[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: etherconf or ifupdown problem with subnets

Michael Stone wrote:

> >> >If you're arguing that the introduction of CIDR means that we should use
> >> >smaller defaults than the old ranges, then a /30 should be the default.
> >> 
> >> No! 
> >> 
> >Why not?
> Because I think /30 is far more uncommon than /24. 
You'd be surprised.

Nearly all the customer netblock allocations we give to customers are
/28, /29 or /30.  Getting a /24 off RIPE for a single customer is for
all intents and purposes, impossible.

> >> >I can't see any argument for any other default range, /24 is far too big
> >> >for any real networks to be "useful".
> >> 
> >> Curious. I'm surrounded by /24's here. 
> >> 
> >Are they RFC1918 space, or CIDR space?
> CIDR. I don't know what gives you the idea that a 2-machine subnet is
> ideal.
The most sensible default is the lowest bound, surely?  Less likely to
cause problems?

/24 is an arbitrarily *large* netblock.

> >However, if you don't give a netmask either, I don't see why the old
> >classful defaults are wrong.
> Because they are utterly meaningless on the modern internet. If someone
> doesn't specify a netmask in the install process we should either abort
> or try a few other tricks (like icmp netmask request) rather than using
> a meaningless default.
Perhaps, but you either have to error, or provide some kind of sensible
default.  I do not believe that /24 is sensible across the board.

Scott James Remnant     Have you ever, ever felt like this?  Had strange
http://netsplit.com/      things happen?  Are you going round the twist?

Attachment: signature.asc
Description: This is a digitally signed message part

Reply to: