[Asterisk-Users] Power Alarm Error - Help

[Rich Adamson]

> > I had a similar problem with power.
> > I connected Asterisk to a Belkin UPS 1200VA and the the server would boot up
> > and asterisk would load but the T1s on the Quad T1 card failed to come up. I
> > placed a loop on the card and still no change. Finally, I removed the UPS
> > and the T1s came up.
> > 
> > Do know if this will help you, but the T1 card seems to be delicate with
> > power.
> 
> I've been having similar trouble with one of my units.  I put a UPS on the
> system and it seemed to get better, but how that module fails regularly.
> 
> Incidentally, when a module dies and holds the circuit open, I can top
> asterisk, unload the kernel modules, reload them, run ztcfg, and restart * w/o
> restarting the system.  However, it does require * to stop and that is
> annoying.
> 
> I moved my problematic phone to a different fxs module and all seems fine.  We
> have a similar problem with a TDM/FXS module in a different location.
> 
> I've written digium support, but they are kinda slow in responding.

It sounds like the same problem that many of us have seen over the last
several months. Digium support has been trying to identify the root cause
and has been very quiet in terms of comments, etc. Replacing the little
fxo/fxs module has apparently corrected the problem in some cases, but
its fairly obvious the tdm card still has a problem that has not yet
been diagnosed/fixed. Dropping/restarting the drivers is the only known
fix (bypass) thus far.

If you look back over the previous posts relative to tdm problems, you'll
see lots of comments about ups's, power supplies, pstn lines, etc, etc. 
The majority of these have indicated the problem takes a fair amount of 
time (eg, days or weeks) before recurring, suggesting the problem might 
be related to a memory leak, temperature, or some other cause that 
involves 'time'.

Shutting down an * system to install a ups, etc, disturbs just about
everything associated with the intermitant problem. Jumping to a 
conclusion that a ups (etc) fixed a problem is essentially ignoring
all of the other disruptions that happened during the ups install.