Quick question, does there exist a practice of charging customer for IP address blocks used? My theory is that the first Class C is included with the service, but I’m wondering what happens when the customer wants 2,3,4 or more?
Shane
Shane,
There is a practice on that (At least here.). Generally we provide a Class C to our customers at no additional charge, but we have been charging recently for the use of additional blocks. After all, we have to pay those charges to ARIN, and we do need to defer those costs down to the customer if they are going to use a chunk of the address space. At some point we’ll need to get more, and that only increases are costs. Gone are the days when the carrier’s eat all the side costs.
Derek
MessageShane,
The best practice is to follow the ARIN guidelines. This will make it much
easier for you to get your next block of address space. That means:
- Slow start - issue folks what they can justify, not a /24.
- Issue more space upon request, provided that justification is given
- Multihomed customers require no justification for a /24
- Do not issue more than a /21 to a customer. At that point, they can do
directly to the RIR.
Charging is up to you - you are really just charging for your own services
in administering the address space, and perhaps passing through the cost
from ARIN. Most folks do not charge for IP space, and it's never something
I've been personally comfortable with.
- Daniel Golding
Why in this day and age, 9 years after the invention of CIDR, are we still
refering to "class C"'s?
Haha. Mighty good question. No good answer.
Derek
From: Richard A Steenbergen [mailto:ras@e-gerbil.net]
Sent: Thursday, September 05, 2002 1:48 PM
To: Derek Samford
Cc: 'Owens, Shane (EPIK.ORL)'; nanog@nanog.org
Subject: Re: IP address fee??> Shane,
> There is a practice on that (At least here.). Generally
we
> provide a Class C to our customers at no additional charge, but we
have
Why in this day and age, 9 years after the invention of CIDR, are we
still
refering to "class C"'s?
--
Richard A Steenbergen <ras@e-gerbil.net>
PGP Key ID: 0x138EA177 (67 29 D7 BC E8 18 3E DA B2 46 B3 D8 14 36 FE
B6)
Why in this day and age, 9 years after the invention of CIDR, are we still
refering to "class C"'s?
Because we used up "class B"s?
Alex
Possibly because that is what they are still teaching them as in
school?
Seriously... I'm not sure that the teachers I had for networking and
systems admin had ever heard of CIDR.
The textbooks hadn't. It was a nice bump in the learning curve when I
hit the real world.
*********** REPLY SEPARATOR ***********
Shane,
There is a practice on that (At least here.). Generally
we
provide a Class C to our customers at no additional charge, but we
have
Why in this day and age, 9 years after the invention of CIDR, are we
still
refering to "class C"'s?
--
Richard A Steenbergen <ras@e-gerbil.net>
PGP Key ID: 0x138EA177 (67 29 D7 BC E8 18 3E DA B2 46 B3 D8 14 36 FE
B6)
I've never seen a text book which had any relevance to modern networking
which didn't cover CIDR.
Perhaps if we all made a conscious effort to avoid using the term, new
people who are learning from the examples they see around them would stop
picking up on it as "how things work".
History is nice, but not knowing when to give up and move on is just sad.
At least as importantly, why do 254 addresses get provided where the
actual need might not warrant that quantity?
Tony
Tony Tauber wrote:
> > Shane,
> > There is a practice on that (At least here.).
> > Generally we provide a Class C to our customers at no
> > additional charge, but we have
>
> Why in this day and age, 9 years after the invention of CIDR, are we
> still refering to "class C"'s?At least as importantly, why do 254 addresses get provided where the
actual need might not warrant that quantity?
Because it's easier to do the reverse DNS? Sorry to contribute to the
general noise, but that answer's close to the truth.
Tony Tauber wrote:
>
>
> > > Shane,
> > > There is a practice on that (At least here.).
> > > Generally we provide a Class C to our customers at no
> > > additional charge, but we have
> >
> > Why in this day and age, 9 years after the invention of CIDR, are we
> > still refering to "class C"'s?
