|
RE: [Detnet] WG Review: Deterministic Networking (detnet)
2015-09-21 10:15:53
PWEs are definitely within the scope of technologies to consider/study further.
At some point of the charter massaging PWEs were part of it or at least
suggested to be included explicitly. I would include PWEs and even GRE to the
list of possible candidate data plane technologies just that those do not come
as a “surprise” or us being restricted not to work on solutions based on them.
- Jouni
From: detnet [mailto:detnet-bounces(_at_)ietf(_dot_)org] On Behalf Of Andrew G.
Malis
Sent: Friday, September 18, 2015 9:53 AM
To: Adrian Farrel
Cc: detnet WG; Lou Berger; IETF Discussion; IESG
Subject: Re: [Detnet] WG Review: Deterministic Networking (detnet)
Adrian and Lou,
When I first read the draft charter, my immediate reaction was that the scope
of work would be deterministic IP and MPLS flows layered over a deterministic
Ethernet infrastructure as defined by IEEE. This would probably be pretty
straightforward work.
However, your conversation got me to read the charter more closely, and while
the word "pseudowire" isn't used, the inclusion of the PALS WG in the charter
implies to me that the scope of work could include the transport of
deterministic Ethernet flows (as defined by IEEE) within pseudowires carried by
arbitrary IP and/or MPLS infrastructures. All of a sudden, the work is much
less straightforward. If this is indeed part of the scope of work, it should be
explicit in the charter (or explicitly excluded if not).
Cheers,
Andy
On Fri, Sep 18, 2015 at 12:26 PM, Adrian Farrel
<adrian(_at_)olddog(_dot_)co(_dot_)uk<mailto:adrian(_at_)olddog(_dot_)co(_dot_)uk>>
wrote:
Thanks Lou,
I think we're in agreement that a new WG would be helpful as a place to
coordinate whatever work is needed and to provide a discussion forum. This is
partly because there is no existing place where this work wold not provide a
distraction.
It is possible that the location for the work is RTG if the applicability
document is describing the applicability of some of the control plane protocols,
although applying RSVP would possibly put it in TSV. And if the work is applying
IP, it might be in INT. Not so sure that this is a really important issue.
But I am still left looking at the current charter text and thinking it is not
describing the applicability statement that we are discussing. If my paragraph
that you quoted describes the work well, can we do some serious edits to the
charter to make it substantially clearer what the WG is actually doing. I might
suggest removing nearly all of the text and replacing it with a short paragraph
that says something like what I wrote (with perhaps a few more words). Currently
I find the text confusing in scope and very open to misinterpretation.
Thanks,
Adrian
-----Original Message-----
From: Lou Berger
[mailto:lberger(_at_)labn(_dot_)net<mailto:lberger(_at_)labn(_dot_)net>]
Sent: 18 September 2015 16:52
To:
adrian(_at_)olddog(_dot_)co(_dot_)uk<mailto:adrian(_at_)olddog(_dot_)co(_dot_)uk>;
ietf(_at_)ietf(_dot_)org<mailto:ietf(_at_)ietf(_dot_)org>;
iesg(_at_)ietf(_dot_)org<mailto:iesg(_at_)ietf(_dot_)org>
Cc: 'detnet WG'
Subject: Re: WG Review: Deterministic Networking (detnet)
Hi Adrian,
I have to say that there were times that I felt the same as you, and
questioned what a DetNet WG would / should do. I think you hit the key
points in your mail and the main work that needs to be done is to say
how all the pieces fit together when operating over IETF owned data
planes, i.e. IP and MPLS, without modification. I think your last
paragraph summarizes the work to be done quiet well.
Are you sure that this work is more than an applicability statement that
shows
how existing tools are used to achieve the desired function. This document
might
cover data plane, OAM, packet classification, configuration, control plane,
security, etc. That would be useful work and would probably need a WG to
achieve
the necessary discussion.
This answers the question that the work belongs in the IETF in some WG,
but doesn't say that a new WG is needed. I came the conclusion that a
new WG is needed to ensure that the overall solution "works" and that
the data plane details are sufficiently defined.
Does this help?
Lou
On 9/18/2015 11:38 AM, Adrian Farrel wrote:
Hi IESG,
I am struggling to understand why this work is being proposed in the Routing
Area. Actually, I am slightly struggling to understand why it is being
proposed
for the IETF.
This is not say I don't think a WG is needed, but the only work I see
described
here is a documentation of data plane work and an "overall architecture". I
assume that any modification to a layer 2 data plane will be carried out by
the
SDO that owns that data plane. In particular, if changes to Ethernet are
needed,
they will be done in the IEEE. So, that leaves us with work at L3 for which
the
proposed charter text says IP or MPLS. Now, it seems to me that any change
to
IP
or MPLS in the forwarding plane is alarming, and also that any change to IP
would need to be done in the Internet Area.
At the same time, the charter explicitly puts discussion of control plane
out of
scope.
Are you sure that this work is more than an applicability statement that
shows
how existing tools are used to achieve the desired function. This document
might
cover data plane, OAM, packet classification, configuration, control plane,
security, etc. That would be useful work and would probably need a WG to
achieve
the necessary discussion.
