PRI trunk crossover clock problem
Submitted by sewlist on Tue, 03/04/2008 - 4:55am.
Hi There,
We have got a PRI back to back with a Samsung PBX(7200), everything works good, but seldom , The channel restarts and drop all existing calls for that 2 second restart
I've done some search and its appears to be a CRC issue, where the clock gets out of sync and restarts if to many errors.
I've tried numerous settings , but still no luck, its starts happening after about 15 min of operation and then happens random.
My zaptel.conf
# Span 1: WCT1/0 "Wildcard TE12xP Card 0"
span=1,1,0,ccs,hdb3,crc4
# termtype: te
bchan=1-15,17-31
dchan=16
# Span 2: WCT1/1 "Wildcard TE12xP Card 1"
#span=2,2,0,ccs,hdb3,crc4
span=2,1,0,ccs,hdb3,crc4
# termtype: te
bchan=32-46,48-62
dchan=47
My span 1 is fine, but span 2 is the bastard.
Is there any way I can reduce this?, Could it be a cable? or is it possible to turn CRC checking off on the samsung and trixbox?, will this makes things worse?
Any advice would be apreciated
Thanks
Sew
Potensial Resolution
Tue, 03/04/2008 - 5:47am
Timing
Tue, 03/04/2008 - 5:50am
Dear Sew
From span=1,1,0,ccs,hdb3,crc4 we can tell that you sync to your telco, the second 1 means you get the timing from them if this is the case the other PBX is expecting timing.
So span two should look like this span=2,0,0,ccs,hdb3,crc4 so you send timing to the PBX
Hope that helps
http://www.voip-info.org/wiki/index.php?page=Zaptel.conf+span+syn...
Al.
Current Settings
Tue, 03/04/2008 - 11:54pm
Thanks for the info Scott
Here is my current config on span 2 , The random channel restart still happens
# Span 2: WCT1/1 "Wildcard TE12xP Card 1"
span=2,0,0,ccs,hdb3,crc4
# termtype: te
bchan=32-46,48-62
dchan=47
What u mean by line clocking on the pbx, u mean the samsung side must be set for line clocking?, I think the techs has set it to get timing from me, said I must emulate my telco for them
Thanks again
S
That would be line clocking.
Wed, 03/05/2008 - 12:49am
That would be line clocking. I am not sure how good the clock is from your PRI card. It sounds like you are getting slips.
I would hate to think you have to put a CSU between them to fix up the clock. Let's see what every one else has to say.
Scott
--
Scott
aka "Skyking"
More Info
Thu, 03/06/2008 - 4:29am
ok, i've done some more digging,
I've turned crc off both side(before scott said was bad idea), problem persist, the samsung can only get timing from me, they can't be master
looking at /var/log/asterisk/full i see a possible cause of the span restarts
PRI got event: HDLC Abort (6) on Primary D-channel of span 2
This happens the whole time, Googling this event, there was few suggestions
Some ppl suggest its a IDE problem which cause intterupts issues, using hdparm -u1 (unmask intterupt ), one guy actually replace motherboard to fix the issue. Other suggest Intel issues with network cards
As last resort ill replace the M/B
some info on the system
Currently the server is P4 1.8Ghz Compaq brand
hdparm /dev/hda
/dev/hda:
multcount = 16 (on)
IO_support = 3 (32-bit w/sync)
unmaskirq = 1 (on)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
readahead = 256 (on)
geometry = 65535/16/63, sectors = 78165360, start = 0
# cat /proc/interrupts
CPU0
0: 2494416 local-APIC-edge timer
1: 449 IO-APIC-edge i8042
2: 0 XT-PIC cascade
8: 1 IO-APIC-edge rtc
12: 118 IO-APIC-edge i8042
14: 9311 IO-APIC-edge ide0
15: 22011 IO-APIC-edge ide1
153: 2478957 IO-APIC-level uhci_hcd:usb1, wcte12xp
161: 0 IO-APIC-level Intel 82801DB-ICH4
169: 2476223 IO-APIC-level uhci_hcd:usb3, wcte12xp
177: 0 IO-APIC-level uhci_hcd:usb2
193: 25071 IO-APIC-level eth1
209: 0 IO-APIC-level ehci_hcd:usb4
NMI: 0
LOC: 2494386
ERR: 0
MIS: 0
Any advice would be appreciated
Thanks again
S
I wish I had a T1 test set
Thu, 03/06/2008 - 11:12am
I wish I had a T1 test set on the line.
By any chance do you have a friend with a T-Bird or other T1 analyzer?
HDLC is the encapsulation format for the data inside the D-channeling. This is where all the signaling takes place. Loosing the HDLC is a symptom not the cause.
I am not sure how clocking is derived on the PRI card. I doubt the card has a good free running OCXO on it, probably never designed as a timing source.
The more I think about this the more a CSU would solve all of your problems. You can time right off it.
Let me see if I have any left in the warehouse, We tossed a ton of them over the years.
Scott
--
Scott
aka "Skyking"
IRQ info
Fri, 03/07/2008 - 1:17am
lspci -vb
01:02.0 Ethernet controller: Digium, Inc. Unknown device 8001 (rev 11)
Subsystem: Digium, Inc. Unknown device 8001
Flags: bus master, medium devsel, latency 32, IRQ 9
I/O ports at d000
Memory at d0022000 (32-bit, non-prefetchable)
Capabilities: [c0] Power Management version 2
01:03.0 Ethernet controller: Digium, Inc. Unknown device 8001 (rev 11)
Subsystem: Digium, Inc. Unknown device 8001
Flags: bus master, medium devsel, latency 32, IRQ 5
I/O ports at d100
Memory at d0020000 (32-bit, non-prefetchable)
Capabilities: [c0] Power Management version 2
Only other one with IRQ 9 is
00:1e.2 Multimedia audio controller: Intel Corporation 82801FB/FBM/FR/FW/FRW (ICH6 Family) AC'97 Audio Controller (rev 05)
Subsystem: Micro-Star International Co., Ltd. Unknown device b010
Flags: bus master, medium devsel, latency 0, IRQ 9
I/O ports at e500
I/O ports at e600
Memory at d01c1000 (32-bit, non-prefetchable)
Memory at d01c2000 (32-bit, non-prefetchable)
Capabilities: [50] Power Management version 2
Want me to disable onboard audio?
Dont know if the aterisk CLI zap show status helps
Description Alarms IRQ bpviol CRC4
Wildcard TE12xP Card 0 OK 24 0 0
Wildcard TE12xP Card 1 OK 26 0 0
Out of interest sake, I 've swopped pc to a P4 3Ghz with 512MB ram, and almost no change about HDLC issue, everything changed except the disk ( had to do this 5am this morning to avoid PRI downtime ) :-)
HDLC abort issue was suggested by a guy who posted this
http://www.ctunion.com/node/95
CSU
Fri, 03/07/2008 - 1:23am
http://kb.digium.com/entry/13
Fri, 03/07/2008 - 10:22am
http://kb.digium.com/entry/138/
http://kb.digium.com/entry/95/
Make a loop back plug and run the test. If you don't pass, disable audio and run the test again. You can try stopping other programs or switching hardware until you pass the test.
If you pass the test, run the tests again but plug your cross over cable into the your card, and put the loop back plug on the end of the cable.
If you still pass the test, then your hardware is good. Though you could still try upgrading to asterisk 1.4, then use "hardhdlc" in your zaptel.conf.
I'm not sure what else you could try, other than scott's suggestions.
Member Since:
2007-10-08