Flintlock should know when a mVM dies

[Callisto13]

What happened:
An unrelated config issue caused the firecracker process to die shortly after boot. It appeared that flintlock did not notice this (at least according to the logs).

What did you expect to happen:
Flintlock should notice that an mVM had died, and it should at least log that this has happened.

How to reproduce it:
Going off this commit (nothing special about it, just head of main rn and happens to contain an issue. you could also repro with a new commit and manually kill the pid probably) create a mVM with the config in hack/scripts/payload/CreateMicroVM.json.

You should see the firecracker process start, then after a few moments die. The firecracker logs will talk about no init found, but that is not important here. The issue is that flintlock did not appear to notice.

Anything else you would like to add:
We do not have Get merged right now, so curious to know what the status knows.

Environment:

• flintlock version:
• OS (e.g. from /etc/os-release):

[richardcase]

We also need to think about a unresponsive vm.