[PDB-tech] Netnod Stockholm LANs same IX

[Martin J. Levy]

One questioned could be ... does this provide a mechanism to handle IX fabric renumbering? Presently we add the new cidr back to the existing fabric. Is there a codified way to handle this?

Martin 

> On Apr 15, 2016, at 10:51 PM, Chris Caputo <ccaputo at alt.net> wrote:
> 
>> On Sat, 16 Apr 2016, Matt Griswold wrote:
>> * "Martin J. Levy" <mahtin at mahtin.com> [160414 13:52 -0700]:
>>> +1 
>>> 
>>> This could not be handled by PDB1.0 cleanly however 2.0 can
>>> handlev it. Same for some otter IXs like this.
>>> 
>>> How do we update the DB to add data to the MTUs and/or VLAN numbers
>>> columns and more? Should NETNOD request this?
>> 
>> Netnod can (and should ;) do it themselves, yes - the new permissions
>> model supports full control by IX operators. Any other IXPs in similar
>> situations should as well and may email support at peeringdb.com if
>> they'd like help doing so.
>> 
>> I'm not a fan of the IX LAN separation - it was used in part to help
>> with the Euro-IX/IX-F database sync, and seemed like a great idea. In
>> practice I think it overly complicates everything it touches to offer a
>> tiny bit more description to what is really just an edge case.
>> 
>> Dropping it in favor of each LAN being a separate IX, while moving the
>> fields on the LAN record to the base IX, seems to be a much better
>> solution. The only issue I see in getting rid of the separate LANs on
>> the same IX record is association to a parent of sorts, and I think the
>> Org + area covered does that adequately.
>> 
>> Would love to hear any feedback on that - it will be one of the first
>> topics for the newly forming Product Committee.
> 
> Separation of ix records by VLAN sounds fine to me.  Right now looking at:
> 
>  https://peeringdb.com/ix/13
> 
> is confusing because there are addresses from multiple VLANs being shown 
> in the same list without a VLAN label.  The Altopia entries being a prime 
> example.
> 
> If not separation, maybe a selector on the "Peers at this Exchange Point" 
> would work.
> 
> Thanks,
> Chris
>