Am 05.02.2011 19:27, schrieb /dev/rob0:

> Perhaps so. And on average the TTL for cache hits will be roughly 
> half what you'd have when doing the recursion yourself.

but it's only ONE request and your are not the onnyl one
refreshing his caches

>>> * This cached requests are much faster and fewer as full recursion
> 
> Faster: possibly a bit. But if you have one locally-connected 
> recursive nameserver serving your network, its cache hits will be 
> quite a bit faster than any external nameserver.

And my two nameservers in the network are cahcing the ansers
from the forwarder too - so what is the problem?

> Fewer: definitely not, because as above, on average you will see 
> roughly half the TTL.

does not matter, se above

>>> * It reduces the load of the root-Servers
> 
> Maybe a little bit, yes. But most of those records have very long 
> TTL, which is designed to reduce the load. Don't go predicting gloom 
> and doom and the collapse of the Internet: it won't happen. The 
> global root is able to handle it.

there are not only root-servers involved

It makes a differnce fpr the tld-servers and finally for
the authoritative servers - you have only the half
of all what happens in your mind

> Yes, a local caching nameserver using external forwarders might 
> present a few benefits. It's also potentially subject to cache 
> poisoning, and that being a cache which you do not control.

nothing is perfect

> Recently I knew a NetworkSolutions-registered domain which expired 
> because of the registrant's oversight. NetSol redirects NS for 
> expired domains to some ns1/ns2.pendingrenewaldeletion.com. They set 
> a very long TTL for these NS records, at least 3 days, IIRC.

cismet

> ns1.pendingrenewaldeletion.com. and its partner have a wildcard A 
> record for zone "." Try it, query any name you can think of:
> 
>   this.is.fun.but.TLD.does.not.exist. 7200 IN A       209.62.105.19

so these are idiots

> To bring this almost back on topic, use of forwarders for zone "." 
> will mean, in many if not most cases, that queries to Spamhaus are 
> blocked, as Charles pointed out. Try it:
>     $ dig 2.0.0.127.zen.spamhaus.org. any @4.2.2.2 # Level3
>     $ dig 2.0.0.127.zen.spamhaus.org. any @8.8.4.4 # Google
> 
> Maybe *your* forwarder is not blocked yet. Have you seen what happens 
> when your Zen queries suddenly stop working?

spmahus *loool*

are there really peopole outside using them?
sorry, but since they blocked "nic.at", the registry for
austrian domains because spamhaus meant nic.at have to delete
a domain form where comes spam they are persona non grata

> A few years back I had 
> MX on Comcast cable, and their nameservers were blocked without 
> warning, while I was gone for a few days. I returned to a massive 
> spam mess, of course.

http://www.barracudanetworks.com/ns/?L=de

Get a (virtual) appliance, their filters are fine and
their one blacklist does never block anybody

> Comcast, another good BAD example: they hijack all NXDOMAIN, 
> returning a wildcard A record. Wham, suddenly your DNSBL queries 
> return positive, and you're blocking everything!

again: everybody who configures a namesever this way is stoopid

> Will your forwarder do something like this to you?

if my ISP would touch any dns-answer he would not be my ISP

> Will they notify you in advance of if if they do?

no, but i would notify them!

> Will they warn you before Spamhaus starts blocking them?

again: it is braindead using blacklists from people
which are blocking a tld-registry without understand
nic.at can not delete a domain because some guy say
"soma is coming from there"

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to