Adrian
-----Original Message-----
From: IETF-Announce
[mailto:ietf-announce-bounces(_at_)ietf(_dot_)org<mailto:ietf-announce-bounces(_at_)ietf(_dot_)org>]
On Behalf Of
The
IESG
Sent: 18 September 2015 15:51
To: IETF-Announce
Cc: detnet WG
Subject: WG Review: Deterministic Networking (detnet)
A new IETF working group has been proposed in the Routing Area. The IESG
has not made any determination yet. The following draft charter was
submitted, and is provided for informational purposes only. Please send
your comments to the IESG mailing list (iesg at ietf.org<http://ietf.org>)
by 2015-09-28.
Deterministic Networking (detnet)
------------------------------------------------
Current Status: Proposed WG
Assigned Area Director:
Deborah Brungard
<dbrungard(_at_)att(_dot_)com<mailto:dbrungard(_at_)att(_dot_)com>>
Mailing list
Address: detnet(_at_)ietf(_dot_)org<mailto:detnet(_at_)ietf(_dot_)org>
To Subscribe: https://www.ietf.org/mailman/listinfo/detnet
Archive: https://mailarchive.ietf.org/arch/browse/detnet/
Charter:
The Deterministic Networking (DetNet) Working Group focuses on
deterministic data paths that operate over Layer 2 bridged and Layer 3
routed segments, where such paths can provide bounds on latency, loss,
and packet delay variation (jitter), and high reliability. The Working
Group addresses Layer 3 aspects in support of applications requiring
deterministic networking. The Working Group collaborates with IEEE802.1
Time Sensitive Networking (TSN), which is responsible for Layer 2
operations, to define a common architecture for both Layer 2 and Layer
3. Example applications for deterministic networks include professional
and home audio/video, multimedia in transportation, engine control
systems, and other general industrial and vehicular applications being
consider by the IEEE 802.1 TSN Task Group.
The Working Group will initially focus on solutions for networks that
are under a single administrative control or within a closed group of
administrative control; these include not only campus-wide networks but
also can include private WANs. The DetNet WG will not spend energy on
solutions for large groups of domains such as the Internet.
The Working Group will specify an overall architecture that encompasses
the data plane, OAM (Operations, Administration, and Maintenance), time
synchronization, management, control, and security aspects which are
required to enable a multi-hop path, and forwarding along the path, with
the deterministic properties of controlled latency, low packet loss, low
packet delay variation, and high reliability. The work applies to
point-to-point (unicast) and point-to-multipoint (multicast) flows which
can be characterized in a manner that allows the network to 1) reserve
the appropriate resources for the flows in advance, and 2) release/reuse
the resources when they are no longer required. The work covers the
characterization of flows, the encapsulation of frames, the required
forwarding behaviors, as well as the state that may need to be
established in intermediate nodes. Candidate Layer 3 data plane
technologies that may be used, without modification, include: IP and
MPLS.
The working group will document which deployment environments and types
of topologies are within (or outside) the scope of the DetNet
architecture. This work focuses on the data plane aspects and is
independent from any path setup protocol or mechanism. The data plane
will be compatible with the work done in IEEE802.1 TSN.
The Working Group's scope explicitly excludes modifications of transport
protocols, OAM, Layer 3 forwarding, encapsulations, and control plane
protocols.
DetNet is chartered to work in the following areas:
Overall architecture: This work encompasses the data plane, OAM,
time synchronization, management, control, and security aspects.
Data plane: This work will document how to use IP and/or MPLS to
support a data plane method of flow identification and packet
forwarding over Layer 3.
Data flow information model: This work will identify the information
needed for flow establishment and control and be used by a
reservation protocol or by YANG data models. The work will be
independent from the protocol(s) used to control the flows
(e.g. YANG+NETCONF/RESTCONF, PCEP or GMPLS).
Identification of additional YANG models: This work will document
device and link capabilities (feature support) and resources
(e.g. buffers, bandwidth) for use in device configuration and status
reporting. Such information may also be used when advertising the
deterministic network elements to a control plane. Control plane
related information will be independent from the protocol(s) which
may be used to advertise this information (e.g. IS-IS or GMPLS
extensions). Any new YANG models will be coordinated with the
Working Groups that define any augmented base models.
As needed, problem statement: This effort will establish the
deployment environment and deterministic network requirements.
As needed, vertical requirements: This effort will detail the
requirements for deterministic networks in various industries, for
example, professional audio, electrical utilities, building
automation systems, wireless for industrial applications.
To investigate whether existing data plane encryption mechanisms can
be applied, possibly opportunistically, to improve security and
privacy.
The WG coordinates with other relevant IETF Working Groups, including
CCAMP, PCE, PALS, TEAS, OSPF, IS-IS, TSVWG, and 6TisSCH. As the work
progresses, requirements may be provided to the responsible Working
Group, e.g. PCE, TEAS, and CCAMP, with DetNet acting as a focal point to
maintain the consistency of the overall architecture. The WG will liaise
with appropriate groups in IEEE and other Standards Development
Organizations (SDOs).
WG deliverables include:
As standard track or informational RFCs
Overall architecture
Data plane specification
Data flow information model
YANG model augmentations
WG sustaining/informational documents may include:
These documents may not necessarily be published, but may be
maintained in a draft form or on a collaborative Working Group wiki
to support the efforts of the Working Group and help new comers:
Problem statement and (constrained) deployment environments
User-driven use cases
Milestones:
|
|