vicidial.org

by **Trying** » Tue Jan 11, 2011 9:49 am

Hi

I installed 3.1.4 64bit. Vicidial is on and the system sees my Sangoma cards. Should I be worried about the following errors?:

Stopping Asterisk...

smg_ctrl: Stopping running processes...
smg_ctrl: safe_sangoma not running...
smg_ctrl: sangoma_brid is stopped
Removing PID files...done

Stopping Wanpipe...

Unloading Dahdi modules...

Removing old configuration files...

Copying new Wanpipe configuration files...

Copying new Dahdi configuration file (/etc/dahdi/system.conf)...

Copying new Chan-Dahdi configuration files (/etc/asterisk/chan_dahdi.conf)...

Starting Wanpipe...

Starting WAN Router...
Loading WAN drivers: wanpipe done.
Starting up device: wanpipe1

wanconfig: WAN device wanpipe1 driver load failed !!
: ioctl(wanpipe1,ROUTER_SETUP) failed:
: 22 - Invalid argument

Wanpipe driver did not load properly
Please check /var/log/wanrouter and
/var/log/messages for errors

Starting up device: wanpipe2

wanconfig: WAN device wanpipe2 driver load failed !!
: ioctl(wanpipe2,ROUTER_SETUP) failed:
: 22 - Invalid argument

Wanpipe driver did not load properly
Please check /var/log/wanrouter and
/var/log/messages for errors

Configuring interfaces: w1g1 w1g1: unknown interface: No such device

done.
Configuring interfaces: w2g1 w2g1: unknown interface: No such device

done.

Loading Dahdi...
DAHDI Tools Version - 2.4.0
DAHDI Version: 2.4.0
Echo Canceller(s):
Configuration
======================

SPAN 1: CCS/HDB3 Build-out: 0 db (CSU)/0-133 feet (DSX-1)
SPAN 2: CCS/HDB3 Build-out: 0 db (CSU)/0-133 feet (DSX-1)

62 channels to configure.

DAHDI_SPANCONFIG failed on span 1: No such device or address (6)
Error executing command:
dahdi_cfg -v

Would you like to continue?
1. No - exit
2. YES
[1-2, ENTER='No']:

I get the same errors on all my 4 dialers.

by **Trying** » Tue Jan 11, 2011 10:30 am

I guess I have to worry about it because I cannot make or receive calls

by **williamconley** » Tue Jan 11, 2011 10:53 am

FYI: Both core2Quad 2G RAM running the test at the same time on servers in the same cluster ...

Idle box:

Code: Select all: dahdi_test Opened pseudo dahdi interface, measuring accuracy... 99.999% 99.995% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.997% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.999% 99.997% 99.998% 99.998% 99.998% 99.999% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.998% 99.997% 99.998% 99.999% ^C --- Results after 36 passes --- Best: 99.999 -- Worst: 99.995 -- Average: 99.998000, Difference: 99.998000

loaded box (30+ agents, 200 channels, 100 calls):

Code: Select all: dahdi_test Opened pseudo dahdi interface, measuring accuracy... 99.923% 99.928% 99.992% 99.637% 99.644% 99.635% 99.638% 99.635% 99.988% 99.634% 99.998% 99.648% 99.983% 99.641% 99.910% ^C --- Results after 15 passes --- Best: 99.998 -- Worst: 99.634 -- Average: 99.788940, Difference: 99.990983

Call quality is still adequate (ie: no complaints from manager or agents)

by **Kumba** » Tue Jan 11, 2011 12:11 pm

You have to configure the sangoma cards. The system does not know anything about what you are plugging in nor can it find out. You need to know this before you start using it.

You can run wancfg_dahdi and the sangoma wizard will come up to try and walk you through configuring it.

by **Trying** » Tue Jan 11, 2011 12:58 pm

Kumba, I know I must configure the cards.

