[asterisk-users] Disconnect on PRI ignored?

Matthew Fredrickson creslin at digium.com
Wed Jul 16 12:13:59 CDT 2008


Alexander Zielke wrote:
> Hi List,
>
> i recently set up a system with a TE410P. Everything works, except that 
> disconnects don't seem to be processed.
>
> Here is what i get:
>
>     -- SIP/2025-08245ac8 is ringing
>     -- SIP/2025-08245ac8 is ringing
>     -- SIP/2025-08245ac8 is ringing
> < Protocol Discriminator: Q.931 (8)  len=13
> < Call Ref: len= 2 (reference 23819/0x5D0B) (Originator)
> < Message type: DISCONNECT (69)
> < [08 02 80 90]
> < Cause (len= 4) [ Ext: 1  Coding: CCITT (ITU) standard (0)  Spare: 0  
> Location: User (0)
> <                  Ext: 1  Cause: Normal Clearing (16), class = Normal 
> Event (1) ]
> < [1e 02 82 88]
> < Progress Indicator (len= 4) [ Ext: 1  Coding: CCITT (ITU) standard 
> (0)  0: 0  Location: Public network serving the local user (2)
> <                               Ext: 1  Progress Description: Inband 
> information or appropriate pattern now available. (8) ]
> -- Processing IE 8 (cs0, Cause)
> -- Processing IE 30 (cs0, Progress Indicator)
> q931.c:3779 q931_receive: call 23819 on channel 6 enters state 12 
> (Disconnect Indication)
>     -- SIP/2025-08245ac8 is ringing
>     -- SIP/2025-08245ac8 is ringing
> ...
>
> I just made a call from the outside to a local SIP-Phone, but when the 
> outside call hangs up, the Phone keeps ringing.
> The call will only hangup, if i take the call, or wait for the call to 
> time out.
>
> The only similar thing i found is the bug at 
> http://bugs.digium.com/view.php?id=9588, but that seems fixed in 1.4.21.1.
> Did anyone else experienced something like that?
>   

If you are using libpri-1.4.4, you should either downgrade to 1.4.3 or 
upgrade to 1.4.5.  A new default behavior was introduced in 1.4.4 (which 
should have been optional, not default) which causes a channel to be 
left open until the RELEASE timer expires when a DISCONNECT is received 
with Inband progress information avaiable.

Matthew Fredrickson



More information about the asterisk-users mailing list