Re-scanning for available devices/connections after failure

[Daenyth Blank <daenyth gmail com>]

Hi all,

I came to the list earlier asking what card would be the best for a
remote deployment and ended up following your suggestion of the Sierra
598U, and I'm having a small issue with the card sometimes. Yesterday
one of the deployed machines went offline and didn't reconnect until
rebooted. When I examined the logs, networkmanager seemed to think
that the card had become disconnected, then tried to reload it,
failed, and then stayed offline from there. It's actually offline
again right now so I can't provide much of the logs. Here's some that
may be slightly out of order pulled from my scrollback:
http://gist.github.com/461679

Since these machines are remotely deployed, it's not always possible
to manually trigger a reconnect. What options do I have if NM doesn't
attempt to automatically reconnect? Currently I'm thinking I might put
in a cron job that will check connectivity and restart nm if not
connected. Is there a cleaner way to make it re-scan for available
devices/connections?

Thanks,