[aklug] Re: DNS Root Failure?

On 02/12/11 23:52, barsalou wrote:
> This might help answer the question:
>
> http://www.root-servers.org/map/
>
> and this talks about your specific question
>
> http://en.wikipedia.org/wiki/Root_nameserver
>
> Mike B.
>
> Quoting Christopher Howard <christopher.howard@frigidcode.com>:
>
>> Okay, humour me a bit, but I'm curious:
>>
>>> From what I understand, DNS works like so:
>> 1. DNS Client communicates requests to recursive DNS server
>> 2. Recursive DNS server contacts root DNS server to get authoritative
>> list for top-level tld (com, edu, org, etc)
>> 3. Recursive DNS server gets list from authoritative server for that
>> domain, to get list for subdomain.
>> ...
>> And so on and so on until an IP address for the host can be returned to
>> the client. Though most of these steps can be skipped if the request is
>> cached and the TTL has not expired.
>>
>> Now, lets say the root server in Anchorage, which as far as I know is
>> the only one in Alaska, suddenly goes kaput. Are all the other DNS
>> servers in Alaska programmed to automatically send their requests to
>> another root server out-of-state? Or is the delay while server operator
>> figure out what the heck is going on and change this manually? Can the
>> clients all continue to operate normally until the TTL expires?
>>

>

Could you explain one point regarding the "anycast" functionality: Do
all the servers in an anycast cluster actually share the same IP
address? Or does a packet go to a separate "anycast" address and then
get re-addressed to go the closest server?

-- 
frigidcode.com
theologa.indicium.us
---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.