>
> At least as importantly, why do 254 addresses get provided where the
> actual need might not warrant that quantity?Because it's easier to do the reverse DNS? Sorry to contribute to the
general noise, but that answer's close to the truth.
these days you can easily delegate reverse using CIDR with BIND ...
http://www.faqs.org/rfcs/rfc2317.html
-chris
Thus spake "Richard A Steenbergen" <ras@e-gerbil.net>
>
> Possibly because that is what they are still teaching them as in
> school?
>
> Seriously... I'm not sure that the teachers I had for networking and
> systems admin had ever heard of CIDR.
>
> The textbooks hadn't. It was a nice bump in the learning curve when I
> hit the real world.I've never seen a text book which had any relevance to modern networking
which didn't cover CIDR.
Sadly, most texts I've read, and certainly all the current courseware I've
looked at, still teach classful addressing and subnetting as the primary method
with a sidebar on CIDR as the "new" method.
Perhaps if we all made a conscious effort to avoid using the term, new
people who are learning from the examples they see around them would stop
picking up on it as "how things work".History is nice, but not knowing when to give up and move on is just sad.
The term class C sticks because it's so useful; you'll note that class [AB]
aren't used much colloquially. This is how English evolves.
S
> At least as importantly, why do 254 addresses get provided where the
> actual need might not warrant that quantity?Because it's easier to do the reverse DNS? Sorry to contribute to the
general noise, but that answer's close to the truth.
http://www.faqs.org/rfcs/rfc2317.html
Easier maybe... But with classless delegation of IN-ADDR.ARPA
this should not be an issue any longer.
Thus spake "Tony Tauber" <ttauber@genuity.net>
> > Shane,
> > There is a practice on that (At least here.).
> > Generally we provide a Class C to our customers at no
> > additional charge, but we have
>
> Why in this day and age, 9 years after the invention of CIDR, are we
> still refering to "class C"'s?At least as importantly, why do 254 addresses get provided where the
actual need might not warrant that quantity?
Because ARIN doesn't verify end-users actually need all the addresses SWIPed to
them, and the more addresses an ISP SWIPs, the lower the cost per address and
the easier it is to get more.
There is at least one provider which assigns a /23 to each customer circuit even
if the customer has their own IP space. I was unable to get a reasonable
explanation other than "policy".
S
Shane:
I think an important question would be what level of service are they
buying. Including 255 address with a T3 would be very reasonable, less so
with a T1, not very reasonable with DSL, and ridiculous with a dial-up
account.
There is generally a charge for additional IPs with DSL (or co-location)
services because it is so cheap. You don't usually find this with T1 and
above. But everyone's pricing is different.
about 2 years ago, interviewing fresh graduates for jobs, i found that they
were still being taught classful networking at many colleges.
it was a fairly depresssing discovery.
richard
I think an important question would be what level of service are they
buying. Including 255 address with a T3 would be very reasonable, less so
with a T1, not very reasonable with DSL, and ridiculous with a dial-up
account.
I must be missing something. Why would you expect need for IP addresses to
correlate with bandwidth? I can see a company buying a DS3 for a single
web/application server or load balancer. I can see an apartment building with
120 network jacks getting a T1.
It may make business sense to bundle more 'free IPs' with packages that cost
more money. But the actual allocation must be based upon demonstrated need.
Read your agreement with ARIN.
DS
Current CCNA Exam Description:
http://www.cisco.com/warp/public/10/wwtraining/certprog/testing/current_exams/640-607.html
I submit that the comonly used definition of "Class C" has changed from
"An address in the class C range" to "a block of addresses aligned on a
/24 boundary".
My guess of the real underlying reason is that saying "I need a full class
C" or "I need a block of [4,8,16,32,64] addresses" seems to be a lot
easier to say in a clear fashion over the phone or in person than "I need
a slash-twentyfour".
- Forrest W. Christian (forrestc@imach.com) AC7DE