Mirja,
I haven't got feedback from Collin on the proper wording to be inserted.
Do you think if adding the following sentence (in purple) to Section 3.2 will
be good enough?
“This method requires multiple copies of the same packet to all NVEs that
participate in the VN. If, for example, a tenant subnet is spread across 50
NVEs, the packet would have to be replicated 50 times at the source NVE.
Obviously, this approach creates more traffic to the network that can cause
congestion when the network load is high. This also creates an issue with the
forwarding performance of the NVE.
In addition, if ECN is enabled, NVEs should following the guidelines
described in [ECN-ENCAP].”
Please let us know as soon as possible, so we can upload the new version to
move the draft forward.
Thank you. Linda
-----Original Message-----
From: Linda Dunbar
Sent: Tuesday, October 10, 2017 2:52 PM
To: 'Mirja Kuehlewind (IETF)' <[email protected]>; 'Colin Perkins'
<[email protected]>
Cc: The IESG <[email protected]>; [email protected]; Matthew Bocci
<[email protected]>; [email protected]; [email protected];
[email protected]
Subject: RE: Mirja Kühlewind's Discuss on draft-ietf-nvo3-mcast-framework-10:
(with DISCUSS)
Collin,
Per Mirja, can you provide a sentence to explain the following?
- If there is no congestion caused by replication by the NVE for multicast
frame, how is ECN impacted?
- Is the impact only specific to the replication for multicast data frame or is
the impact same to uni-cast data frame as well?
Thanks, Linda
-----Original Message-----
From: Mirja Kuehlewind (IETF) [mailto:[email protected]]
Sent: Tuesday, October 10, 2017 2:00 AM
To: Linda Dunbar <[email protected]<mailto:[email protected]>>
Cc: The IESG <[email protected]<mailto:[email protected]>>;
[email protected]<mailto:[email protected]>; Matthew Bocci
<[email protected]<mailto:[email protected]>>;
[email protected]<mailto:[email protected]>;
[email protected]<mailto:[email protected]>;
[email protected]<mailto:[email protected]>
Subject: Re: Mirja Kühlewind's Discuss on draft-ietf-nvo3-mcast-framework-10:
(with DISCUSS)
Hi Linda,
the point is not that this may cause congestion or to make NVO3 ECN enabled.
The points are that if packet replication is used that may break some
assumption that the congestion control that is already in use for multicast
traffic makes (see Colins mail). On ECN there is actually nothing specific
about using ECN in an NVO3 network; you can probably just do it. However if you
use tunneling it’s good to point people at
draft-ietf-tsvwg-ecn-encap-guidelines-09. That’s all.
Maybe you can go back to Colin and ask him to propose some short text.
Mirja
> Am 09.10.2017 um 23:34 schrieb Linda Dunbar
> <[email protected]<mailto:[email protected]>>:
>
> Mirja,
>
> Actually sending more copies of the same packet, especially multicast
> packets, rarely cause congestion.
> Congestion only occurs when links’ utilization reach close to 100% as most
> switches/routers today can handle wire speed forwarding. The rule of thumb of
> deploying network is 50% link utilization, especially in data center where
> links can be added very easily.
>
> In addition, Application based multicast is less than 2% of total traffic.
> Anyway, to make Collin happy, I already inserted one sentence in the
> following original paragraph (in purple) in -10 version which has been
> uploaded:
>
> “This method requires multiple copies of the same packet to all NVEs that
> participate in the VN. If, for example, a tenant subnet is spread across 50
> NVEs, the packet would have to be replicated 50 times at the source NVE.
> Obviously, this approach creates more traffic to the network that can cause
> congestion when the network load is high. This also creates an issue with the
> forwarding performance of the NVE.”
>
> ECN for NVO3 network is totally different subject. It will take a lot to
> describe ECN clearly for NVO3 network. To the least, it is not even clear of
> the scenario, use cases, and implication for ECN for NVO3.
>
> Therefore, we don’t believe it is within the scope of this draft to discuss
> or reference ECN for this draft.
>
> Linda Dunbar
>
> -----Original Message-----
> From: Mirja Kühlewind [mailto:[email protected]]
> Sent: Monday, October 09, 2017 10:57 AM
> To: The IESG <[email protected]<mailto:[email protected]>>
> Cc:
> [email protected]<mailto:[email protected]>;
> Sam Aldrin <[email protected]<mailto:[email protected]>>; Matthew
> Bocci
> <[email protected]<mailto:[email protected]>>;
> [email protected]<mailto:[email protected]>;
> [email protected]<mailto:[email protected]>;
> [email protected]<mailto:[email protected]>
> Subject: Mirja Kühlewind's Discuss on draft-ietf-nvo3-mcast-framework-10:
> (with DISCUSS)
>
> Mirja Kühlewind has entered the following ballot position for
> draft-ietf-nvo3-mcast-framework-10: Discuss
>
> When responding, please keep the subject line intact and reply to all email
> addresses included in the To and CC lines. (Feel free to cut this
> introductory paragraph, however.)
>
>
> Please refer to https://www.ietf.org/iesg/statement/discuss-criteria.html
> for more information about IESG DISCUSS and COMMENT positions.
>
>
> The document, along with other ballot positions, can be found here:
> https://datatracker.ietf.org/doc/draft-ietf-nvo3-mcast-framework/
>
>
>
> ----------------------------------------------------------------------
> DISCUSS:
> ----------------------------------------------------------------------
>
> Based on the feedback provided by the tsv-art review (Thanks Colin!) I would
> like to see a paragraph or short section that discusses how replication as
> used in section 3.2 and 3.3 can impact multicast congestion control and also
> provides a pointer to draft-ietf-tsvwg-ecn-encap-guidelines-09 in case ECN is
> supported in the NVO network which can likely be the case in data center
> scenarios.
_______________________________________________
nvo3 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/nvo3
