[AusNOG] My Predictions for the ISP Industry

Terry Sweetser terry+AusNOG at skymesh.net.au
Wed Mar 14 10:10:47 EST 2012


Noggers,

My main 3 problems are:

[1] Sat NSP#1 has *no* ipv6 migration policy or plan in place,
[2] Sat NSP#2 has been told ipv6 migration is "too expensive" by their MSP,
[3] Fibre NSP has not rolled out support for DSCP Mapping on ipv6.

SkyMesh has an ipv6 /32, "the only address space we'll ever need this
century" and the SkyMesh backbone has ipv6 deployed.

SkyMesh has already chosen DS Lite when problems [1], [2] and [3] go away.

Now, [3] goes away some time this year, [2] maybe in 2014/2015 (too
late) and [1] will just keep wasting time and money, as usual.

(In fairness, [3] can do ipv6 now, over IPoE or PPPoE, dscp support
disappears under all pppoe and/or ipv6.)

Right now, I'll looking into getting 2+ old Cisco G2's and running them
exclusively doing NAT, and getting back a /17 from [1].

That /17 will last only 6 months on the new [2] network ... as they
force SkyMesh to use a /30 per client.

Terence C. Sweetser
Engineering Manager

SkyMesh Pty Ltd
Licensed Telecommunications Carrier
ABN 62 113 609 439
37 Baxter Street
FORTITUDE VALLEY Q 4006


On 14/03/12 00:03, Skeeve Stevens wrote:
> So, as a continuation to my previous reply, here are my predictions on
> what will most likely happen in this region (not just Australia).
>
> - APNIC runs out of IPv4 (We are here)
> - ISPs start to run out of IPv4 (We're almost here)
> - ISPs try to buy IPv4 from those who have it spare *(Starting to
> happen now - March 2012 - mark the date)*
> - Cost of IPv4 starts to rocket like gold (from now onwards)
> - At some point they realise it will cost less to build solutions to
> extend IPv4 - such as Carrier Grade NAT (CGN)/Large Scale NAT
> (LSN) technologies - all of which are evil and very broken - but there
> is no choice
> - ISPs will demand their vendors come up with CGN/LSN solutions -
> Vendors will love it and charge a premium for broken products.  Some
> may even come up with standards breaking solutions which do actually work.
> - ISPs will implement broken CGN/LSN solutions which will prolong the
> life of their IPv4 pools
> - ISPs will start to charge end users to have a real IPv4 address - to
> pay for the acquisition of more
> - Many end-users will end up with reduced functionality until they can
> move to IPv6
> - Some ISPs will start to use address space that does not belong to
> them.  People start to scramble to keep their networks going
> - Some ISPs will start to implement native IPv6 long after others to
> catch up but will find their customers have gone elsewhere
> - Users who don't like the broken-ness will replace their old routers
> with IPv6 ones and go to ISPs who have native IPv6
>
> In the end, the end users don't care about IP addressing (Generalising
> - 98% don't). They want to type a name in, and get what they want.  It
> doesn't matter to them how it works. Businesses may care a bit more,
> but unlikely.  There is NO killer app for IPv6 yet and there is
> unlikely to be one for quite sometime.
>
> To all the ISPs out there who are waiting on a business case for IPv6
> - you are fools... I am sorry, but it is as simple as that.  There
> will NEVER EVER be any valid business care that matters "right now",
> the time when you need to be thinking about it - except one - Staying
> in business.
>
> But, it will take most ISPs a year or two to fully integrate IPv6 into
> their networks, and those who haven't started doing it yet, might as
> well be planning to shut down their businesses because in the next
> year or two, it will be too late... smart ISPs like Internode will
> have already done it (but probably still have work to do).  Telstra
> has been spending millions on it for years now to get ready for it -
> but they are in a different position to everyone else - they HAVE the
> money to do whatever solutions they need to do to survive.
>
> At some point there is going to be many people selling snake oil,
> promising to make things better, selling dodgy solutions that many
> will buy into and end up in a mess and go bust.
>
> At eintellego we've been implementing IPv6 solutions for our customers
> for 2 years now, but even many of them have only partially embraced
> IPv6 - no matter what we tell them.  It is our experience that it
> takes a good year to fully get up to speed, even in a small ISPs
> network.  Things that need to be considered are:
>
> - Transit (Native IPv6)
> - Core network (Routers, Switches, Firewalls, Appliances)
> - Server Operating Systems (Windows, Linux, VMware, etc)
> - Application awareness (DNS, Mail, Web, Authentication, etc etc)
> - Operational Support Systems (Provisioning, Accounting, Network
> Management, etc)
> - Device support - KVM, PDU, Printers, VoIP, Cameras, Building
> Management, NTU's, blah blah blah
> - Access technologies - Dial (its still used!), xDSL, Ethernet
> (easiest!), Wireless, Mobile, NBN (can it even do v6 yet?!)
> - Redeveloping user facing systems (DNS, web, etc etc)
> - Training staff - helpdesk, provisioning staff, engineers, etc
>
> All of which has to be paid for - by management who has no idea what
> on earth is going on and why they need to pay for it.  They will ask
> for the business case; the techies will look confused as they try to
> justify the coming armageddon to the bean counters..... 
>
> Even the little tier 3 reseller ISPs have most of the above list that
> they need to be seriously considering.
>
> Many of the tiny-small ISPs I know about need a large part of
> infrastructure replaced - and they are the least likely to be able to
> afford it.
>
> We've had no more IPv4 available from APNIC for over a year now.
>  There are still only a few carriers in Australia who are offering
> IPv6 - some countries still don't even have it at all - especially in
> the Pacific region.
>
> What is the easy way out?  There isn't any.  Some of my fellow
> directors at The Internet Society of Australia are still quite
> passionate about IPv6 evangelism... me... I'm rather meh (whatever)
> about it these days.  The way I view things is that IPv6 is going to
> come, whether anyone likes it or not, and this situation is akin to
> telling people that they should have safe sex or shouldn't smoke.
>  Many people - those who think they know better - they need to
> experience what is going to happen for themselves.
>
> Speaking on behalf of me, an industry veteran, the CEO of eintellego,
> a Director of ISOC-AU and the ISP Liaison, community member, and
> whatever other roles I do.
>
>
> *Skeeve Stevens, CEO*
>
> eintellego Pty Ltd
> skeeve at eintellego.net
> <mailto:skeeve at eintellego.net> ; www.eintellego.net
> <http://www.eintellego.net>
>
> Phone: 1300 753 383 ; Fax: (+612) 8572 9954
>
> Cell +61 (0)414 753 383 ; skype://skeeve
>
> facebook.com/eintellego <http://facebook.com/eintellego>
>
> twitter.com/networkceoau
> <http://twitter.com/networkceoau> ; www.linkedin.com/in/skeeve
> <http://www.linkedin.com/in/skeeve>
>
> PO Box 7726, Baulkham Hills, NSW 1755 Australia
>
>
> The Experts Who The Experts Call
> Juniper - Cisco – Brocade - IBM
>
>
> _______________________________________________
> AusNOG mailing list
> AusNOG at lists.ausnog.net
> http://lists.ausnog.net/mailman/listinfo/ausnog

-- 
Terence C. Sweetser
Engineering Manager

SkyMesh Pty Ltd
Licensed Telecommunications Carrier
ABN 62 113 609 439
37 Baxter Street
FORTITUDE VALLEY Q 4006

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ausnog.net/pipermail/ausnog/attachments/20120314/47c47be1/attachment.html>


More information about the AusNOG mailing list