The output I gave you above is after I configured it via wancfg_dahdi. Yet again I have used the same configuration on the same cards many times before on Vicibox installs without any issues.

by **williamconley** » Tue Jan 11, 2011 1:29 pm

3.1.4 came out today with some location changes for the sangoma stuff ... may improve your success rate. thanks kumba!

by **Trying** » Tue Jan 11, 2011 1:30 pm

Here is some output I noticed in var/log/messages:

Jan 11 15:05:47 dialer1 kernel: [ 1046.842828] Processing WAN device wanpipe1...
Jan 11 15:05:47 dialer1 kernel: [ 1046.842833] wanpipe1: Locating: A101/1D/A102/2D/4/4D/8 card, CPU A, PciBus=7, PciSlot=2
Jan 11 15:05:47 dialer1 kernel: [ 1046.842841] wanpipe1: Found: A101/1D/A102/2D/4/4D/8 card, CPU A, PciBus=7, PciSlot=2, Port=0
Jan 11 15:05:47 dialer1 kernel: [ 1046.842917] wanpipe1: AFT PCI memory at 0xD6100000
Jan 11 15:05:47 dialer1 kernel: [ 1046.842921] wanpipe1: IRQ 20 allocated to the AFT PCI card
Jan 11 15:05:47 dialer1 kernel: [ 1046.842952] wanpipe1: Starting AFT 2/4/8 Hardware Init.
Jan 11 15:05:47 dialer1 kernel: [ 1046.842974] wanpipe1: Enabling front end link monitor
Jan 11 15:05:47 dialer1 kernel: [ 1046.842979] wanpipe1: Global Chip Configuration: used=1 used_type=1
Jan 11 15:05:47 dialer1 kernel: [ 1046.846294] wanpipe1: Global E1 Front End configuration
Jan 11 15:05:47 dialer1 kernel: [ 1046.847324] wanpipe1: AFT Data Mux Bit Map: 0x01234567
Jan 11 15:05:47 dialer1 kernel: [ 1046.847337] wanpipe1: Configuring DS DS26521 E1 FE
Jan 11 15:05:47 dialer1 kernel: [ 1046.847340] wanpipe1: Port 1,HDB3,CRC4,120OH
Jan 11 15:05:47 dialer1 kernel: [ 1046.847342] wanpipe1: Clk Normal:0, Channels: FFFFFFFF
Jan 11 15:05:47 dialer1 kernel: [ 1046.847345] wanpipe1: Sig Mode CCS
Jan 11 15:05:47 dialer1 kernel: [ 1046.850562] wanpipe1: Rx Sensitivity Gain 43dB (default).
Jan 11 15:05:47 dialer1 kernel: [ 1046.850623] wanpipe1: Enable transmit AIS alarm
Jan 11 15:05:47 dialer1 kernel: [ 1046.851068] wanpipe1: Front end successful
Jan 11 15:05:47 dialer1 kernel: [ 1046.851293] wanpipe1: Front End Interface Ready 0x40000000
Jan 11 15:05:47 dialer1 kernel: [ 1046.851297] wanpipe1: Failed to register device in HW Echo Canceller module!
Jan 11 15:05:47 dialer1 kernel: [ 1046.851301] wanpipe1: E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.851425] wanpipe1: E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.851432] wanpipe1: Enable transmit AIS alarm
Jan 11 15:05:47 dialer1 kernel: [ 1046.851452] wanpipe1: AFT communications disabled! (Dev Cnt: 1 Cause: Device Down)
Jan 11 15:05:47 dialer1 kernel: [ 1046.851460] wanpipe1: E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.851468] wanpipe1: Enable transmit AIS alarm
Jan 11 15:05:47 dialer1 kernel: [ 1046.851488] wanpipe1: AFT communications disabled! (Dev Cnt: 1 Cause: Device Down)
Jan 11 15:05:47 dialer1 kernel: [ 1046.851518] wanpipe1: TASKQ Not Running
Jan 11 15:05:47 dialer1 kernel: [ 1046.851520] wanpipe1: Global Chip Shutdown Usage=1, Used_type_cnt=1
