[arin-ppml] [ppml] Policy to help the little guys
Jo Rhett
jrhett at svcolo.com
Mon Apr 28 17:20:12 EDT 2008
On Mar 20, 2008, at 10:18 PM, Ted Mittelstaedt wrote:
> The last time we bought bandwidth I had to argue with the salesguy
> to get them to understand we wanted a full unrestricted table sent
> to us.
> By default, they filtered their outgoing table at the /24 boundary,
> and that was just the beginning, they sent a worksheet listing all of
> the default filters they put in additionally to this.
I've been involved with trying to help some smaller operators with
the exact opposite problem. The large providers appear to be taking
a "you can do anything you want with the routes we give you" hands
off approach. I actually never experienced the problem you are
mentioning. Was this a tier-1 or a regional ISP? (I don't have much
if any experience with the regionals)
We don't have problems with routes sent, but a lot of the peer-1s
seem to have major problems tagging their routes according to their
own documentation. I spent at least an hour a week dropping notes to
our peers/uplinks about problems with their community strings.
> Yes I guess there's networks out there who are still in the Dark
> Ages. I would submit that when the fertilizer impacts the rotating
> air-moving device that they and their customers will get a very
> rapid education in the Correct Way to do things. And if they and
> their customers are too incompetent to understand what is going
> on, they will go bankrupt and the rest of us will be well rid of them.
>
> So, my question still stands - are we discussing theory or practice?
If incompetence or lack of quality testing actually killed off
companies, the market would be very different from what it is today.
It's a fairly easy argument that the largest vendor of network gear
has the least amount of QA in their releases. By your logic they
would have failed by now, and it just hasn't happened.
Although I like your logic ;-)
--
Jo Rhett
senior geek
Silicon Valley Colocation
Support Phone: 408-400-0550
More information about the ARIN-PPML
mailing list