[rpd] Re: Factors affecting in-region utilization - way forward?

Sat Jul 19 14:24:00 UTC 2014

well

On 19 July 2014 16:48, Mukom Akong T. <mukom.tamon at gmail.com> wrote:

> *** speaking as mysefl i.e. no hats ***
>
>
> On Sat, Jul 19, 2014 at 3:31 PM, Seun Ojedeji <seun.ojedeji at gmail.com>
> wrote:
>>
>>
>> "changing the subject line to allow for proper followup"
>> On Sat, Jul 19, 2014 at 11:36 AM, Noah Maina <mainanoa at gmail.com> wrote:
>>
>>> Which is why I was arguing that LIRs shouldn't seat on the space
>>> allocated but use it extensively. If customers want ips give it to them.
>>>
>> Lots of LIR don't practice this, they kind-of encourage NATing by giving
>> their users a few public IP and when the user ask for more, they attach
>> a fee to it. A typical ISP gives max of a /28 for free and everything
>> beyond attracts an extra fee. You can imagine a typical institution with
>> over 40,000 users assigned a /28 for use.
>>
>
>
> I'd say a good engineer would do the math of "get my own space from
> AFRINIC" vs "pay for more IPs from the provider". A very important point
> is that moment when the contract is being negotiated - engineers should
> give input in to this. I'd say two important points need to be made:
>
> a) The institutions is paying for a circuit and transit
> b) The institution should connect via BGP if they so wish do
>
> This should counter that default position from some providers where a
> connection means a CPE with a /28 and a NAT configuration.
>
>
>
Service Providers - typically have different internet service packages
which are mostly delivered using different access technologies to target
different markets.

For example a service provider may have a package which assigns /29 IPv4
addresses to a 2048 / 512 kbps down/up shared internet connection delivered
over ADSL. Thats normal :) because such services are designed for home
users. They assume the service will be used by not more than 4 - 10
connected workstations. This is also mostly programmed with the
provisioning systems and seen more as a business policy or rule.

If an institution should go for such a service they should not expect a /24
IPv4 or higher to be assigned to them.

It only becomes an issue when institutions purchase dedicated circuits and
they are not assigned enough IPv4 (e.g. /24). It will surprise you that
service providers mostly assign /24 to their corporate customers but these
customers only use about 5 - 8 IPs and just NAT the rest of their network
even though the service provider routes the /24 to their demarcation point.

Lesson: most internet users are only concerned about being connected and
how fast their internet connection is. Only a few care to know whether they
are NATed or have a direct transparent IP connection. This is universal.

The bitter truth is the chunk of the IPv4 addresses in general are consumed
by data centers and hosting companies worldwide. Lets think about how to
get more of such consumers established in our region and stop brooding over
IPs and double standard policies.

