I'm planning to run/have some quite sensitive data on a VPS, however, knowing well that the host could at any time access the data (rare but as long as it could happen in theory, I would very much like to have some sort of contingency planning down). Could it be possible to practice some sort of "damage control"?
For example, the method mentioned in: https://www.howtogeek.com/123568/how-to-get-email-notifications-whenever-someone-logs-into-your-computer/
Would it be enough, as one would be notified whenever someone else accesses the server? Or can a host easily bypass it regardless? For example via a "maintenance" and then extraction, where internet would be off?
Although since most VPS providers claim 99% up-time and assuming that an "evil host" would probably initially check what's on a machine through internet first, inevitably having to log in. Would this be an okay method regardless?
Unless of course they have a very elaborate and near-impossible to detect monitoring software installed on every machine, which for example, occasionally takes screenshots, giving the provider information on which machine could be "interesting" to inspect further... although could this also, simply be combated by a re-formatting before usage?
Looking forward to any and all feedback