implementing load mitigation logics for tracee-ebpf

[mtcherni95]

[NDStrahilevitz]

Just a thought, we can export the current threshold as a metric in Stats, this way we can track the health probe's action in prometheus.

[rafaeldtinoco]

@mtcherni95 and @NDStrahilevitz.. we've been discussing multiple things and ideas about fixing userland filtering (first by bringing the values from tracee-rules signatures to the current filtering mechanisms we have, and then, later, moving things to eBPF filtering). Should we close this for now and re-open once we have a clear picture ? Either way I think Nadav can use this as a base for his PRs on filtering improvements.

[mtcherni95]

I believe we should either keep it or integrate this PR in another one. Anyways I think that the event pressure engine according to priorities is something we do want to have in Tracee. IMHO I wouldn't mind to keep this PR open, refinish it and push it. By default the feature could temporarily be set to off.

[yanivagman]

can be linked to #1343

[rafaeldtinoco]

Closing this to inactivity.