[Asterisk-Users] Stability and motherboard questions with TE406P and TE410P

Kyle Sexton ks at mocker.org
Wed Mar 22 14:50:43 MST 2006


I am having a problem with asterisk not being stable enough for production
use.  I have two cards, the digium TE406P, and the TE410P.  The TE410P is
the primary card that I am using but I would like to move to the TE406P for
the echo cancellation and more flexibility of PCI slots available.

General config info:

3 PRIs, configured as such:

span=1,2,0,esf,b8zs
span=2,0,0,esf,b8zs
span=3,1,0,esf,b8zs
span=4,0,0,esf,b8zs

I have modified the 2 and 1 timing source declarations to several different
configurations but am unsure what the proper way to configure that is.  Is
my current setup correct?  Both span one and span three have timing on the
line from the carrier.

Rundown of issues seen with these cards:

TE410P:
- zttest will never report 100% for me across different motherboards
(Supermicro P8SCT, Dell 850)
- Crash/instability of about once per two weeks where I have to power cycle
the server, i.e. phone calls stop working and a reboot fixes it


TE406P:
- zttest runs flawlessly on this card, 100% across the board
- PRIs will go from up and working fine, to "Provisioned, Down, Active"
after the server has been up for around 10 minutes, this may be related to
rxfax and txfax being installed?  Has anyone had an issue with this specific
card?  We have had this experience across multiple motherboards.

After working with two different cards and across multiple motherboards I am
starting to lose faith on the stability of the Asterisk platform, but I know
others are having lots of success.  My boss is looking for something that
says "With the TE406P, we have had zero issues on X motherboard", does
anyone have any recommendations?  Has anyone else had stability issues with
the digium 4 port cards?

Thanks,
Kyle
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.digium.com/pipermail/asterisk-users/attachments/20060322/dd22275d/attachment.htm


More information about the asterisk-users mailing list