next steps

[Dave Taht]

I am trying to capture all the relevant information. I'm not done yet,
and need to get some dinner.

The diagram will be pulled from git and built every 15 minutes, and
placed here in
svg, png, and ps formats.

http://huchra.bufferbloat.net/~d/sdcpe-demo-doc/

(in other words: hit reload on
http://huchra.bufferbloat.net/~d/sdcpe-demo-doc/shownet.png
)

I have put the start at the doc and diagram and generator in github

git clone git://github.com/dtaht/sdcpe-demo-doc.git # for a read-only copy

You can build it with 'make' if you have graphviz installed.

if you have a github account, let me know what it is, and I'll give
you r/w access.



On Sun, Mar 18, 2012 at 4:08 PM, Francis Dupont <fdupont at isc.org> wrote:
> I have some concerns about the image:
>  - the name AFTR CGN is incorrect: an AFTR is a kind of CGN,
>  I suggest to use the SD-AFTR names

Changed to 'SD-AFTR 1... and 2'

>
>  - we don't know what is the hardware used for the SD-AFTRs

The laptops

>  - the internal network (where is the switch has not the given
>  IPv4 prefixes, BTW it has one IPv6 prefix (in 2001:db8::/32)
>
>  - we need the Ethernet (and IPv6 and IPv4) addresses of the 2 SD-AFTRs

Noted.

>
>  - we need one IPv6 address for the SD-AFTR service (and the IPv4
>  address too but we don't know it yet)

We need to simulate using my ipv6 allocation in the lab, which is a
2001:4f8:fff8:/48

everything above 2001:4f8:fff8:800 is currently available.


>  - I'd like to get the IPv4 addresses for the 2 SD-B4s (the usage
>  is to give the address and the prefix length)

We need those routable and from isc's it dept, at this point?

>
> So to summary we have 100 bits of info in 46093 bytes...

I am hoping the alternate format - as in git, as also a raw .gv, has a
suitable compression ratio.

>
> Regards
>
> Francis Dupont <fdupont at isc.org>

I realize I am not done the diagram or data collection, I do have some
open questions on the graph, and I'm going to dinner.


-- 
Dave Täht
SKYPE: davetaht
US Tel: 1-239-829-5608
http://www.bufferbloat.net