handling rules in SIP
<xref target="RFC3261"/>.</t>
<t><list style="hanging">
<t>Note that the forgoing text means that "Event: foo; id=1234"
would match "Event: foo; param=abcd; id=1234", but not "Event:
foo" (id does not match) or "Event: Foo; id=1234" (event portion
does not match).</t>
</list></t>
<t>This document does not define values for event-types. These
values will be defined by individual event packages, and MUST be
registered with the IANA.</t>
<t>There MUST be exactly one event type listed per event header field.
Multiple events per message are disallowed.</t>
</section>
<section title=""Allow-Events" Header Field">
<t>Allow-Events is added to the definition of the element
"general-header field" in the SIP message grammar. Its usage is described
in <xref target="allow-events"></xref>.</t>
</section>
<section title=""Subscription-State" Header Field">
<t>Subscription-State is added to the definition of the element
"request-header field" in the SIP message grammar. Its usage is described
in <xref target="subscriber-notifies"></xref>.</t>
</section>
</section>
<section title="New Response Codes">
<section title=""202 Accepted" Response Code">
<t>The 202 response is added to the "Success" header field
definition. "202 Accepted" has the same meaning as that defined in
HTTP/1.1 <xref target="RFC2616"></xref>.</t>
</section>
<section title=""489 Bad Event" Response Code">
<t>The 489 event response is added to the "Client-Error" header field
field definition. "489 Bad Event" is used to indicate that the
server did not understand the event package specified in a "Event"
header field.</t>
</section>
</section>
<section title="Augmented BNF Definitions">
<t>The Augmented BNF definitions for the various new and modified
syntax elements follows. The notation is as used in SIP <xref
target="RFC3261"></xref>, and any elements not defined in this section
are as defined in SIP and the documents to which it refers.</t>
<figure>
<artwork><![CDATA[
SUBSCRIBEm = %x53.55.42.53.43.52.49.42.45 ; SUBSCRIBE in caps
NOTIFYm = %x4E.4F.54.49.46.59 ; NOTIFY in caps
extension-method = SUBSCRIBEm / NOTIFYm / token
Event = ( "Event" / "o" ) HCOLON event-type
*( SEMI event-param )
event-type = event-package *( "." event-template )
event-package = token-nodot
event-template = token-nodot
token-nodot = 1*( alphanum / "-" / "!" / "%" / "*"
/ "_" / "+" / "`" / "'" / "~" )
; The use of the "id" parameter is deprecated; it is included
; for backwards compatibility purposes only.
event-param = generic-param / ( "id" EQUAL token )
Allow-Events = ( "Allow-Events" / "u" ) HCOLON event-type
*(COMMA event-type)
Subscription-State = "Subscription-State" HCOLON substate-value
*( SEMI subexp-params )
substate-value = "active" / "pending" / "terminated"
/ extension-substate
extension-substate = token
subexp-params = ("reason" EQUAL event-reason-value)
/ ("expires" EQUAL delta-seconds)
/ ("retry-after" EQUAL delta-seconds)
/ generic-param
event-reason-value = "deactivated"
/ "probation"
/ "rejected"
/ "timeout"
/ "giveup"
/ "noresource"
/ "invariant"
/ event-reason-extension
event-reason-extension = token
]]></artwork>
</figure>
</section>
</section>
</middle>
<back>
<references title="Normative References">
&rfc2119;
&rfc2434;
&rfc2616;
&rfc2779;
&rfc2848;
&rfc3261;
&rfc3265;
&rfc4483;
&draft-ietf-sip-gruu;
</references>
<references title="Informative References">
&rfc3515;
&rfc3840;
&rfc3903;
&rfc4485;
&rfc4660;
&rfc5057;
</references>
<section title="Acknowledgements">
<t>Thanks to the participants in the Events BOF at the 48th IETF meeting
in Pittsburgh, as well as those who gave ideas and suggestions on the
SIP Events mailing list. In particular, I wish to thank Henning
Schulzrinne of Columbia University for coming up with the final
three-tiered event identification scheme, Sean Olson for miscellaneous
guidance, Jonathan Rosenberg for a thorough scrubbing of the -00 draft,
and the authors of the "SIP Extensions for Presence" document for their
input to SUBSCRIBE and NOTIFY request semantics.</t>
<t>
I also owe a debt of gratitude to all the implementors who have
provided feedback on areas of confusion or difficulty in the
original specification. In particular, Robert Sparks' Herculean
efforts organizing, running, and collecting data from the SIPit
events have proven invaluable in shaking out specification bugs.
Robert Sparks is also responsible for untangling the dialog usage
mess, in the form of RFC 5057.
</t>
</section>
<section title="Open Issues">
<section anchor="bug711" title="Bug 711: Allow-Events can't express template support">
<t>
OPEN ISSUE: There are several things we can do here. I have not proposed
on in particular; I would prefer to solicit initial feedback from
implementors regarding what has been developed and deployed so far.
