Re: [amqp] [management] Proposal for event notifications (initial outline)

[Rob Godfrey <rob.j.godfrey@gmail.com>]

Wrote this (while on the plane) in reply to the earlier proposal.
While I've been in the air I see you've put out a revised proposal
which I shall look at on my next flight...

> On 13 February 2015 at 19:12, Alan Conway <aconway@redhat.com> wrote:
>         After some encouraging feedback, here it is again specified
>         and
>         simplified (thanks Jakub). I'm still interested in feedback on
>         the spec
>         or potential implementation issues. I will probably be
>         implementing this
>         or something like it shortly.
>
>         Cheers
>         Alan.
>
>         ----
>         # AMQP Management Event Notification
>
>         Management clients often need to know when the state of a
>         managed entity
>         has changed.  This can be for monitoring or logging purposes,
>         or to
>         react when something becomes ready for use.
>
>         Polling the management agent for changes with READ or QUERY
>         operations
>         is inefficient. Clients either poll too often, which overloads
>         the
>         agent, or too seldom which causes needless delays in the
>         system.
>
>         Event notification allows management clients to subscribe for
>         continuous
>         notification of selected events as soon as the agent becomes
>         aware of
>         them.
>
>         ## Event Types
>
>         An Event Type is a named set of attributes. The name is a
>         case-sensitive
>         string. Implementers MAY define their own Event Types which
>         MUST be
>         named using a reverse domain name prefix owned by the
>         implementer, e.g.
>         "com.example.broker.queueDeleted".

Are event types releative to entity types or not? I can't help feeling it would
be advantageous to define a more concrete relationship between entity types and
event tyoes. In the example given would it make more sense to have a "deleted"
event type which can be generated relative to entities of given types (such as
com.example.queue). Obviously we would then want to have a special (or null)
entity type for those event types which do not relate sepcifically to a managed
entity.

Would we want to expose the meta data of which event types
can be generate by entities of a given entity type? Do we otherwise retrun data
about event types through the meta data available at the management node, or is
the meta data (such as which attrbutes an event type carries) simply assumed to
be conveyed by an external mechanism?

>
>         Every Event Type MUST have an attribute "eventType" which
>         carries the
>         Event Type name as a string. Implementers may define
>         additional required
>         or optional attributes for their Event Types.
>
>         Event Types do not have operations, annotations or extensions.
>         No
>         standard event types are defined by this specification.
>
>         ## Notification Messages
>
>         A notification message contains a single event. The message
>         subject MUST
>         be the Event Type name.  The application-properties MUST
>         include all
>         required attributes for that Event Type, in particular
>         "eventType" with
>         the Event Type name.  They MAY include any optional attributes
>         for the
>         Event Type.
>

If we decide to relate event types to entity types, I think we'd here want to
define the standard fields for the entity type and entity id.

As an aside, do events have unique ids or sequence numbers, or timestamps? At
the very least perhaps we should give names for these attributes even if they
are not supported by every event type.

>         ## Subscribing for Event Notification
>
>         To subscribe for notification messages, a client creates a
>         link from the
>         address "$management.events" with a *filter* (TODO xref)
>         describing the
>         events that it is interested in.  Closing the link ends the
>         subscription.
>

Are you suggsting that every management node have a corresponding ".events"
node, or that this applies soley to the special case $management?  Is the
support of events mandatory for every management node?

One thought here is that we should more explicitly define an address/topic
hierarchy and use wilcard pattern matching.  In the bindmap group Shawn has
suggested (and we agreed in principle on the core TC call last time) to use
MQTT-like pattern matching.  This could give us something like <management node
name>/<event type>/<entity type>/<entity id> where # and + could be used as the
matching wildcards.  This would allow someone to subscribe for "all delete
events" or "all events related to queues" or "all events related to queue id X".

>         The filtering capabilities depend on the filters provided by
>         the agent.
>         An agent SHOULD provide the APACHE.ORG:SELECTOR described in
>         <http://www.amqp.org/specification/1.0/filters>.  This allows
>         clients to
>         filter on arbitrary SQL-like expressions over all the
>         attributes of the
>         event including "eventType".
>

Even if we decide for SQL like filters then we probably want to look at what the
JMS mapping spec comes up with here, I don't think the Apache filter is quite
what we want.



>         Since the event type MUST also be included in the message
>         subject, even
>         simple filters (for example the legacy filters described in
>         <http://www.amqp.org/specification/1.0/filters>) MAY be used
>         to filter
>         by event type.
>
>         Any filter that can operate on the subject or
>         application-properties of
>         a message will be able to work with event notification
>         messages, but for
>         flexibility and interoperability implementors SHOULD provide
>         APACHE.ORG:SELECTOR.
>
>         ----
>
>         # Notes, not part of the spec:
>
>         Needs cross references to relevant sections on filters etc.
>
>         This design does not allow multiple events to be batched in a
>         single
>         message. IMO an implementation should batch messages at the
>         transport
>         (into TCP packets) NOT batch events into a message. Batching
>         in messages
>         complicates the client, defeats transport level batching (by
>         forcing
>         "fat" messages on the transport) and rules out (or badly
>         complicates)
>         the use of AMQP filters to select events.
>
>         I considered an alternative design where the client sends a
>         subscription
>         request with a reply-to address and the agent then sends
>         notification
>         messages to the reply-to address.
>
>         I rejected that design for the following reasons:
>
>         1. It is impossible in general to tell when such a subscriber
>         goes away
>         if it crashes or fails to send an explicit unsubscribe
>         message.
>         Implementations must already deal with closing links when
>         clients
>         disconnect or heartbeat out so using links solves this
>         problem. Even in
>         an indirect topology (like link-routed Qpid dispatch networks)
>         the
>         routers must proxy link closure to be AMQP-correct.
>
>         2. We should use AMQP standard features to solve AMQP
>         problems, not
>         re-invent the wheel in a form only useful for management.
>         Filters/selectors were designed for exactly this purpose we
>         should make
>         them work for us.
>
>
>
>
>         ---------------------------------------------------------------------
>         To unsubscribe from this mail list, you must leave the OASIS
>         TC that
>         generates this mail.  Follow this link to all your TCs in
>         OASIS at:
>         https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
>
>
>