I’d be grateful if anyone here could comment knowledgeably on an idea
that I have had, as to whether it should be possible or not.

Consider two Asterisk boxes, each with one or more E1s on EuroISDN.
Each box has a different telephone number that hunts across all its
E1 channels. In addition there is another number that hunts across
all the channels on all the boxes.

A call comes in to one of the boxes and is answered. After some
interaction with the caller, the box decides that the call needs
to be handled by the other box.

I don’t want just to relay the call through to the second box using
IAX or SIP or an additional PSTN channel. What I would like to do is
to redirect the call in the PSTN so that it ends up connected only to
the second box. That is why each box would have its own telephone
number as well as the global access number for all boxes.

Obviously, as well as the ability to redirect the call transparently
to a channel on the second box, I also need the second box to be able
to identify that this call is redirected, and not a fresh call from
outside. Presumably I sould use CLI in some way, and/or perhaps
USERUSERINFO.

Is the above possible using ISDN? If not, would it be possible using
SS7? Is SS7 support still active in Asterisk?

Any insight would be appreciated!

Cheers
Tony

Hi,
I’m looking for some advice on how to solve DTMF issues.
I have 2 boxes, one which is the connection to the PSTN (PSTN) through
PRIs and SIP trunks, and a second (PBX) which has UAs registered to
it.
We have a customer that has an existing pbx that we trunk analog lines
to using a GXW-4008.
The GXW is set to dtmfmode inband. This seems to provide the best outbound DTMF.

The issue I’m currently having is with inbound DTMF.
PBX and PSTN are connected through a standard sip trunk. Both machines
are on the same physical switch.

Here are the results I’ve seen:

PBX < -> PSTN using rfc2833 | Incoming call on PRI | DTMF on pbx
voicemail system fails (dup/missing digits)
PBX < -> PSTN using inband | Incoming call on PRI | DTMF on pbx
voicemail system is correct

PBX < -> PSTN using rfc2833 | Incoming call on SIP | DTMF on pbx
voicemail system is correct
PBX < -> PSTN using inband | Incoming call on SIP | DTMF on pbx
voicemail system is correct

All asterisk versions are 1.4.35.
PRI card is a Sangoma A104 with HW DTMF detection.

Does asterisk just have a problem converting the DTMF from the
D-channel to rfc2833?
The DTMF log looks ok (I dialed ’642′), so I’m not sure where the
issue is coming in.

[Jul 4 21:05:44] DTMF[9769] channel.c: DTMF begin ’6′ received on Zap/15-1
[Jul 4 21:05:44] DTMF[9769] channel.c: DTMF begin passthrough ’6′ on Zap/15-1
[Jul 4 21:05:44] DTMF[9769] channel.c: DTMF end ’6′ received on
Zap/15-1, duration 100 ms
[Jul 4 21:05:44] DTMF[9769] channel.c: DTMF end accepted with begin
’6′ on Zap/15-1
[Jul 4 21:05:44] DTMF[9769] channel.c: DTMF end passthrough ’6′ on Zap/15-1
[Jul 4 21:05:45] DTMF[9769] channel.c: DTMF begin ’4′ received on Zap/15-1
[Jul 4 21:05:45] DTMF[9769] channel.c: DTMF begin passthrough ’4′ on Zap/15-1
[Jul 4 21:05:45] DTMF[9769] channel.c: DTMF end ’4′ received on
Zap/15-1, duration 100 ms
[Jul 4 21:05:45] DTMF[9769] channel.c: DTMF end accepted with begin
’4′ on Zap/15-1
[Jul 4 21:05:45] DTMF[9769] channel.c: DTMF end passthrough ’4′ on Zap/15-1
[Jul 4 21:05:45] DTMF[9769] channel.c: DTMF begin ’2′ received on Zap/15-1
[Jul 4 21:05:45] DTMF[9769] channel.c: DTMF begin passthrough ’2′ on Zap/15-1
[Jul 4 21:05:46] DTMF[9769] channel.c: DTMF end ’2′ received on
Zap/15-1, duration 100 ms
[Jul 4 21:05:46] DTMF[9769] channel.c: DTMF end accepted with begin
’2′ on Zap/15-1
[Jul 4 21:05:46] DTMF[9769] channel.c: DTMF end passthrough ’2′ on Zap/15-1

Thanks.

