ring group fails ringing some sip phones

[goodmanuel]

Hi,

I have the following problem:

Incoming route set to ring group 600 (default ring group) containing
extensions 11,12,13,14,15,16,18,19 mode "ringall"
extensions are all SIP Phones except 19 that is fxs/1 port.
Each SIP extension (all same brand and model Snom300) is working, call out
and in from other extensions and out to pstn via fxo ports are ok.

But when call is incoming from pstn and goes to ring group 600 SIP
extensions 16 and 18 won't ring.

After I check the asterisk CLI I always get this (I put the part is most
interesting and point to the problem):

-- Executing [s@macro-dial:7] Dial("PIKA/fxo/1",
"SIP/11&SIP/12&SIP/13&SIP/14&SIP/15&SIP/16&SIP/18&PIKA/fxs/1|60|trM(auto-blkvm)")
in new stack
-- Called 11
-- Called 12
-- Called 13
-- Called 14
-- Called 15
-- group 1/1 channel 0/0 type=5
-- Called fxs/1
-- PIKA/fxs/1 is ringing
-- SIP/15-1017bf58 is ringing
-- SIP/14-10177f30 is ringing
-- SIP/11-102e3a70 is ringing
-- SIP/13-10331d50 is ringing
-- SIP/12-1030c258 is ringing
-- SIP/15-1017bf58 is ringing
-- SIP/14-10177f30 is ringing
-- SIP/11-102e3a70 is ringing
-- SIP/13-10331d50 is ringing
-- SIP/12-1030c258 is ringing
-- SIP/15-1017bf58 is ringing
-- SIP/14-10177f30 is ringing
-- SIP/11-102e3a70 is ringing
-- SIP/13-10331d50 is ringing
-- SIP/12-1030c258 is ringing
-- SIP/12-1030c258 answered PIKA/fxo/1

as you can see the dial command is correct but the ringing list exclude the
last 2 SIP extensions in the list of ring group in fact they don't ring. In
their place appear "-- group 1/1 channel 0/0 type=5" that I don't understand
the meaning
fxs/1 is always working even is last in ring group priority
If I exchange extensions 16 and 18 with 11 and 12 in the priority list of
ringgroup (ringall) 11 and 12 won't ring and 16 and 18 will ring.

What happens??

I attach as follows the software version in Warp: (PADS 2.2.5 image
installed)

Mon Aug 23 10:26:24 EDT 2010
PADS version 2.2.5-6
toolchain version 1.0.2
Kernel version 2.6.31.7-7
Skeleton version 1.0.0-20
warp-locales 1.0.1
Busybox 1.10.3
mysql 5.1.30
sox version 14.2.0
curl version 7.19.2
Zaptel version 1.4.9.2
Asterisk version 1.4.25.1
asterisk-addons version 1.4.8
HMP version 2.8.11-1
LCD LIB version 1.0.0-8
Astmanproxy version 1.1.3
pikagsm 1.0.22989
chan_gsm 1.0.22989
chan_pika version http://svn.pikatech.com/chan_pika/tags/3.8/3.8.7.6
crond included, refer to busybox version
daemontools 0.76
Dhcpcd version 3.2.3
dnsmasq version 2.47
dosfstools 3.0.9
Dropbear version 0.50
e2fsprogs 1.41.0
LibXML version
libiconv 1.12
gettext 0.16.1
LIBIDN version 1.9
libpng 1.2.36
freetype2 2.3.9
PHP version 5.2.9
FreePBX version 2.7.0
gdbm version 1.8.3
GHOSTSCRIPT version 8.62
lighttpd version 1.4.19
NTP included
perl version 5.10.0
PHP Pear version 3.5.6
ssmtp 2.61
tftpd version 0.48
libtiff version 3.8.2
PIKA Update Utilities 2.0.34
zoneinfo 1.0.0
IMAGE_CREATION_DATE=Mon Aug 23 11:03:43 EDT 2010


thanks for any help

Manuel

[mrecoskie]

Hi Manuel,

Interesting. I have not encountered this problem before using ring groups in FreePBX (and even non-FreePBX configurations). That being said I am typically using BRI not FXO.

Out of curosity what are the SIP endpoints that you are communicating with? As a hunch - is your SIP traffic going through some device that imposes a restriction on the number of simultaneous SIP calls? I know some IP DECT phones where this can be the case.

The "-- group 1/1 channel 0/0 type=5" is a log produced by the Pika channel when trying to locate a specific channel. (in this case type 5 being 'phone') Have you tried removing the fxs port from the list to see if this had any affect? Have you tried moving the fxs port in the list?

Just some thoughts.

[fsinetworks]

Hi,

With pika warp asterisk i encoured this problem with bri and with fxo.

