Re: [v6ops] Incremental Deployment of IPv6-only Wi-Fi for IETF Meetings

[Lee Howard <[e-mail]>]

On 7/17/17, 10:27 AM, "v6ops on behalf of JORDI PALET MARTINEZ"
<[e-mail] on behalf of [e-mail]> wrote:

>
>4) The result is the same, but people don’t need to spend time. We
>configure the Jool so it reports EVERY usage, so we can then process that
>file to detect WHAT was using the CLAT.

Your proposal is to log the 5-tuple of every connection (using the CLAT)
on the IETF network?
Any privacy concerns?

>5) The result is the same as the suggested experiment, but NOT breaking
>anything, not asking the participants to report anything, but having
>BETTER collection of data about what will be broken if the CLAT is not
>there (anything that goes into the CLAT).

I’m not sure that logging the 5-tuple (if that’s what you mean to log)
gives you that information.
For instance, if Outlook for Mac isn’t working, and we don’t know why,
then you may not see anything; the DNS lookup may happen over IPv6, and
OfM barfs without ever sending a packet. At most you’ll see a packet to an
SMTP/POP/IMAP port, but it won’t tell you whether it was the client or the
server that failed.
I’m not sure how to know of VPN failures where the DNS goes through the
VPN. 

>
>
>Real world networks (end-users and corporate which are not related to
>IETF), will not (in 3-5 years), be willing to replace all the apps and
>devices that fail with just NAT64. They need a CLAT or similar support in
>their LANs.

I don’t know how many corporate apps/devices fail with IPv6 + NAT64. Maybe
some legacy business systems, but that’s not the case for all, and won’t
be in 3-5 years. Even if it is, it may well make more sense for an
enterprise to use IPv6 on client networks, and surround legacy systems
with NAT64.

No, I don’t expect 90% of users will be doing IPv6-only in their homes or
corporate LANs in 3 years, but I wouldn’t be surprised if it’s 10-20%.

Lee