</t>
<t>
The key problem is that support of template event packages can't be
expressed in a complete yet bounded fashion. It would not be
reasonable, for example, to require notifiers that support winfo
on arbitrary packages to include an "Allow-Events" header field
with contents like "presence, presence.winfo, presence.winfo.winfo,
presence.winfo.winfo.winfo, presence.winfo.winfo.winfo.winfo,
presence.winfo.winfo.winfo.winfo.winfo...."
</t>
<t>
One alternative would be to list event packages and template event
packages, without explicitly indicating which templates can be applied
to which other packages. In such a case, the preceding example would
be collapsed to "Allow-Events: presence, winfo". The notifier may
have local policy that limits how they can be combined -- but we have
plenty of other places where protocol allows something, but policy
forbids it.
</t>
</section>
<section title="Remove 202 Response Code?">
<t>
In practice, the 202 response code defined in RFC 3265 has proven to be
nearly useless, due to its redundancy with the "pending" state, and
its interaction with the HERFP problem. Given that 202 must be treated
as 200 if an implementation does not understand it: would removing the
202 response code cause any issues for current implementations?
</t>
</section>
<section title="Timer L and Resubscribes">
<t>
<xref target="subscription-confirmation"/> defines a new
Timer L that is used upon initial subscription to bound the
amount of time that a subscriber needs to wait for a NOTIFY.
Should this also apply to resubscribes? On one hand, the
mechanism is not as necessary, since the subscriber already
has a negotiated expiration time associated with the
subscription. On the other hand, if no NOTIFY arrives in
64*T1, it is highly likely that the notifier has gone
off the rails, which means that the subscriber can safely
clean up state associated with that subscription. The key
question involved in applying Timer L to resubscriptions is
whether doing so makes subscriptions unnecessarily brittle.
</t>
</section>
</section>
<section title="Changes">
<t>
This section, and all of its subsections, will be consolidated into
a single "Changes Since RFC 3265" section prior to publication.
Bug numbers refer to the identifiers for the bug reports kept on file
at http://bugs.sipit.net/.
</t>
<section title="Changes since RFC 3265">
<section title="Bug 666: Clarify use of expires=xxx with terminated">
<t>
Strengthened language in <xref target="subscriber-notifies"/> to
clarify that expires should not be sent with terminated, and
must be ignored if received.
</t>
</section>
<section title="Bug 667: Reason code for unsub/poll not clearly spelled out">
<t>
Clarified description of "timeout" in <xref target="subscriber-notifies"/>.
(n.b., the text in <xref target="polling-state"/> is actually pretty clear
about this).
</t>
</section>
<section title="Bug 669: Clarify: SUBSCRIBE for a duration might be answered with a NOTIFY/expires=0">
<t>
Added clarifying text to <xref target="notifier-notifies"/> explaining
that shortening a subscription to zero seconds is valid. Also added
sentence to <xref target="duration"/> explicitly allowing shortening
to zero.
</t>
</section>
<section title="Bug 670: Dialog State Machine needs clarification">
<t>
The issues associated with the bug deal exclusively with the
handling of multiple usages with a dialog. This behavior
has been deprecated and moved to <xref target="dialog-reuse"/>.
This section, in turn, cites <xref target="RFC5057"/>, which
addresses all of the issues in Bug 670.
</t>
</section>
<section title="Bug 671: Clarify timeout-based removal of subscriptions">
<t>
Changed <xref target="notifier-notifies"/> to specifically cite
Timer F (so as to avoid ambiguity between transaction timeouts
and retransmission timeouts).
</t>
</section>
<section title="Bug 672: Mandate expires= in NOTIFY">
<t>
Changed strength of including of "expires" in a NOTIFY from
SHOULD to MUST in <xref target="notifier-notifies"/>.
</t>
</section>
<section title="Bug 673: INVITE 481 response effect clarification">
<t>
This bug was addressed in <xref target="RFC5057"/>.
</t>
</section>
<section title="Bug 677: SUBSCRIBE response matching text in error">
<t>
Fixed <xref target="event-header field"/> to remove incorrect
"...responses and..." -- explicitly pointed to SIP for
transaction response handling.
</t>
</section>
<section title="Bug 695: Document is not explicit about response to NOTIFY at subscription termination">
<t>
Added text to <xref target="dialog-handling"/> indicating that
the typical response to a terminal NOTIFY is a "200 OK".
</t>
</section>
<section title="Bug 696: Subscription state machine needs clarification">
<t>
Added state machine diagram to <xref target="subscribing"/> with
explicit handling of what to do when a SUBSCRIBE never shows up.
Added definition of and handling for new Timer L to
<xref target="subscription-confirmation"/>.
Added state machine to <xref target="notifier-notifies"/> to
reinforce text.
</t>
</section>
<section title="Bug 697: Unsubscription behavior could be clarified">
<t>
Added text to <xref target="notifier-refresh"/> encouraging (but
not requiring) full state in final NOTIFY message. Also added
text to <xref target="unsubscribing"/> warning subscribers that
full state may or may not be present in the final NOTIFY.
</t>
</section>
<section title="Bug 699: NOTIFY and SUBSCRIBE are target refresh requests">
<t>
Added text to both <xref target="subscribe"/> and <xref target="notify"/>
explicitly indicating that SUBSCRIBE and NOTIFY are target refresh
methods.
