[patio11]

Hosts requiring a reboot due to issues with the underlying physical hardware are inevitable for VMs. It happens to me approximately once per instance-year on Rackspace (and before that, Slicehost). [Edit: Slightly better approximation: ~5 incidents in the last ~20 instance-years.]

The general level of seriousness with regards to operations is not constant across all VM providers. Let's put it this way: there exist companies where the heart-and-soul of the business is supporting $5 a month hobbyists and there are companies where the heart-and-soul of the business is business-grade services. Choose appropriately.

[benguild]

This. Some VPS providers don't care at all about downtime since the services are so cheap. Then, scaling that cheap service to hundreds of dollars a month for more resources doesn't necessarily guarantee any increase in quality... maybe just an increase in attention if you make a fuss