Kacheong Poon wrote:
As part of the NWAM project, we are investigating the support
of IPv4 LLA in Solaris. One big issue with IPv4 LLA support
is the multi-homing nature of most Solaris machine.
Most ?
Most laptops and desktops (which IMO is where LLA is most valuable)
generally aren't multi homed.
To make the issue more complicated, the RFC mentions that a
system supporting LLA but with routable address should be able
to communicate with another system using LLA if both systems
have interfaces on the same "link." This means that LLA capable
systems can talk to each other using either routable address or
LLA. The following is a simple case,
Eeek! This just feels wrong to me for some reason.
IPv6 used to have a "similar" behavior as the above, the on-link
assumption. But it was thought to be bad and has been removed.
So I would recommend that despite what the RFC says this is not
something worth implementing.
Here are some questions.
1. Is LLA support essential to Solaris networking that we need
to include it and make it work?
It would be very helpful for laptops.
It would also be very very helpful for machines running TX since an LLA
could be used to the global zone <-> local zone X11 traffic.
Especially it said TX machine is a laptop.
2. If we really need to support it, is it OK for it to "work
in some cases but not all?" The RFC does not have a solution
to make it work with multiple interfaces.
I think so.
The way I'd expect this to work is like this:
NWAM is setup without any static addresses.
It tries DHCP on all interfaces and gets no leases.
It uses the "first"/"preferred" interface and does LLA on that.
If you are using LLA only a single interface is brought up by NWAM.
3. Is it OK to have certain limitations, such as only supporting
LLA in one interface? Or not support the communication
between LLA and routable address? Or ... This is to allow
us to have a more "consistent" failure mode.
I think some LLA support is better than nothing. I think the LLA
routing would just cause confusion and some very strange routing
problems for humans.
--
Darren J Moffat
_______________________________________________
networking-discuss mailing list
[email protected]
