Hi Albert,           [thanks for trying to answer the question of uses]

> Manfredi, Albert E wrote:
>> -----Original Message-----
>> From: Arnaud Ebalard [mailto:[EMAIL PROTECTED] 
>> 
>>> Can you please give me in one or two sentences (i.e.
>>> little effort) the specific purpose/use those people have.
>>> This is the only thing i keep asking for on the list and
>>> no one has still answered with specific use.
>> 
>> Testing paths between routers that would otherwise be unused in the
>> current topology

Certainly my english but i don't understand that argument. Are you
stating that source routing can be useful to create traffic on link that
are not used otherwise? It can't be that. Can you clarify that point?

>>, creating separate paths between dual-homed hosts connected
>> redundantly to a single network infrastructure

If you consider end-host connected to a specific "Home Network"
infrastructure, this already exists in MIPv6. It works great, even
with IPsec protection. Last evolution (available under Linux for
instance) allows transparent movement of IPsec tunnel endpoints when a
change from a link to another occurs. As a positive side effect, you
even get address stability.

The point here is that if you only want dual-homed hosts connected to
another network you control, you don't need any help from the core, you
only need some help in _your_ network. This function is provided by the
Home Agent in MIPv6 or by your correspondent. I don't state it is the
solution to all problems on earth, but if you don't want to select
specific weird pathes in the core, i.e. only need to get the best
routing path from you current source address to the destination, source
routing in the network is unneeded.

>> routing certain types of signals on separate paths from other types
>> of signal (e.g. multimedia streams routed differently from text or
>> file transfers).

Can you elaborate on that. I imagine that the path in the core will
quite be the same for all flows (modulation below) and that the penalty
occurs mainly between the different connection means you consider. If
you take satcom, wifi or basic ethernet, you get different prefixes
(hence different source addresses) that allow you to select how your
packet will leave the source (this obviously works in both directions,
if you select the right destination). 

[Correct me where i get things wrong] Now, if you expect differences in
the routing at some point (where source routing should occur), this
means that the operator is able and is willing to provide different QoS
for your flows. There is no way you won't be charged for that. This 
basically means that you have some commercial relationship with it and
also basically exclude providers you don't know directly. IMHO, this
also implies that those operators are near your source (your provider ?)
and/or near your destination. Based on all of that, there are probably
other ways to get QoS applied to your flows, and earlier in the network
than expecting source routing to provide that: asking for different
prefixes, for specific guarantees on specific flows based on protocol,
flow label, destination, ...

As a side note, adding MTU, parsing and soft processing is probably not
the best way to get the best results for your flows.


Jeroen Massar <[EMAIL PROTECTED]> writes:

> On your "own" (thus in some way controlled by you) network, or on the
> public Internet where you have totally no control over those networks,
> and mostly you will not get control as those ASN's are somebody elses?
>
> If your own, then the security question becomes lighter as a special tag
> could be added to packets whereby routers know that these packets should
> be allowed to do these tricks, otherwise it won't work as you don't own
> those people's routers.
>
> And is this being really used or only for research/testing? [y/n]

Couldn't agree more.

Regards,

a+

--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to