[asterisk-r2] suddenly system has become unstable

[Raúl Torres Duque]

Linux Kernel: 2.6.18-92.1.6.el5
Asterisk version: 1.4.21.2
Zaptel Version: 1.4.12.1
Wanpipe versión: 3.2.7.1
Sangoma A101 card

This is an "oldie" that had been working perfectly up until recently. It has a twin sister that is not presenting any issues at the moment.
Telco (Telmex) says the line is OK.
Even though the card is green and there are no alarms, the system stops processing the calls all of the sudden. By resetting the services (Wanpipe, Zaptel and Asterisk), calls can be made normally again. It'll work for a variable amount of time (from a couple of hours to days) and then it will break again. We have no clue whatsoever what could be happening and as I said if not for the twin system working fine we would've tried to upgrade this one. Sangoma has not come back to me to verify is not a hardware issue.

=====Zapata.conf=====
[trunkgroups]

[channels]
language=es
context=default
usecallerid=yes
hidecallerid=no
callwaiting=yes
usecallingpres=yes
callwaitingcallerid=yes
threewaycalling=yes
transfer=yes
canpark=yes
cancallforward=yes
callreturn=yes
echocancel=yes
echocancelwhenbridged=yes
relaxdtmf=yes
rxgain=0.0
txgain=0.0
;group=0
callgroup=1
pickupgroup=1
busydetect=yes
busycount=4
;hanguponpolarityswitch

immediate=no

;Sangoma A101 port 1 [slot:0 bus:17 span:1] <wanpipe1>
context=from-pstn
group=0
signalling=mfcr2
mfcr2_variant=mx
mfcr2_get_ani_first=no
mfcr2_max_ani=10
mfcr2_max=4
mfcr2_category=national_subscriber
mfcr2_call_files=yes
mfcr2_logdir=span1
mfcr_logging=all
mfcr2_mfback_timeout=-1
;mfcr2_metering_pulse_timeout=-1
mfcr2_metering_pulse_timeout=500  <--I changed this from another post but it made no difference

channel =>1-10

=====EXCERPT FROM /var/log/messages=====
Aug  2 08:58:18 pbx kernel: wanpipe1: OOF alarm is ON
Aug  2 08:58:18 pbx kernel: wanpipe1: E1 disconnected!
Aug  2 08:58:18 pbx kernel: wanpipe1: RAI alarm is ON
Aug  2 08:58:18 pbx kernel: wanpipe1: RED alarm is ON
Aug  2 08:58:18 pbx kernel: wanpipe1: AFT communications disabled!
Aug  2 08:58:18 pbx kernel: wanpipe1: Starting TDMV 1ms Timer
Aug  2 08:58:19 pbx kernel: wanpipe1: RAI alarm is OFF
Aug  2 08:58:20 pbx kernel: wanpipe1: RAI alarm is ON
Aug  2 08:58:39 pbx kernel: wanpipe1: RAI alarm is OFF
Aug  2 08:58:39 pbx kernel: wanpipe1: OOF alarm is OFF
Aug  2 08:58:39 pbx kernel: wanpipe1: RED alarm is OFF
Aug  2 08:58:40 pbx kernel: wanec1: The H100 slave has lost its framing on the bus!
Aug  2 08:58:41 pbx kernel: wanec1: The H100 slave has lost its framing on the bus!
Aug  2 08:58:45 pbx kernel: wanpipe1: E1 connected!
Aug  2 08:58:45 pbx kernel: wanpipe1: AFT communications enabled!
Aug  2 08:58:45 pbx kernel: wanpipe1: AFT Global TDM Intr
Aug  2 08:58:45 pbx kernel: ADDRCONF(NETDEV_CHANGE): w1g1: link becomes ready
Aug  2 08:58:45 pbx kernel: wanpipe1: Global TDM Ring Resync
Aug  2 08:58:45 pbx kernel: wanpipe1: Card TDM Rsync Rx=0 Tx=2
Aug  2 08:58:46 pbx kernel: wanpipe1: RAI alarm is OFF
Aug  2 08:58:46 pbx kernel: wanpipe1: OOF alarm is OFF

=====File /var/log/asterisk/mfcr2/span1/chan-1-forward-3-20090803153702.call (failed call)=====
[15:37:02:711] [Thread: 3078372240] [Chan 1] - Call started at Mon Aug  3 15:37:02 2009 on chan 1
[15:37:02:711] [Thread: 3078372240] [Chan 1] - ABCD Tx >> [SEIZE] 0x00
[15:37:02:711] [Thread: 3078372240] [Chan 1] - ABCD Raw Tx >> 0x01
[15:37:02:712] [Thread: 3078372240] [Chan 1] - ABCD Tx >> [CLEAR FORWARD] 0x08
[15:37:02:712] [Thread: 3078372240] [Chan 1] - ABCD Raw Tx >> 0x09
[15:37:10:730] [Thread: 3084327824] [Chan 1] - calling callback on chan 1
[15:37:10:731] [Thread: 3084327824] [Chan 1] - Seize Timeout Expired!
[15:37:10:731] [Thread: 3084327824] [Chan 1] - Protocol error. Reason = Seize Timeout, R2 State = Clear Forward Transmitted, MF state = MF Engine Off, MF Group = Forward MF init
[15:37:10:731] [Thread: 3084327824] [Chan 1] - DNIS = 5517668510, ANI = 3080, Last MF Signal =

On this call we got the "All circuits are busy now" prompt

Thanks in advance for the help
Raúl

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-r2/attachments/20090803/1e8e172d/attachment.htm