Hi Everyone,
I am currently happly running Asterisk 1.0.9 (the final 1.0 branch?) at www.iax.ie with sipfriends and iaxfriends.
I’m obviously very interested in moving our servers to 1.2.5.
I upgraded one server last night and I came across a problem and I dropped back to 1.0.9.
I’m using TE405P cards with 4 X E1 (euroISDN) PRI lines into the cards - one card per server.
When I upgraded to 1.2.5 what happened was this:
When I dialed in from a PSTN line into the upgraded asterisk server it looked, on the CLI, like it was answering and playing prompts to me but from the PSTN side of it it just kept ringing and then, after a couple of rings, the PSTN gave a busy / unavailable tone and on the asterisk CLI is said “Channel X/XX, span X got hangup” and that was it. When I downgraded back to 1.0.9 all was OK again.
I’m using the newer firmware version of the TE405Ps from Digium.
Does anyone know how to fix this problem? Otherwise I’ll be stuck on 1.0 forever
thanks,
Derek Conniffe
Hi Derek
What does the pri debug show? I have TE cards running 1.2 fine. Also what does zapata.conf look like? Have you got access to a test set so you could monitor the channels.
Ian
Hi Ian,
I’ll have to check pri debug - I didn’t do that (ever before).
My /etc/asterisk/zapata.conf is the same one as I’m using for asterisk 1.0.9 (I only made a few alterations for Realtime SIP/IAX settings but did’nt load any new sample configurations).
My zapata.conf, with all comments removed, on this server is: -
[trunkgroups]
trunkgroup => 1,16
trunkgroup => 2,47
trunkgroup => 3,78
trunkgroup => 4,109
spanmap => 1,1,0
spanmap => 2,2,0
spanmap => 3,3,0
spanmap => 4,4,0
[channels]
context=PRI-NTL
switchtype=euroisdn
signalling=pri_cpe
usecallerid=yes
hidecallerid=no
callwaiting=no
restrictcid=no
threewaycalling=no
immediate=no
echocancel=yes
group=1
channel=>1-15,17-31
group=2
channel=>32-46,48-62
group=3
channel=>63-77,79-93
group=4
channel=>94-108,110-124
rxwink=300 ; Atlas seems to use long (250ms) winks
rxgain=0.0
txgain=0.0
OK looks ok except for rxwink. this shouldnt be here its a T1 timer not a e1 thing. I dont know why but seems to be left in by maby people.
Ian
thanks Ian,
I’ll try without that timing line later and if the problem is still there I’ll check the pri debug output
Derek