[Asterisk-Users] Cannot get two TE410Ps to operate correctly in the same machine

[Rich Adamson]

> [SNIP]
> 
> > What if you have a single port T1/E1 card from digium? No decision to
> > be made really; you're going to sync from the other end if it goes 
> > higher in the hierarchical chain. If the port goes to a box consider
> > lower in the chain, then the distant box should be configured to
> > accept clock sync.
> 
> How about the following topology for a bit of discussion:
> 
> Telco -> T100P -> * Server A -> IAX2 -> * Server B -> T100P -> PRI PBX
> 
> In this case, we have Server A pulling timing from the telco, ensuring 
> they are in Synch. On the other side of the IAX2 link, we have Server B 
> generating timing onto the span for the PBX to listen to. T100P is set to 
> emulate an NI-2 switch.
> 
> Seems to work without much of a problem, as it should! ;)

Certainly nothing wrong with that approach. The key is the telco -> A
is configured properly at both ends of that span so both ends understand
exactly who is syncing to who.  Same with B -> PBX, but that is a 
totally independent engineering decision unrelated to whatever 
was decided upon for telco -> A.

The T100P emulation is irrelavent in any case, as is the IAX2 link. The
IAX2 link does not (and could not) pass accurate span clock sync data 
under any circumstances.

As sort of a side note, if the PBX has other spans heading towards a telco
it might be worthwhile to consider having B sync from the PBX. That all
depends 100% on exactly how the PBX is constructed internally (which
might be difficult to determine). "If" the PBX electronics were designed 
(internally) in such a way as to attempt to sync all T1 interface clocks 
together (regardless of how many T1 cards might be installed), and,
the PBX has a span heading towards a telco, then having B sync from the
PBX is probably a better choice. But, if the PBX offers a config choice
on _each_ span interface to 'accept' sync (or essentially generate sync),
then pick whatever choice you'd like as long as both ends of that span
are subject to the same engineering decision.