[asterisk-bugs] [JIRA] (PRI-168) DISCONNECT with Progress Indicator #8
armeniki (JIRA)
noreply at issues.asterisk.org
Sun Apr 27 23:27:18 CDT 2014
[ https://issues.asterisk.org/jira/browse/PRI-168?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=217629#comment-217629 ]
armeniki edited comment on PRI-168 at 4/27/14 11:25 PM:
--------------------------------------------------------
Hi Richard,
Yes, you're right, in this case we would be only interested in Point 1, ie: the *sending* aspect of the DISCONNECT with Progress Indicator process. This is because in this case Asterisk, being configured as "pri_net", is acting as the "Telco" for the BCM400.
I looked at the ETS ISDN standards document and it does appear that there is a deficiency as you mention as the Standards do mention what I am referring to in Section 5.3.4.1.
(http://www.etsi.org/deliver/etsi_i_ets/300100_300199/30010201/01_60/ets_30010201e01p.pdf) In fact, this is what the network-side timer T306 is for, which is already part of the LibPRI. (Page 152, Table 9.1 of the standard.)
Just to be certain, Nortel have a really handy tool for monitoring their PBX and the PRI connections to it, called BCM Monitor (I'm sure you know of it). So, I had a look on both the production system connected to the Telco and the test system connected to Asterisk whilst dialing a busy number. Both times I did not hangup, I simply waited for the network to disconnect me.
On the Telco setup, as soon as I dialed the number, the message User busy 17 came up and I heard the engaged signal, this continued for 30 seconds, which I'm told is the standard duration.
On the Asterisk setup, as soon as I dialed the number, the message User busy 17 briefly appeared and I got disconnected straight away without hearing anything.
Here's a screen shot of the Telco connection:
http://i1100.photobucket.com/albums/g408/armeniki/telephone/ISDN-telco.png
Here's a screen shot of the Asterisk connection:
http://i1100.photobucket.com/albums/g408/armeniki/telephone/ISDN_asterisk.png
If you look towards the end, you'll notice with the Telco connection that at:
*22:20:45 > CC > Cref Origin: Cref 3 DISCONNECT User busy*
--There is a Progress Indicator attached to it. So in essence, the network is sending the PI #8, waiting 30 seconds for T306, then sends a final DISCONNECT (similar to what Hangup(17) does in the present case of Asterisk).
However, on the Asterisk system at:
*22:27:32 > CC > Cref Origin: Cref 1 DISCONNECT User busy*
--There is no PI #8 coming int with the DISCONNECT and so the PBX responds with the RELEASE_CC.
So indeed this is the reason why we are having this issue.. the lack of those important *03 00 81 88* HEX characters coming through (Progress Indicator #8) with the DISCONNECT.
---
PS: I have updated the title of this thread to reflect the issue more accurately.
was (Author: armeniki):
Hi Richard,
Yes, you're right, in this case we would be only interested in Point 1, ie: the *sending* aspect of the DISCONNECT with Progress Indicator process. This is because in this case Asterisk, being configured as "pri_net", is acting as the "Telco" for the BCM400.
I looked at the ETS ISDN standards document and it does appear that there is a deficiency as you mention as the Standards do mention what I am referring to in Section 5.3.4.1.
(http://www.etsi.org/deliver/etsi_i_ets/300100_300199/30010201/01_60/ets_30010201e01p.pdf) In fact, this is what the network-side timer T306 is for, which is already part of the LibPRI. (Page 152, Table 9.1 of the standard.)
Just to be certain, Nortel have a really handy tool for monitoring their PBX and the PRI connections to it, called BCM Monitor (I'm sure you know of it). So, I had a look on both the production system connected to the Telco and the test system connected to Asterisk whilst dialing a busy number. Both times I did not hangup, I simply waited for the network to disconnect me.
On the Telco setup, as soon as I dialed the number, the message User busy 17 came up and I heard the engaged signal, this continued for 30 seconds, which I'm told is the standard duration.
On the Asterisk setup, as soon as I dialed the number, the message User busy 17 briefly appeared and I got disconnected straight away without hearing anything.
Here's a screen shot of the Telco connection:
http://i1100.photobucket.com/albums/g408/armeniki/telephone/ISDN-telco.png
Here's a screen shot of the Asterisk connection:
http://i1100.photobucket.com/albums/g408/armeniki/telephone/ISDN_asterisk.png
If you look towards the end, you'll notice with the Telco connection that at:
*22:20:45 > CC > Cref Origin: Cref 3 DISCONNECT User busy*
--There is a Progress Indicator attached to it. So in essence, the network is sending the PI #8, waiting 30 seconds for T306, then sends a final DISCONNECT (similar to what Hangup(17) does in the present case of Asterisk).
However, on the Asterisk system at:
*22:27:32 > CC > Cref Origin: Cref 1 DISCONNECT User busy*
--There is no Progress Indicator attached and the PBX simply responds with the RELEASE_CC.
So indeed this is the reason why we are having this issue.. the lack of those important *03 00 81 88* HEX characters coming through (Progress Indicator #8) with the DISCONNECT.
---
PS: I have updated the title of this thread to reflect the issue more accurately.
> DISCONNECT with Progress Indicator #8
> -------------------------------------
>
> Key: PRI-168
> URL: https://issues.asterisk.org/jira/browse/PRI-168
> Project: LibPRI
> Issue Type: New Feature
> Security Level: None
> Affects Versions: 1.4.13
> Reporter: armeniki
> Assignee: Richard Mudgett
> Severity: Minor
>
> Hi everyone,
> As you know, currently when the Hangup() command is used in the Asterisk dial plan, it will tear down both the far and and near end of the call and audio amongst other things. In addition, there is a way of indicating the PRI Cause Code to the user by entering the code within the parenthesis ie: Hangup(1) or Hangup(16), etc.
> Here is the issue: We have a PBX connected to an E1/PRI from the Telco. This is a production system and does not use Asterisk. On this system, whenever someone dials a number which is busy, for example, the phone's display will show the standard PRI message "user busy" and the user will hear an engaged signal (busy signal). Likewise, if a wrong number is dialled, the display will show "unallocated num" and another tone or message will be heard.... and so on.
> Currently, we're doing tests on the same type of system connected to Asterisk via an E1/PRI and have found that this does not happen. Basically, if the user dials a busy number, they will hear a busy signal but they won't see the "user busy" message. Alternatively, if we change the extension script a bit, we can issue a Hangup(17) to make the phone show "user busy" but then there's no way to play the busy signal because the phone/channel hangs up.
> SO....
> Would it be possible to perhaps create a new function called "SendPRICause()" (kind of like the SendText function for SIP phones) so that we can use that instead of Hangup()?
> This way the use can see the message and hear whatever recording needs to be played to them at the same time.
> An example of its usage could be:
> exten => s-DN_CHANGED,1,Progress()
> exten => s-DN_CHANGED,1,SendPRICause(22)
> exten => s-DN_CHANGED,n,Playback(/var/lib/asterisk/sounds/tel/sorry-number-changed,noanswer)
> exten => s-DN_CHANGED,n,Hangup()
> --------------------------
> Cheers,
> Armen
--
This message was sent by Atlassian JIRA
(v6.2#6252)
More information about the asterisk-bugs
mailing list