Re: Confirmation to advance: draft-ietf-6man-ipv6only-flag-05

[Nick Hilliard <[e-mail]>]

David Farmer wrote on 28/05/2019 13:53:
> Personally, my issue is IPv4 LL, in particular, service 
> advertisement/discovery traffic using IPv4 LL. This service 
> advertisement/discovery traffic can be sizable and is already 
> problematic for many WiFi networks so the elimination of such traffic 
> that is by definition futile is an unqualified win in such situations.

quantifying this, if you are able to use ethertype ACLs, all the host 
will see is its own mdns transmits.  These are typically small scale (< 
1pps measured over longer timescales); they would be coalesced with 
other transmits on wifi devices, and the transmit mechanism for these 
multicast frames would be wifi unicast to the AP, i.e. cheap.  There 
would be no transmits from AP to host.

If RFC2563 were used to block ipv4 LL autoconfig, then LL traffic would 
in theory disappear and all you'd be left with would be the occasional 
DHCP request.

I.e. in both cases, background ipv4 chatter would drop to operationally 
insignificant levels.

In respect of unqualified wins, there ain't no such thing as a free 
lunch.  Each option has its costs and will be a balance between 
complexity, protocol purity, efficiency, security, ease of 
implementation, scope of application and many other things.

Operationally, dropping 95%+ of ipv4 background chatter would be a great 
win.  This is trivial with ethertype ACLs, and my suspicion is that 
widescale implementation of RFC2563 would give comparable results.  The 
remaining chatter is not very important in the scale of things.

> Also, the testing that I did shows that most newer clients no longer 
> support RFC2563. I'd be happy to be wrong on this and I'm not 100% sure 
> my test was valid, but even if I'm wrong this shows that making RFC2563 
> work as desired isn't trivial.

that may well be the case, but the protocol exists.  It's difficult to 
justify a new protocol when an existing standard which does 95%+ of what 
the new one proposes hasn't been widely implemented.

Nick