</t>
</section>
<section title="Bug 722: Inconsistent 423 reason phrase text">
<t>
Changed reason code to "Interval Too Brief" in
<xref target="notifier-initial-subscribe"/> and
<xref target="notifier-refresh"/>, to match 423 reason
code in SIP <xref target="RFC3261"/>.
</t>
</section>
<section title="Bug 741: guidance needed on when to not include Allow-Events">
<t>
Added non-normative clarification to <xref target="allow-events"/>
regarding inclusion of Allow-Events in a NOTIFY for the one-and-only
package supported by the notifier.
</t>
</section>
<section title="Bug 744: 5xx to NOTIFY terminates a subscription, but should not">
<t>
Issue of subscription (usage) termination versus dialog termination
is handled in <xref target="RFC5057"/>. The text in <xref target="notifier-notifies"/>
has been updated to summarize the behavior described by 5057, and cites it
for additional detail and rationale.
</t>
</section>
<section title="Bug 752: Detection of forked requests is incorrect">
<t>
Removed erroneous "CSeq" from list of matching criteria in <xref target="forking"/>.
</t>
</section>
<section title="Bug 773: Reason code needs IANA registry">
<t>
Added <xref target="reason-codes"/> to create and populate IANA registry.
</t>
</section>
<section title="Bug 774: Need new reason for terminating subscriptions to resources that never change">
<t>
Added new "invariant" reason code to <xref target="subscriber-notifies"/>,
ABNF syntax.
</t>
</section>
<section title="Clarify handling of Route/Record-Route in NOTIFY">
<t>
Changed text in <xref target="proxy"/> mandating Record-Route in
initial SUBSCRIBE and all NOTIFY messages, and adding "MAY" level
statements for subsequent SUBSCRIBE messages.
</t>
</section>
<section title="Eliminate implicit subscriptions">
<t>
Added text to <xref target="notifier-subscribe"/> explaining
some of the problems associated with implicit subscriptions,
normative language prohibiting them.
Removed language from <xref target="notify"/> describing
"non-SUBSCRIBE" mechanisms for creating subscriptions.
Simplified language in <xref target="notifier-notifies"/>,
now that the soft-state/non-soft-state distinction is
unnecessary.
</t>
</section>
<section title="Deprecate dialog re-use">
<t>
Moved handling of dialog re-use and "id" handling
to <xref target="dialog-reuse"/>. It is documented only
for backwards-compatibility purposes.
</t>
</section>
<section title="Rationalize dialog creation">
<t>
<xref target="dialog-handling"/> has been updated to specify that
dialogs should be created when the NOTIFY arrives. Previously, the
dialog was established by the SUBSCRIBE 200, or by the NOTIFY
transaction. This was unnecessarily complicated; the newer rules are
easier to implement (and result in effectively the same behavior
on the wire).
</t>
</section>
<section title="Refactor behavior sections">
<t>
Reorganized <xref target="node-behavior"/> to consolidate behavior along
role lines (subscriber/notifier/proxy) instead of method lines.
</t>
</section>
<section title="Clarify sections that need to be present in event packages">
<t>
Added sentence to
<xref target="event-packages"/>
clarifying that event packages are expected to include
explicit sections covering the issues discussed in this
section.
</t>
</section>
<section title="Make CANCEL handling more explicit">
<t>
Text in <xref target="cancel"/> now clearly calls out
behavior upon receipt of a CANCEL. We also echo the
"...SHOULD NOT send..." requirement from <xref target="RFC3261"/>.
</t>
</section>
<section title="Remove State Agent Terminology">
<t>
As originally planned, we anticipated a fairly large number of
event packages that would move back and forth between end-user devices
and servers in the network. In practice, this has ended up not being
the case. Certain events, like dialog state, are inherently hosted at
end-user devices; others, like presence, are almost always hosted in
the network (due to issues like composition, and the ability to deliver
information when user devices are offline). Further, the concept of
State Agents is the most misunderstood by event package authors. In
my expert review of event packages, I have yet to find one that got
the concept of State Agents completely correct -- and most of them
start out with the concept being 100% backwards from the way RFC 3265
described it.
</t>
<t>
Rather than remove the ability to perform the actions previously
attributed to the widely misunderstood term "State Agent," we have
simply eliminated this term. Instead, we talk about the behaviors
required to create state agents (state aggregation, subscription
notification) without defining a formal term to describe the
servers that exhibit these behaviors. In effect, this is an
editorial change to make life easier for event package authors;
the actual protocol does not change as a result.
</t>
<t>
The definition of "State Agent" has been removed from
<xref target="definitions"/>.
<xref target="migration"/> has been retooled to discuss
migration of subscription in general, without calling
out the specific example of state agents.
<xref target="aggregation"/> has been focused on
state aggregation in particular, instead of state
aggregation as an aspect of state agents.
</t>
</section>
</section>
</section>
</back>
</rfc>
| PAFTECH AB 2003-2026 | 2026-04-23 20:48:39 |