Re: [DNA] Definition of "Link Up" and "Link Down" events?

[Bernard Aboba <aboba@internaut.com>]

> > "High probability" is different from "the link layer is in a state where
> > IP packet transmission is not possible".  The former depends on link
> > quality, whereas the latter depends on the link layer state machine.  As a
> > result, the former is likely to be highly transient, while the latter is
> > likely to be less so.
>
> Ok,  so is ~(IP packet transmission not possible) == (IP packet transmission
> possible)? From your previous emails, I have the impression that you think
> this is an oversimplification, is that right?

IP packet transmission "possible" is an indication that the link
layer will accept an IP packet for transmission.  Whether it actually
arrives depends on the link quality.

Where we have perhaps over-simplified is in assuming that links are either
"up" or "down" in terms of their frame loss (e.g. either low loss or high
loss in both directions).  In some cases (e.g. outdoor mesh networks) the
data shows that links can be in intermediate states as well as exhibiting
considerable asymmetry.

However, those over-simplifications won't affect DNA as long as our definition
of "Link Up" is not dependent on frame loss.  However, it is not
clear to me that other "link indications" are as robust.  For
example, are indications such as "Link Going Down" dependent on frame
loss, are they indicative of a decision to bring the link down?



>
> > > >Sure, but I'd suggest that we should be clear that this additional info
> is
> > > >not a determinant of "Link Up" and "Link Down" events.
> > >
> > > Hmm, what about the case where 802.1x authentication must complete prior
> to
> > > "Link Up"? That has nothing whatsoever to do with the quality of the
> link
> > > from a frame transmission standpoint?
> >
> > Since EAP supports retransmission, and on 802.11 802.1X frames are sent to
> > a unicast address, it is possible for 802.1X authentication to complete
> > even in situations where there is high frame loss.  So it is possible for
> > an 802.11i link to reach the "Link Up" state even though link quality is
> > poor.
> >
>
> >From the IP stack's standpoint, "Link Up" on a link where 802.1x is required
> means: "802.1x is complete and the controlled port is now open for traffic".
> Even if the link is of stellar quality, the IP stack can't get any packets
> out prior to that. If the state is: "802.1x is complete and has failed, the
> controlled port is closed" means the link is still down. This particular
> state transition has nothing to do with the link quality, from the IP
> stack's viewpoint, even though the transition of the state machine regarding
>  whether frames can successfully get across the link does affect whether or
> not 802.1x completes.
>
> If 802.1x is not required on the link, then the transition of interest to
> the IP stack is controlled by the link quality alone.
>
> So my point was, link quality is not the only factor that influences whether
> the driver can report "Link Up" to the IP stack. Sorry for being not clear.
>
>             jak
>
>