Test your connectivity for World IPv6 Day

Rémi Després remi.despres at free.fr
Tue Jun 7 09:27:06 CEST 2011


Hi Tore,

More comments in line.


Le 7 juin 2011 à 00:20, Tore Anderson a écrit :

> * Rémi Després
> 
>> Fixing known problems of PMTU discovery before trying to do better
>> than 1280 is an alternative that I recommend, especially for servers
>> that have important contents to deliver.
> 
> It's quite hard to discover the blackholes (and by extension fix them)
> if you're defaulting to 1280.

Which blackholes?
- With 1280, there isn't any related to MTU's. Right?
- Non-MTU-related blackholes shouldn't be harder to detect with 1280 assumed to be the default PMTU.
- Creating blackholes to better detect them seems strange to me.  

> And if 1280 becomes the de-facto standard,
> the blackholes are free to proliferate.

Again, which blackholes?

> That said, it might be a
> sensible way to minimise the set of potential problems when first
> publishing AAAAs.

Indeed.
That's the point.

> For what it's worth we changed www.vg.no from 1280 to 1500 a couple of
> months ago.

Unfortunate IMHO, if it applies to servers that all IPv6-capable clients should be able to work with.

> No complaints so far.

So what?
Promoting IPv6 shouldn't IMHO consist in waiting for consciously created complaint causes.
Many users may chose to disable IPv6 rather than bothering to complain (to whom? with which diagnostic? ...).

> In any case, the few users that have
> HE/SixXS/etc. tunnels can take care of themselves. If it breaks, they
> get to keep both parts.

Too bad for IPv6!

> For real deployments, on the other hand ... well, I'm hoping no serious
> ISP or content provider will willingly put their end users or web sites
> behind MTU-impaired links or tunnels. Recipe for disaster if you ask me.

What do you mean, precisely, by "MTU-impaired links or tunnels"?

Regards,
RD


> 
> -- 
> Tore Anderson
> Redpill Linpro AS - http://www.redpill-linpro.com/
> Tel: +47 21 54 41 27



More information about the ipv6-ops mailing list