Re: [Teas] [Detnet] Help wanted in detnet

["Adrian Farrel" <[e-mail]>]

Hey John,
 
I realise that your email was two months ago, but I don't see any responses in
the archive and so I thought I would try to understand why by re-reading your
email.
 
I think it is very probably completely unclear to everyone who is not following
Detnet exactly what you're asking for.
 
Reading your email (and none of the Detnet documents) I am not clear whether you
plan to transport deterministic Ethernet over some as yet unspecified data plane
(which you might choose or enhance based on its capabilities), whether you
propose to map the features of the transported deterministic Ethernet to
features of the transport, whether you hope to map the features of the
deterministic Ethernet to features of the transport so that only the data is
actually transported, or what. You could, for example, use a transport that has
no scheduling and queueing support, but that is also non-blocking.
 
Indeed, the word "transport" probably immediately made half of your audience
think one thing (transport network) and the other half another thing (transport
layer protocol).
 
Of course, I went the extra mile ;-) and looked at your slide. Sadly no light
was shed.
 
I suspect the answer you're looking for is not "why not use deterministic
Ethernet as your transport?" and similarly "have you thought of using a photonic
network?"
 
Can you better phrase your question for the audience?
 
Thanks,
Adrian
 
From: detnet [mailto:[e-mail]] On Behalf Of John Dowdell
Sent: 15 November 2016 05:32
To: [e-mail]; [e-mail]; [e-mail]
Subject: [Detnet] Help wanted in detnet
 
Hi all in mpls and teas groups (consolidating earlier posting only on mpls)
 
Over in detnet wg we are working on transporting deterministic Ethernet (e.g.
from 802.1TSN LANs) across the wide area for applications such as industrial
control and audio/video broadcast. One of the key assumptions is that the
underlying transport already has sufficient scheduling and queueing support to
allow proper extension of time-sensitive (or even time-critical) traffic over,
for example, MPLS.
 
To help us validate that assumption, please can someone volunteer to help us?
Feel free to come over to the detnet list. Recommendations on exactly which
flavour(s) of MPLS and/or configurations and/or gotchas would be most welcome.
 
The slides in question are here
https://www.ietf.org/proceedings/97/slides/slides-97-detnet-detnet-data-plane-02
.pptx, slide 4.
 
Many thanks in anticipation.
 
John