On 22.02.2019 10:11, Chris Smith via users wrote:
I certainly don’t need anything as complex as all
that. I’m not dealing with anything mission-critical as such and I don’t need it to work
seamlessly to the user either, I just need it to be automated. The more I think about
this, the more I realise that all I need is an automated way to detect a problem and
reboot the server. Ideally, this would happen from an external VPS or other system being
able to poke Xen (in case the server cannot restart itself) and ideally this would be tied
in to the existing BitFolk monitoring services (if possible) to avoid re-inventing the
wheel.
Have you considered watchdog? If likely problems would be detectable
from the VM then watchdog can run custom tests and trigger reboot or
attemt to fix things.
- OM