Home » Forums » Vendor Forums - Non-Certified » Rhino Equipment

Kernel Panic faults on Rhino R4FXO-EC

thefunksterdeluxe
Posts: 29
 Member Since:
2010-01-19
Submitted by thefunksterdeluxe on Thu, 03/25/2010 - 6:17am.

I have a TB 2.8 and just installed 2 R4FXO cards. I enabled them with yum, setup-rhino & setup-pstn. Everything seemed to be working.

But intermittently, the server would lock up. When I went through debug, I noticed this happening every time I hit the 4th or 8th channel. I have unplugged the analog lines for now from those two (one on each card), but I thought they would work the same as the other ports on the card.

These two are spaced differently (a little further away) from the working ports on the card. Any idea what is going on with these?

Also - Rhino support had me build the driver from their ftp source but it didn't help.

Thanks

‹ Trixbox 2.2.12 OR 2.4.X with Rhino RCB8Fxx Line not hanging Up selling rhino PCI r2t1+ec for $350, like new! ›


jfinstrom
Posts: 2008
 Member Since:
2007-03-07
We were a bit unclear about
Thu, 03/25/2010 - 8:16am

We were a bit unclear about the issue. I had worked with Andrew (of trixbox) who had a system panicking on boot and a source build fixed. The described issue was simply 2.8 was panicking it which sounded the same. With the provided details here I wouldn't have gone in that direction as a first step necessarily. That said I am not sure why the engineer placed the ports that way I imagine as a point of reference when describing the port numbers to an end user. The ports them self are identical there is no special reason for that gap. The card it self impersonates a Digital (T1) card to DAHDI so the communication to DAHDI is the same on all 4 channels. The channel 4/8 thing is likely a simple coincidence. There was a weird counter issue in a former driver that never made public release where the DAHDI changed channel counts from 0 based to 1 based numbering but if you are running the latest driver build this would not be an issue. If you can provide the surrounding logs or provide an SSH connection through out official support channels we can do some more digging and see if we can narrow this down.

Regards,

--



thefunksterdeluxe
Posts: 29
 Member Since:
2010-01-19
logs
Thu, 03/25/2010 - 8:43am

I just plugged those two analog lines back in and tested. My first call triggered the failure:

from /var/log/asterisk/full:
[Mar 25 04:02:06] VERBOSE[5535] logger.c: == Parsing '/etc/asterisk/logger.conf': [Mar 25 04:02:06] VERBOSE[5535] logger.c: == Found
[Mar 25 04:02:06] VERBOSE[5535] logger.c: Asterisk Event Logger restarted
[Mar 25 04:02:06] VERBOSE[5535] logger.c: Asterisk Queue Logger restarted
[Mar 25 08:25:29] VERBOSE[8835] logger.c: -- Starting simple switch on 'DAHDI/8-1'
[Mar 25 08:29:03] VERBOSE[3381] logger.c: Asterisk Event Logger Started /var/log/asterisk/event_log
[Mar 25 08:29:03] VERBOSE[3381] logger.c: Asterisk Dynamic Loader Starting:
[Mar 25 08:29:03] VERBOSE[3381] logger.c: == Parsing '/etc/asterisk/modules.conf': [Mar 25 08:29:03] VERBOSE[3381] logger.c: == Found
[Mar 25 08:29:03] VERBOSE[3381] logger.c: == Parsing '/etc/asterisk/dnsmgr.conf': [Mar 25 08:29:03] VERBOSE[3381] logger.c: == Found
[Mar 25 08:29:03] VERBOSE[3381] logger.c: == Parsing '/etc/asterisk/http.conf': [Mar 25 08:29:03] VERBOSE[3381] logger.c: == Found

It hit DAHDI 8-1 and I have Linux configured to reboot when seeing a Kernel Panic so I don't need to hard reset it.

I don't know if it matters, but the analog lines going in to the Rhino cards are supplied by an Avaya PBX because my Asterisk/Rhino gear is serving as a POC lab. They are set up as outbound modem lines. I swapped some of them before plugging them back in to see if it was a line problem, which it doesn't appear to be.

Thanks



thefunksterdeluxe
Posts: 29
 Member Since:
2010-01-19
Found a solution
Thu, 03/25/2010 - 12:02pm

http://www.trixbox.org/forums/trixbox-forums/help/kernel-panic-28...

gem7 had the same problem on a 4 port card, calling the last port would cause kernel panics. He set 'echocancel=no' in chan_dahdi.conf

I just did this and it worked, though I don't understand the implications.

Thanks everyone.



poison
Posts: 46
 Member Since:
2006-08-04
Call rhino
Fri, 03/26/2010 - 6:25am

Log everything and call Rhino. If you give them the right info they are usually very helpful in fixing anything causing a problem. I had a problem once with one of their PRI cards that caused a panic and they were able to figure it out quickly. I would think its a driver from their site that may fix this. I would not think turning echocancel=no is a good idea.



thefunksterdeluxe
Posts: 29
 Member Since:
2010-01-19
Yeah
Tue, 03/30/2010 - 11:50am

I tried that. No dice. Part of it is on me because my employer requires an Act of Congress to get inward access so I can't provide ssh to the helpdesk.

This is a lab/proof-of-concept server so I have no issue with leaving EC disabled. But I did find out yesterday that the trixbox appliance I'm looking at comes preloaded with Sangoma modules so I will reapply my focus elsewhere. Thanks for James, poison & and all who tried assisting.



Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.