>
>
>On Thu, 2003-11-13 at 16:50, Freddi Hansen wrote:
>
>
>>>> >Freddi Hansen wrote:
>>>
>>>
>>>>> >> with boards from Aculab, we are replacing Aculab
boards with Digium
>>>>> >> boards BUT we would need more
>>>>> >> Digium boards IF we could use both Digium and
Aculab cards in the same
>>>>> >> server. The reason being that
>>>>> >> TE410P doesn't support SS7-ISUP so we continue
using only Aculab cards
>>>>> >> in the servers that must support
>>>>> >> SS7/ISUP.
>>>>
>>>>
>>>> >
>>>> >Do you use the Aculab SS7/ISUP together with Asterisk
somehow?
>>>> >
>>>> >/Olle
>>>
>>>
>>>
>>> Not yet, I am still porting applications from our old properitary
box to
>>> the Asterisk base. There's a lot of stuff that work diffrent
and I have to
>>> add quite a bit of code to run a site that consist of an array of
'*' servers.
>>> Issues are like database replication, locating which server the SIP
user is registered on
>>> a.s.o. The old systems are build around Aculab cards and some of
theses systems use
>>> the SS7 but not in a mix with '*' yet. We did some analasys
of the SS7 issue before we
>>> started on the '*' road. We decided that we need different
approaches depending upon
>>> the size of the '*' site.
>>>
>>> Big site (upto 16 servers) would use the Milborne box from
Datakinetec in UK. It's a
>>> signalling converter that does all the heavy SS7 stuff. Each server
would still carry
>>> 2 TE410P cards. The channel driver would be rather simple since you
get the signalling
>>> over tcp (Q931) in the format like 'incoming call in port 5
timeslot 10'.
>>>
>>> For smaller systems the Aculab card with 2 E1 lines and 2 ethernet
interfaces could be a solution.
>>> The last release from them that I used did allow you to define a
CIC code map that
>>> would span 2 extra TE410P cards and still let the Aculab handle the
signalling.
>>> This does ofcourse still require a new '*' channel driver
but it wouldn't need to deal
>>> with the SS7 stack (only Q931).
>>>
>>> I hope that I soon will get some time to play with this stuff
again.
>>> Freddi
>>
>>
>
>Instead of bothering the channel driver, sounds like you just need to
>modify libpri then to handle the Q931 coming from a source other than a
>zap channel. Would be much simpler I hope to handle it this way than to
>write a special driver.
> -- Steven Critchfield
>
I realized by re-reading my mail that it can mislead a bit. The Q931
statemachine is embedded on the Aculab card and '*'
interface will have to be through the standard Aculab API which more
or less sits on top of a Q931. I am still a newbie
to '*' so I am not ready to take a headdive into that project yet.
b.r.
Freddi
