> Hmm. How about mounting /var/log from an MMC or CF or NFS location?
I don't have a wifistix-CF so that's not possible in the short term. I'm a bit concerned that sending the messages over the network via NFS might change things too much -- and if network connectivity turns out to be the issue it probably wouldn't help :-)
Is there an easy way to redirect the messages to the serial port? I've got a laptop connected to a unit in the field full time.
> As an alternative, use the watchdog driver to just restart if the
That's a bandaid I'll consider if it feels like it's going to take a while to get to the real issue. It would at least let me get more info on the rest of the system without having to do almost daily manual reboots!
I spent some time exploring how robust the wifistix is in the face of bad connectivity. It actually looked pretty good! Doing something as vile as removing the antenna during operation resulted in a graceful recovery once it was replaced. The driver was smart enough to turn off the yellow LED when the signal was lost due to missing antenna. Upon replacing the antenna the LED began flashing within a couple of seconds and connectivity was restored. Everything else I tried in the way of wireless torture was equally unsuccessful in reproducing my issue.
I'll keep looking. I'm going to move my bench setup (which has never failed) out into the field to see if the problem appears with a change of environment.
On May 10, 2007, at 7:52 AM, Steve Sakoman wrote:
> I attached a laptop with serial connection to one of the units last
> night in the hopes of being able to use a terminal session to see
> what is going wrong.
> The unit did fail overnight, but the serial connection proved to be
> of no use -- the system was wedged, no serial io :-( And of course
> the console message file disappeared into the ether upon reboot.
> The same bits have been running on another unit on my bench for
> more than two weeks without reboot, so I'm off to eliminate
> variables one by one. I swapped the bench and field units after
> the failure last night. The gods were with me and the swapped unit
> failed this morning in the same way, so that eliminates it being
> related to a particular unit and points to something in the
> environment (power supply, EMI, wifi signal strength,
> temperature, . . .)
> These intermittent failures are lots of fun :-)
Hmm. How about mounting /var/log from an MMC or CF or NFS location?
As an alternative, use the watchdog driver to just restart if the
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
gumstix-users mailing list