[e2e] Addressing format...network and mask

Joe Touch touch at ISI.EDU
Thu Dec 4 12:41:47 PST 2003

Alok Dube wrote:

>>Alok Dube wrote:
>>>>>why do I have to stick to "bit boundaries"?
>>>>>they are meant for routers to lookup faster
>>>>They are the definition of Internet routes as well.
>>>yes....but its the prefix we care about...the masks are just there to
>>>make the FT guys happy.
>>The prefix is determined by the mask.
> i guess you mean the "prefix range is determined by the mask"
> now i say that i pass the range and not the mask...(the mask field in the
> protocol update gives the last Ip of that segement and not the mask).

Yes, as noted before, they are equivalent when bit-boundary aligned.

>>>so given that as working, i could still pass start IP and end IP of
>>>the range.
>>>its still backward compatible...incase one was to propogate subnets,
>>>one could still use the 1st and last made from the mask and build the
>>You can't send your routes outside your subsystem, and they would need
>>to propagate back through to the core Internet interchanges.
> so i set a bit in the protocol updates which says if whats being passed is
> the range or the mask...

And change the forwarding algorithm in the core of the Internet. This 
isn't a local change, unless you advertise bitmasks of aggregates of 
your ranges to the outside world, or limit your system to ranges that 
are equivalent to individual bitmasks.


More information about the end2end-interest mailing list