Finally, i wrote a context in the files /etc/asterisk/extensions_custom.conf


[from-numeris]
exten => s,1,NoOp("On est dans le context numeris")
exten => s,2,Dial(SIP/101&SIP/106,30,tTwW)
exten => s,3,Dial(SIP/101&SIP/106&SIP/104&SIP/105,10,tTwW)
exten => s,4,Dial(SIP/101&SIP/106&SIP/104&SIP/105&SIP/100,10,tTwW)
exten => s,n,Set(LANGUAGE()=fr)
exten => s,n,Goto(s-${DIALSTATUS},1)
exten => s-NOANSWER,1,Voicemail(u101@default)
;exten => s-BUSY,1,Voicemail(b101@default)
exten => s-BUSY,1,Voicemail(u101@default)

exten => 7800,1,NoOp("On est dans le context numeris")
exten => 7800,2,Dial(SIP/101&SIP/106,20,tTwW)
exten => 7800,3,Dial(SIP/101&SIP/106&SIP/104&SIP/105,10,tTwW)
exten => 7800,4,Dial(SIP/101&SIP/106&SIP/104&SIP/105&SIP/100,10,tTwW)
exten => 7800,n,Set(LANGUAGE()=fr)
exten => 7800,n,Goto(7800-${DIALSTATUS},1)
exten => 7800-NOANSWER,1,Voicemail(u101@default)
;exten => 7800-BUSY,1,Voicemail(b101@default)
exten => 7800-BUSY,1,Voicemail(u101@default)

Best regards,
Fabien

[goodmanuel]

Hi Manuel,

Interesting. I have not encountered this problem before using ring groups in FreePBX (and even non-FreePBX configurations). That being said I am typically using BRI not FXO.

Out of curosity what are the SIP endpoints that you are communicating with? As a hunch - is your SIP traffic going through some device that imposes a restriction on the number of simultaneous SIP calls? I know some IP DECT phones where this can be the case.

The "-- group 1/1 channel 0/0 type=5" is a log produced by the Pika channel when trying to locate a specific channel. (in this case type 5 being 'phone') Have you tried removing the fxs port from the list to see if this had any affect? Have you tried moving the fxs port in the list?

Just some thoughts.

Hi mrecoskie,

I'm again over this issue that hadn't been solved so far. The telephones are SNOM 300 and should not impose any restrictions. I have tried removing and moving fxs port in the list without any effect. I have tried to call the ring group internally (not from pstn) and the same result, also I have tried QUeues instead of ring groups and still the same effect or similar (maybe just change the extension who's not ringing but always 5 extensions at most ring at the same time)
I have tried also customised context but whenever I put more than 5 SIP extensions in DIAL command only the first 5 will ring.
Any ideas how to solve or workaround this problem?

[goodmanuel]

Hi Manuel,

Interesting. I have not encountered this problem before using ring groups in FreePBX (and even non-FreePBX configurations). That being said I am typically using BRI not FXO.

Out of curosity what are the SIP endpoints that you are communicating with? As a hunch - is your SIP traffic going through some device that imposes a restriction on the number of simultaneous SIP calls? I know some IP DECT phones where this can be the case.

The "-- group 1/1 channel 0/0 type=5" is a log produced by the Pika channel when trying to locate a specific channel. (in this case type 5 being 'phone') Have you tried removing the fxs port from the list to see if this had any affect? Have you tried moving the fxs port in the list?

Just some thoughts.

Hi mrecoskie,

I m using Snom 300 SIP endpoints. No restrictions. Tried to remove or move fxs port in the list. Tried to move the SIP extensions in the list. The result is change the extensions who ring and the ones who don't. But always 5 SIP extensions at most, and fxs always ok.

[goodmanuel]

Hi mrecoskie, the endpoints are snom 300. No restrictions. Tried to move and remove fxs port in the list with any effects. Seems only 5 SIP endpoints can ring at most in ringroups. Also I tried queues with same effect. Is there any solution or workaround to this problem?

[mrecoskie]

Hi manuel,

I wonder if something like 'call-limit' could be set somewhere? Maybe you can
try grep'ing for it. I know 5 is used in the Asterisk default sip.conf for some items.
Could you post your sip.conf?

Also have you examined the cpu and memory usage of the unit? ('top' and 'free')

Beyond this I would tempted to trace the SIP messages to and from the box for clues as to why this might be happening. Or possibly try a different type of SIP endpoint?

[goodmanuel]

Hi mrecoskie, thank you! Yes I do some further check as you suggested and finally I found the reason that was a problem with RTP ports (that I had limited previously in rtp.conf) and that was not good for asterisk so that in the end log "full" became big and affected warp performance, (at least that is what I think happened). Now I cleared log full, put rtp port back as asterisk default and SIP endpoints started to ring all again. Now I will try to logrotate full log in order to avoid such a problem for any reason in the future.