ICANN off the map

Far be it from me to introduce operational content into the mix of the
usual suspects referring to anyone that doesn't adhere to their worldview
as ranting or being sociopaths, however there doesn't seem to be any
announcements for ICANNs address space at the moment as near as I can tell
from polling the various route servers, rendering them unreachable.

I know of no critical Internet infrastructure service which requires
www.icann.org to be reachable in order to function.

I suspect even cybernothing.org occasionally does maintenance, and it
is unreachable for periods of time. But the Internet survives without
cybernothing.org being reachable.

I don't believe I ever said it did. If you think that I get some sort of
pleasure out of them being down, I don't.

As far as I can tell, DNS for icann.org is functioning as designed.
It has several name servers dispersed across geographical areas and
network topology which continue to give authoritative answers for the
icann.org zone without interruption. Isn't that how its supposed to
work?

ICANN has nothing to do with routing, the world wide web, or lots
of other things on the Internet.

No but the fact that sh ip bgp 192.0.0.0 results in "Network not in table"
from every publicly available route server I've checked has everything to
do with routing thus it would seem that this is somewhat operationally on
topic.

This isn't about ICANN per se, but rather about a good chunk of address space
being unreachable for some unknown reason that has very little to do with
"normal maintenance."

That the people that want to run the Internet are too incompetent to sign
up with Akamai or utilize some other mechanism to ensure that people can
reach their site in the event of an outage like any reasonably
serious business would do is a seperate matter.

But since you bring it up, if they can't run a website, do you want them
running the Internet?

a message of 34 lines which said:

That the people that want to run the Internet are too incompetent to sign
up with Akamai or utilize some other mechanism to ensure that people can
reach their site in the event of an outage like any reasonably
serious business would do is a seperate matter.

ICANN is a customer of Abovenet (apparently, they don't use
multi-homing). Abovenet had connectivity problems that week-end
(Core3.sjc2 rebooting) and that may explain the trouble.

May be ICANN should multi-home to avoid depending on one
provider. But, as someone said, www.icann.org is not critical for the
daily operation of the Internet.

If they are single homed, what's this about:

   ICANN (ASN-L-ROOT)
   4676 Admiralty Way, Suite 330
   Marina Del Rey, California 90292
   US

   Autonomous System Name: L-ROOT
   Autonomous System Number: 20144

   Coordinator:
      Crain, John (JC2032-ARIN) crain@icann.org
      +1 3108239358

   Record last updated on 30-Mar-2001.
   Database last updated on 18-Nov-2001 19:53:59 EDT.

Maybe I'm just being dense - it wouldn't be the first time - but if ICANN
isn't multi homed, or *about* to be multihomed, why are they holding on to
scarce resources like an ASN?

Neither is www.whitehouse.gov or thousands of other sites on the net that
have the sense to use some mechanism to address single point outages.

The only people who claim ICANN has a goal of "running the Internet" are people outside of ICANN. ICANN, itself, has a far more modest and mundane goal.

That you would insist that an organization with non-critical hosts incur the cost of an Akamai service, or the like, rather than simply enjoy basic co-location with an ISP that has a good track record, bespeaks of the mission creep people keep trying to impose on ICANN.

d/

[snip]

> ICANN has nothing to do with routing, the world wide web, or lots
> of other things on the Internet.

No but the fact that sh ip bgp 192.0.0.0 results in "Network not in table"
from every publicly available route server I've checked has everything to
do with routing thus it would seem that this is somewhat operationally on
topic.

This isn't about ICANN per se, but rather about a good chunk of address space
being unreachable for some unknown reason that has very little to do with
"normal maintenance."

% whois -h whois.arin.net NET-ICANN
ICANN
c/o Internet Assigned Numbers Authority (NET-ICANN)
   4676 Admiralty Way, Suite 330
   Marina del Rey, CA 90292-6695
   US

   Netname: ICANN
   Netblock: 192.0.32.0 - 192.0.47.255
[snip]

% host www.icann.org
www.icann.org has address 192.0.34.65

192.0.0.0/24 != 192.0.32.0/20

Sorry to throw cold water of facts onto politics, but if they were
publishing bad DNS info, they stopped.

Actions speak louder than words Dave.

My point exactly.

d/

ps. and no, folks, this exchange will not continue.

That was a typo on my part. It didn't change the situation and much more
thatn 192.0.32.0/24 was affected.

Given that this is ASN-L-ROOT, I'd imagine it's because
while ICANN may be single homed to MFN, L.ROOT-SERVERS.NET isn't.
I'm currently seeing a 2914 path and I'm willing to bet that
there's more that is either down or used solely as backup transit.

  Is it so hard to check before posting?

  --msa

Heh, tell that to some of my users.

Try telling them this: "Calm Down. It's only ones and zeros...."

Wish I knew who said that. :wink: