Your box was one of the units that was affected by a bug we had in some monitoring. Ping checks were failing spontaneously, which is not generally an issue, but each failure was resulting in a leaking file descriptor. Once the number of leaked file descriptors exceeded the OS limit this triggered a restart. This was restarting your box a few times per day.
We patched this issue last night, but your device has automatic updates disabled (we can change that) and I did not realise that your box missed the update until it showed up on my monitoring this morning. So I triggered the update to prevent further bug related restarts over the day. Sorry for not giving you heads up regarding this, how rude huh!!!
We have had some issues perfecting our update process which has meant we have historically pushed updates out much earlier than we perhaps should be so we can monitor them. But the last 4 updates have been 100% successful (WAAAHOOOO) which means we have got the transactional update process downpacked in SphireOS so we can start triggering updates at 2am local time.
For critical updates that just have to go out, we can announce them if we have to push them out during the day.