[atlas] Very short uptime for some probes

Hi all.

I've only been on this list for a relatively short while so this is
most likely in an FAQ I should be digging thru...

I've noticed that the uptime for some of the physical probes we've
installed are very short. According to the control panel, one in
particular has connect times of around 2-3 minutes and disconnect
times of 4-6 minutes. This cycle is repeated every 10 minutes or
so. All day, every day.

Superficially that sort of pattern suggests some sort of continuous
reboot loop.

However, if I ping the probe remotely, I get just an occasional
packet loss. Certainly no long gaps that you'd expect from a
continuous reboot. That I can ping it remotely also suggests that
external connectivity isn't the problem. Besides the periodicity is
just too regular and frequent for a routing loss or similar.

(In any event, if the network it's on is losing connectivity every 5
minutes I would have plenty of other symptoms.)

The problem probe is also reported as getting UDMs assigned and
running them so I'm not quite sure how it manages to squeeze those
tests in if that connectivity status report is accurate.

So my main question is, should I be concerned about the control panel
connectivity reports? And if I should be, is there much I can do
beyond ensuring the device has good connectivity?


Mark.