Jan 11 15:05:47 dialer1 kernel: [ 1046.851523] wanpipe1: Global E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.853561] wanpipe1: Master shutting down
Jan 11 15:05:47 dialer1 kernel: [ 1046.859394] wanpipe2: Starting WAN Setup
Jan 11 15:05:47 dialer1 kernel: [ 1046.859399]
Jan 11 15:05:47 dialer1 kernel: [ 1046.859401] Processing WAN device wanpipe2...
Jan 11 15:05:47 dialer1 kernel: [ 1046.859405] wanpipe2: Locating: A101/1D/A102/2D/4/4D/8 card, CPU A, PciBus=7, PciSlot=2
Jan 11 15:05:47 dialer1 kernel: [ 1046.859414] wanpipe2: Found: A101/1D/A102/2D/4/4D/8 card, CPU A, PciBus=7, PciSlot=2, Port=1
Jan 11 15:05:47 dialer1 kernel: [ 1046.859491] wanpipe2: AFT PCI memory at 0xD6100000
Jan 11 15:05:47 dialer1 kernel: [ 1046.859494] wanpipe2: IRQ 20 allocated to the AFT PCI card
Jan 11 15:05:47 dialer1 kernel: [ 1046.859520] wanpipe2: Starting AFT 2/4/8 Hardware Init.
Jan 11 15:05:47 dialer1 kernel: [ 1046.859543] wanpipe2: Enabling front end link monitor
Jan 11 15:05:47 dialer1 kernel: [ 1046.859547] wanpipe2: Global Chip Configuration: used=1 used_type=1
Jan 11 15:05:47 dialer1 kernel: [ 1046.862860] wanpipe2: Global E1 Front End configuration
Jan 11 15:05:47 dialer1 kernel: [ 1046.863890] wanpipe2: AFT Data Mux Bit Map: 0x01234567
Jan 11 15:05:47 dialer1 kernel: [ 1046.863902] wanpipe2: Configuring DS DS26521 E1 FE
Jan 11 15:05:47 dialer1 kernel: [ 1046.863905] wanpipe2: Port 2,HDB3,CRC4,120OH
Jan 11 15:05:47 dialer1 kernel: [ 1046.863908] wanpipe2: Clk Normal:0, Channels: FFFFFFFF
Jan 11 15:05:47 dialer1 kernel: [ 1046.863910] wanpipe2: Sig Mode CCS
Jan 11 15:05:47 dialer1 kernel: [ 1046.867157] wanpipe2: Rx Sensitivity Gain 43dB (default).
Jan 11 15:05:47 dialer1 kernel: [ 1046.867218] wanpipe2: Enable transmit AIS alarm
Jan 11 15:05:47 dialer1 kernel: [ 1046.867664] wanpipe2: Front end successful
Jan 11 15:05:47 dialer1 kernel: [ 1046.867889] wanpipe2: Front End Interface Ready 0x40010000
Jan 11 15:05:47 dialer1 kernel: [ 1046.867893] wanpipe2: Failed to register device in HW Echo Canceller module!
Jan 11 15:05:47 dialer1 kernel: [ 1046.867897] wanpipe2: E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.868022] wanpipe2: E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.868030] wanpipe2: Enable transmit AIS alarm
Jan 11 15:05:47 dialer1 kernel: [ 1046.868051] wanpipe2: AFT communications disabled! (Dev Cnt: 1 Cause: Device Down)
Jan 11 15:05:47 dialer1 kernel: [ 1046.868059] wanpipe2: E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.868067] wanpipe2: Enable transmit AIS alarm
Jan 11 15:05:47 dialer1 kernel: [ 1046.868087] wanpipe2: AFT communications disabled! (Dev Cnt: 1 Cause: Device Down)
Jan 11 15:05:47 dialer1 kernel: [ 1046.868117] wanpipe2: TASKQ Not Running
Jan 11 15:05:47 dialer1 kernel: [ 1046.868120] wanpipe2: Global Chip Shutdown Usage=1, Used_type_cnt=1
Jan 11 15:05:47 dialer1 kernel: [ 1046.868123] wanpipe2: Global E1 Front End unconfigation!
Jan 11 15:05:47 dialer1 kernel: [ 1046.870163] wanpipe2: Master shutting down

by **williamconley** » Tue Jan 11, 2011 1:40 pm

Failed to register device in HW Echo Canceller module!

