[Atlas-anchors-pilot] Q&A, part 1

[Gert Doering]

Hi,

On Mon, Sep 17, 2012 at 11:28:45PM +0200, Romeo Zwart wrote:
> There is one apparent misunderstanding that I'd like to correct. In the
> list of hardware details that we sent to this list a while ago, the
> amount of RAM specified could easily misread as 8GB, whereas the
> specification really was for 32GB, not 8GB.

While RAM is cheap, I still think we should take a step back and ask
ourselves "what do we want the Atlas Anchor boxes to do, and what amount
of resources will it need to do that".

If we set out to build a generic "RIPE NCC in a box" system, something
with 4x600G SAS hard disks and 32G RAM might make sense.

If we set out to build a replacement for the TTM network, something
slightly bigger than a 50EUR TL-WR1043ND would easily saturate a 100M
port with probe reply traffic.

As stated before, we're not willing to buy a "RIPE NCC in a box" system
out of our own budget - we have no interest paying for a NRTM box or 
a dedicated-for-us K.Root instance (there's enough root servers in our 
region, we currently peer with 3 or 4 instances at national peering points,
and we pay our share of K.Root already with the LIR fees).


If there's enough interest in building a "RIPE NCC in a box" system,
then please stop calling it "Atlas Anchor", because that's not what it
would be.

Sorry if I sound like a spoilsport, but I think this is really the core
of the issue: understand what you're trying to build, and name it so.

Gert Doering
        -- NetMaster
-- 
have you enabled IPv6 on something today...?

SpaceNet AG                        Vorstand: Sebastian v. Bomhard
Joseph-Dollinger-Bogen 14          Aufsichtsratsvors.: A. Grundner-Culemann
D-80807 Muenchen                   HRB: 136055 (AG Muenchen)
Tel: +49 (89) 32356-444            USt-IdNr.: DE813185279