[NNagain] The Whys of the Wichita IXP Project

[Bill Woodcock]

> On Feb 24, 2024, at 15:03, Dave Taht <dave.taht at gmail.com> wrote:
> Isoc has 50k.

ISOC takes $100m/year out of the community; I guess it’s nice that they want to put $50k back.  That doesn’t help anyone build an IXP.  Money isn’t the issue.

> In my case you can safely assume I have 0 dollars in my pocket

Not a problem, that’s basically where everybody trying to get IXPs off the ground is.

> but substantial expertise, going all the way back to the 90s.

That helps, but again, siting and governance are always the big hurdles.  The technical part is like 5% at the end of the process.

> I would enjoy creating and running an IXP, but there are some needed upfront expenses to handle.

Those are the easy part.  The hard part is getting all the parties to agree on site and governance.

> Connected Nation identifies Fort Myers as an ideal location for an IXP.

That doesn’t count for much, since they’re not a potential peer.  What matters is whether there are networks ready to exchange traffic there, and willing to do the work to make that happen.

> Some traceroutes show traffic being backhauled to Naples (about 40 miles away).

Ok… are there any traceroutes that show interprovider traffic being exchanged _closer_ than forty miles away?  Forty miles roundtrip is 430 microseconds.  By itself, that’s not likely to be sufficient motivation for ISPs that serve both locations to build a new IXP, unless their circuits between the two are congested.  

Are there ISPs that serve only Fort Meyers?  Are there ISPs that serve Fort Meyers, and are dependent on transit (i.e. they don’t peer anywhere yet)?

> There is a huge infrastructure of hospital buildings, rich folk, and so on that makes much of the area seemingly ideal for fiber, but comcast has a lock on most of it (and the aging populace is still quite fond of broadcast tv). The internet in my building is not all that good, but various quotes for putting fiber in here have all stalled out. (and most residents just watch tv). MANY new buildings are under construction however....

None of that, by itself, is going to help, because none of that is related to the exchange.  All that matters for the exchange is that you have three ASes willing to exchange traffic in a location.  If you do, you can potentially get an IXP going.  If you don’t, then there’s no reason for an IXP, and building the shell of one won’t convince anyone to come exchange traffic there.

> So I figure step 1, in starting an IXP, is talk to the ISPs in the area, and also the local government and potential anchors like hospitals and schools?

Only if the hospitals and schools have ASNs.  Start with the ISPs.

                                -Bill

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Message signed with OpenPGP
URL: <https://lists.bufferbloat.net/pipermail/nnagain/attachments/20240224/587a88bd/attachment.sig>