At the moment, the default behavior of the watcher is to ping an external server, and reboot if it's unresponsive for a set number of tries (currently set at 15 minutes). It turns out this is not desirable (we have other, better ways of testing network connectivity of devices), so in bug 778177 we are making an effort to configure the tegras not to do this. To avoid future confusion and update steps, I believe we should change the default behaviour of the watcher to not ping at all. This will be one less thing to check and worry about when a device in automation isn't working as expected.

Comment on attachment 648363 [details] [diff] [review] Default behaviour of watcher should be not to do connectivity ping at all Review of attachment 648363 [details] [diff] [review]: ----------------------------------------------------------------- very simple. No need to write a lot of code for this if we don't need to.