> As for Saun's experience that is outa this world. ..eish...that provider
>>> sounds old school but thats wats up! !!
>>>
>> You can say that again, it may shock you to know that the provide is a
>> major ISP in the continent!
>>
>
>
> Use the power of community to name and shame. I'm sure your institution is
> not the only one at the receiving end of such policy,  when multiple client
>  institutions collectively demanded the proper thing am sure they'll take
> notice.  If the university is part of the BWC in Nigeria, that is a good
> platform.
>
>
> I've also seen a REN try to stop its member university from having and
> advertising their own space. Fortunately, I did tell them that wasn't the
> purpose of a REN and backed up by some universities which had their own
> space and the REN seems to have backed off from that position.
>
>
>
>>  And for the old NAT, its here to stay. ..you wanna kick out out....start
>>> from its inception which is collage and academys where Fundamentals of
>>> Routing are taught lol...NAT/PAT is like a 1st time ideology embibe onto
>>> any aspiring network engineer...then there is the believe that NAT
>>> offers some sort of security lol.
>>>
>>
> :-)  such has to be particularly BAD class. Every Cisco course (ICND,
> CCNA) and I'm sure the AFNOG courses teach Routing First and only later
> teaches NAT as a way to deal with insufficiency of IPv4 addresses.
>
> Sure they are engineers who never had those fundamentals straight and
> somehow ended up running networks - they quickly used NAPT where they
> should have routed (because it is quite easy to set up NAT than configure
> routing)
>
>
>>
>>> Hopefully v6 will kick-it out finally, in generation to come though ;)
>>
>
> The required change from the training perspective is to now say ... "The
> sustainable solution to IPv4 insufficiency is IPv6" as well as have a
> session teaching the contrary consequences of using NAT.
>
>
>>
>> Cheers!
>>
>>>
>>> Cheers.
>>>
>>> Noah
>>> On 19 Jul 2014 09:12, "Andrew Alston" <Andrew.Alston at liquidtelecom.com>
>>> wrote:
>>>
>>>>  *There are quite a number of members who are yet to deploy any subnet
>>>> of the resource allocated to them. There are reasons why this can happen;
>>>> for example, the upstream provider of a member (which I am contact) attach
>>>> a recurring fee to block advertisement. This to me was quite  surprising
>>>> and we are still trying to avoid that cost either through convincing the
>>>> current provider or moving on to another!*
>>>>
>>>>
>>>> *Nevertheless, I don't think there is any member in that category that
>>>> will successfully get additional allocation. On a lighter note, this could
>>>> raise a question on usage and whether a policy is required to "ensure"
>>>> usage ;)*
>>>>
>>>> That’s about the most bizarre thing I’ve read in quite a while… as a
>>>> provider I want my members to advertise every block of space they possibly
>>>> have to me – the more they advertise to me, the more traffic flows via me
>>>> to them, the more transit I sell them.  I really don’t understand the logic
>>>> behind some providers.
>>>>  Let’s face facts, IF a provider has customers that have their own
>>>> space and their own ASN, its in the providers interests to encourage
>>>> as much advertisement as possible.  However, on the converse, it is in a
>>>> providers interests to have customers on space assigned by them and
>>>> not running BGP at all (in the latter case, it means the customer probably
>>>> isn’t multi-homed, and for the customer to churn the customer will have to
>>>> renumber, which can be a MAJOR headache, meaning the customer is far less
>>>> likely to move on).
>>>>
>>>>  It’s an interested dialectic, it is in AfriNIC’s (and hence it could
>>>> be argued the communities) interests to have as many people as possible
>>>> with their own space and their own ASN’s.  However, it is in the interests
>>>> of providers to encourage the uptake of space out of their own blocks
>>>> assigned by AfriNIC and discourage this behaviour.  At the same time,
>>>> what amazes me about Africa and the substantive use of NAT, it is NOT in a
>>>> providers interests to have customers behind NAT, and I wonder if this
>>>> isn’t something we could use to promote the uptake of IPv4 on the
>>>> continent.  The simple reality is, a customer behind NAT can churn in an
>>>> instant, the changes required on the customer side are minimal.  However, a
>>>> customer on a providers space that is NOT running NAT and has the
>>>> space all over the place has to renumber which could be a downtime and OPEX
>>>> intensive activity.  (I’ve actually seen research that shows that non-NAT
>>>> customers are FAR less likely to churn, it reduces the churn rate by
>>>> double digit percentage points).
>>>>
>>>>  Thanks
>>>>
>>>>  Andrew
>>>>
>>>>
>>>> ------------------------------
>>>> DISCLAIMER: This email contains proprietary information some or all of
>>>> which may be legally privileged. It is for the intended recipient only. If
>>>> an addressing or transmission error has misdirected this email, please
>>>> notify the author by replying to this email. If you are not the intended
>>>> recipient, you must not use, disclose, copy, print, or rely on this email.
>>>> We cannot accept liability for any statements made which are clearly the
>>>> sender's own and not expressly made on behalf of this company or one of its
>>>> agents.
>>>>
>>>>
>>>> _______________________________________________
>>>> rpd mailing list
>>>> rpd at afrinic.net
>>>> https://lists.afrinic.net/mailman/listinfo.cgi/rpd
>>>>
>>>>
>>
>>
>> --
>> ------------------------------------------------------------------------
>>
>>
>>
>>
>>
>> *Seun Ojedeji,Federal University Oye-Ekitiweb:
>> http://www.fuoye.edu.ng <http://www.fuoye.edu.ng> Mobile: +2348035233535**alt
>> email: <http://goog_1872880453>seun.ojedeji at fuoye.edu.ng
>> <seun.ojedeji at fuoye.edu.ng>*
>>
>> The key to understanding is humility - my view !
>>
>>
>>
>> _______________________________________________
>> rpd mailing list
>> rpd at afrinic.net
>> https://lists.afrinic.net/mailman/listinfo.cgi/rpd
>>
>>
>
>
> --
>
> Mukom Akong T.
>
> http://about.me/perfexcellence |  twitter: @perfexcellent
>
> ------------------------------------------------------------------------------------------------------------------------------------------
> “When you work, you are the FLUTE through whose lungs the whispering of
> the hours turns to MUSIC" - Kahlil Gibran
>
> -------------------------------------------------------------------------------------------------------------------------------------------
>
> _______________________________________________
> rpd mailing list
> rpd at afrinic.net
> https://lists.afrinic.net/mailman/listinfo.cgi/rpd
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.afrinic.net/pipermail/rpd/attachments/20140719/07e6beec/attachment.html>