Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [EFM] OAM loop back / echo server function


I would like to step back and look at the "Why's" of loopback / echo as well as
some of the "How's."

1. The SP needs to know that a link is up and functioning correctly up to a
demarc point. This demarc point would ideally be in the PHY (or MAC) and not in
a higher layer device.

2. Remote OAM fulfills that requirement in almost all cases. The only exception
would be a situation in which there is a data dependent or data rate dependent
fault on the line.

3. Depending on the PHYs chosen (for each of the three media), this situation
may be sufficiently rare enough to ignore.

4. Assuming that data dependent fault finding must be addressed, we need a
proposed mechanisms that specifically solves this problem.

5. If it can be shown that the problem can be solved acceptably using (for
instance) LLC test frames, then we do not need to specify this in 802.3ah. We
could consider making a recommendation or guidance that EFM system vendors
include support for certain other standards. Ultimately the system vendors must
make systems that incorporate all of the standards and functions that their
customers require - 802.3ah cannot change the world - it can only enable the

6. If there is no acceptable solution available then we must propose one. If it
is PHY layer, then we can define & specify it. If it is higher layer then we may
need to make a recommendation to somebody else to solve it.

7. In any situation where we loopback data, we must ensure that the data does
not leak out to pollute the behavior at higher layers.