Re: DSTM

[Marc Blanchet <Marc.Blanchet@hexago.com>]

-- Thursday, June 24, 2004 11:21:30 +0200 Iljitsch van Beijnum
<iljitsch@muada.com> wrote/a ecrit:

> On 24-jun-04, at 5:37, Margaret Wasserman wrote:
> 
> 
>> I have not personally heard any ISP or Enterprise operators indicate 
>> their intention to run IPv6-only backbones any time soon.

- in which part of the planet you talk to?
- in which market are you talking to?

> 
> Probably because the notion that this could be a viable option hasn't
> entered the brains of the decision makers yet.
> 
> If you consider the complexities of building a large IPv4 network,
> especially if you want to avoid NAT but don't have a lot of address
> space, versus the simplicity of building a similar network using IPv6, it
> makes sense that people will want to run an IPv6-only network and tunnel
> IPv4 over it where needed as soon as the necessary mechanisms
> (IPv4-in-IPv6 tunneling, 
available now. implemented. I can show you a demo.

> IPv6 DNS resolver discovery) are available and
> all the hardware can handle IPv6 in the fast path. (I had a few
> conversations with vendors at the summit thing last week and many still
> do IPv6 in software.)
> 
>> I'm also not sure how/why we need a specific DSTM server at all... How 
>> is DSTM superior to using a DHCPv6 option to get the configuration 
>> information needed to set-up a configured IPv4-in-IPv6 tunnel and then 
>> using DHCPv4 over that tunnel?
> 
> I for one am not in favor of making DHCPv6 a dependency.
> 
> I also have another problem with DSTM as it is: it depends on a single
> gateway. That's not good. We need redundancy.

sorry. don't agree. 
 - many ways to acheive redundancy: anycast discovery, vrrp, etc... 
 - other mechanisms do not have any redundancy: i.e. 6to4 border gateway.

Marc.

> 
> It might be better to create an IPv4-in-IPv6 tunneling mechanism that
> treats the IPv6 network as an IPv4 link layer and run standard IPv4
> protocols such as ARP and DHCP over that. (Similar to 6over4 but the
> other way around.) This should work very well in networks that support
> site-wide multicast. (There are some security issues with rogue DHCPv4
> servers but those can be mitigated by filtering broad/multicasts.)
> 
> However, we may also want to consider the situation where a user is
> roaming on a standard issue IPv6 network far away, and wants to tunnel
> back home. There are already various VPN solutions that do this for IPv4
> over IPv4, though, so hopefully we don't have to reinvent the wheel here.
> 
> One remark about the draft: there are some serious line length issues,
> please fix as this makes some parts very hard to read.
> 



------------------------------------------
Marc Blanchet
Hexago
tel: +1-418-266-5533x225
------------------------------------------
http://www.freenet6.net: IPv6 connectivity
------------------------------------------