i guess you'll need to start lookin' that up.

by **Trying** » Wed Jan 12, 2011 3:20 am

I sent the errors through to Sangoma and will let you know what they have to say.

by **Trying** » Thu Jan 13, 2011 2:39 am

Just some interim feedback. Sangoma logged into our servers last night and they will be installing their driver from source tonight. They think it may be a driver issue. Unfortunately the time difference between Canada and South Africa means we can only communicate once a day.

They also noticed the following problem:

Are all your servers openSUSE
11.3? The reason I ask is because when I run: /proc/interrupt I notice
everything running off of one CPU, which might be of concern.

This is also the same on all 4 my dialers and they are all quad cores.

by **williamconley** » Thu Jan 13, 2011 11:43 am

use htop to determine if all processors are in use. i've not had any issue with it. in fact when i run massive sql packages to modify a lot of data ... i can watch the processor in use rotating as it moves along. (Moving 1 million records from one campaign to another, for instance)

by **andre** » Thu Jan 13, 2011 12:49 pm

similar issue with voicetime install (32b and 64b)

by **Trying** » Thu Jan 13, 2011 12:58 pm

I will report back. The Sangoma guys are working on the servers right now.

by **Trying** » Thu Jan 13, 2011 4:05 pm

Here is the feedback from Sangoma:

Okay, so it appears to be an rpm issue. I ran:
-> rpm -qa|grep dahdi

to find the dahdi rpm version and then downloaded the source for that dahdi version from the Vicidial rep. I installed our driver from source, and then pointed it to the source directory for dahdi (which our driver requires when you install from source) and installed perfectly. I also started up the driver and ran a few tested to confirm.

If you run 'wanpipemon -i w1g1 -c Ta' you will see LOF alarms, basically
because I did not configure the line settings properly.

I did not want to make the changes to all your other servers, because I will assume here that Vicidial will have to make new rpms.

For your reference, the source for Dahdi is located in:
-> /usr/src/packages/BUILD/dahdi-linux-2.4.0

and I installed the driver in:
-> /usr/src

I installed the driver using the command './Setup install' from the wanpipe
directory.

Don't be worries about what I had indicated before about all the interrupts
working off one CPU, it has no significance here.

Keep me in the loop if you want more information or need my assistance.

Summary: wanpipe rpm's from Vicidial appear to cause the issue. Installing
our driver from source with corresponding dahdi source resolve the issue

by **Kumba** » Fri Jan 14, 2011 4:30 pm

Wanpipe Voicetime modules load just fine for me. I cannot duplicate that issue therefore cannot fix it. Please open a new thread if you are having voicetime issues.

As far as wanpipe goes, I do not know why it's not loading. Sangoma was not very forthcoming with much assistance.

by **Kumba** » Fri Jan 14, 2011 5:49 pm

Turns out sangoma hard-coded a lot of their calls to the kernel modules and I didn't have a few in the right place. Re-read this post and do the sangoma drivers stuff again:

http://www.vicidial.org/VICIDIALforum/v ... hp?t=15336

It will load after that.

by **Trying** » Sat Jan 15, 2011 2:38 am

Thanks Kumba! I will study it and give it a go when I am back at the office.

by **Trying** » Sat Jan 15, 2011 3:46 am

I am happy to confirm that this indeed sorts out the problem!

vicidial.org

Sangoma Card Not Found + Asterisk not started

Who is online