| Does anyone know of a way, outside of Asterisk, to monitor the health of | those devices?
As Andrew pointed out, the only certain way to ensure that a service is working is to actually exercise the service. I have a mechanism that calls from one system to another, hits a particular extension which writes a timestamp file in /tmp, and then nagios comes along and looks to make sure that the timestamp file is recent. Which doesn't prove that *everything* is working of course. And it's easier to do with PRIs and multiple DIDs. Some months ago, we ran into some problems on our conference bridge, and "solved" them (I'm ashamed to say) with a nightly "restart when convenient" (it seemed to be some sort of corruption over time problem). Can you look at the CDRs periodically and note unused channels? Automatically update the software every night? Power cycle the machine at random times during the day to make sure it's happy? Have the customer schedule a daily conference call that just happens to use you all available channels, and wait for someone to complain they couldn't get it? John
