Re: [Detnet] Resolving the remaining Use Cases Draft Comments - Cellular Radio and Industrial M2M

["Pascal Thubert (pthubert)" <[e-mail]>]

Hello  János

I agree to change the sentence. Zero is not reasonable in the real world, even in wired space : ) What about

"Low packet loss (e.g.,  one in a million, one in a billion, min time between losses, min packets between losses, strictly bounded number of consecutive losses)"
Maybe that's too many examples but when you look up the (old) art of "reliability " you end up with terms such as MTBF...
All the best

Pascal

From: detnet <[e-mail]> On Behalf Of János Farkas
Sent: mercredi 19 décembre 2018 17:13
To: Grossman, Ethan A. <[e-mail]>
Cc: [e-mail]; Suresh Krishnan <[e-mail]>; [e-mail]; Balázs Varga A <[e-mail]>; Maik Seewald (maseewal) <[e-mail]>
Subject: Re: [Detnet] Resolving the remaining Use Cases Draft Comments - Cellular Radio and Industrial M2M

I just tried to resolve the concern around burstless.

Actually, the numbers may not be right.

Checking IEC/IEEE 60802 contributions
https://1.ieee802.org/tsn/iec-ieee-60802-tsn-profile-for-industrial-automation/

Use Cases: http://www.ieee802.org/1/files/public/docs2018/60802-industrial-use-cases-0918-v13.pdf
and traffic type characterization: http://www.ieee802.org/1/files/public/docs2018/60802-ademaj-traffic-type-characterization-1118-v01.pdf
indicate that the most stringent requirement is zero loss.
The next one is how many consecutive packets may be lost
There are more relaxed cases.

This section is: 7.4.  Industrial M2M Asks

Maybe we should focus on the stringent cases.

What about updating the bullet to:

"Low packet loss (e.g., zero, limited number of consecutive packets)"

?

Regards,
Janos

On 12/18/2018 8:31 PM, Grossman, Ethan A. wrote:
Thanks Janos,
Doesn't 0.1-1% sound like an awful lot of packet loss?
Do I understand correctly that you mean:
"Low packet loss (0.1-1 % of packets can be lost, but only a bounded number of consecutive packets can be lost)"

I guess I still don't exactly understand the intent, can you please clarify?
Thanks, and sorry for being dense here.
Ethan.

From: János Farkas <[e-mail]><mailto:[e-mail]>
Sent: Tuesday, December 18, 2018 8:21 AM
To: Grossman, Ethan A. <[e-mail]><mailto:[e-mail]>
Cc: [e-mail]<mailto:[e-mail]>; Maik Seewald (maseewal) <[e-mail]><mailto:[e-mail]>; Balázs Varga A <[e-mail]><mailto:[e-mail]>; [e-mail]<mailto:[e-mail]>; Suresh Krishnan <[e-mail]><mailto:[e-mail]>
Subject: Re: Resolving the remaining Use Cases Draft Comments - Cellular Radio and Industrial M2M

Hi Ethan,

Maybe it is not a god phrase, but the intention with the phrase that the loss should be burstless was that consecutive packets should not be lost.
Perhaps the maximum consecutive loss tolerance parameter describes it better in https://tools.ietf.org/html/draft-ietf-detnet-flow-information-model-02<https://urldefense.proofpoint.com/v2/url?u=https-3A__tools.ietf.org_html_draft-2Dietf-2Ddetnet-2Dflow-2Dinformation-2Dmodel-2D02&d=DwMD-g&c=lI8Zb6TzM3d1tX4iEu7bpg&r=ZcHC6wX_gDwPDcfMaFNZiQ&m=0W0EQyafQ7HdKxGYDUDOn91m44yUJHQre2QdC0Pq5rI&s=xtlwetwJuYWGjTBmpNfLC8CvvhDk4x1YkRjNDTE0rws&e=>. It is the "maximum    number of consecutive packets whose loss can be tolerated"

Maybe updating the bullet in the use cases draft to:

"Low packet loss (0.1-1 %, limited number of consecutive packets)"

?

Regards,
Janos
On 12/16/2018 6:03 AM, Grossman, Ethan A. wrote:
Balazs, Janos:
* Section 7.4

What does burstless mean (especially in the context of low packet loss)?