Asterisk 1.8.3.2

I have been getting this warning constantly on CLI in a call scenario where
I use local channels to connect SIP with PSTN.
I use callfile and local channel to first call a PSTN number and if
answered, use local channel to call SIP phone with music on hold enabled in
Dial string.
If I call PSTN from SIP directly or vice versa I don’t see this warning
coming.
On SIP I have allowed only one codec(alaw).

[Jun 28 15:05:00] WARNING[31016] chan_sip.c: Asked to transmit frame type
slin, while native formats is 0×8 (alaw) read/write = 0×8 (alaw)/0×8 (alaw)

I also tried to yes/no option transcode_via_sln in asterisk.conf without any
success.
Any idea?
Thanks,

You mean say

0=Slave (Use PSTN clock)
1=Master(generate Internal clock)

So best option is 0 for all span if you connected on PSTN right ?

Date: Fri, 27 May 2011 17:27:43 -0300
From: rafaelsnsa@gmail.com
To: asterisk-users@lists.digium.com
Subject: Re: [asterisk-users] DAHDI span timeing source

Hi
The timing source is the clock of the system. When a equipment is 0, the other should be 1. The correct is: 0=slave, 1=master. The default for private systems is “slave”.

Hello.

Considering the following setup:

Legacy PBX –(ISDN)–> Asterisk –(MFC/R2)–> PSTN

When a user dials out, Asterisk receive overlap digits, matches them to an extension and dial the PSTN, completing the call. So far so good.

The issue I’m trying to solve (or at least improve) is the call takes much longer to complete than the users were used to before having Asterisk between the PBX and the PSTN. It happens because the digits are sent to the PSTN only after the extension is matched in the dialplan, and dialing on MFC/R2 takes a few seconds.

Here’s the console log. Notice how it takes 6 seconds from the instant the user starts dialing to the instant the dialed number starts to ring. First 3 seconds is the user manually dialing plus Asterisk absolute timeout. Next 3 seconds are the time Asterisk takes to dial the number to the PSTN and the call be accepted.

[Apr 26 10:57:13] — Accepting overlap call from ’7416′ to ‘‘ on channel 0/1, span 2
[Apr 26 10:57:13] — Starting simple switch on ‘DAHDI/32-1′

*** User finished dialing + Asterisk absolute timeout ***

[Apr 26 10:57:16] — Executing [0145333114657@pbx:1] Answer(“DAHDI/33-1″, “”) in new stack
[Apr 26 10:57:16] — Executing [0145333114657@pbx:2] Dial(“DAHDI/33-1″, “DAHDI/g1/0145333114657″) in new stack
[Apr 26 10:57:16] — Called g1/0145333114657

*** Asterisk starts dialing to the PSTN ***

[Apr 26 10:57:19] MFC/R2 call has been accepted on forward channel 1
[Apr 26 10:57:19] — DAHDI/1-1 is ringing

*** Dialed number finally rings ***

So my question is: is there a way to fully overlap the digits from the user’s phone on the PBX (ISDN) to the PSTN (MFC/R2), eliminating the need to wait for an extension to be matched? I already have overlapdial=yes in both spans, but that didn’t made it. Also googled for it, even searched this list archives but found nothing.

chan_dahdi.conf:

[channels]

signalling=mfcr2
mfcr2_variant=br
mfcr2_get_ani_first=no
mfcr2_max_ani=20
mfcr2_max_dnis=4
mfcr2_category=national_subscriber
mfcr2_logdir=span1
mfcr2_call_files=no
mfcr2_logging=all
mfcr2_mfback_timeout=-1
mfcr2_metering_pulse_timeout=-1
mfcr2_allow_collect_calls=yes
mfcr2_double_answer=no
mfcr2_immediate_accept=no
mfcr2_forced_release=no
mfcr2_charge_calls=yes
language=pt_BR
echocancel=yes
echocancelwhenbridged=no
callgroup=0
pickupgroup=0
group=1
context=telco
overlapdial=yes
channel => 1-15,17-31

switchtype=euroisdn
pridialplan=unknown
prilocaldialplan=unknown
priindication=outofband
signalling=pri_net
busydetect=yes
busycount=5
language=pt_BR
echocancel=yes
echocancelwhenbridged=no
overlapdial=yes
group=2
context=pbx
channel => 32-46,48-62

extensions.conf:

[telco]
exten => _X.,1,Dial(DAHDI/g2/${EXTEN})

[pbx]
exten => _X.,1,Dial(DAHDI/g1/{$EXTEN})