This is a two-part issue.
_Part 1:_
There could be an option to schedule the auto-kill feature based on a calendar time, instead of a minute-based interval. So as to provide a routine 'clean-up' mechanism.
_Part 2:_
Instead of a timer based auto-kill feature, provide monitoring rules. So that, whenever any process [within a specially monitored hitlist] matches these rules, it gets terminated automatically.
An example of such rules would be:
- Rule 1: If status remains 'Not Responding' for 15 seconds, then terminate it
- Rule 2: If memory consumption is more than 8% for 10 seconds, then terminate it.
and so on...
This is a two-part issue.
_Part 1:_
There could be an option to schedule the auto-kill feature based on a calendar time, instead of a minute-based interval. So as to provide a routine 'clean-up' mechanism.
_Part 2:_
Instead of a timer based auto-kill feature, provide monitoring rules. So that, whenever any process [within a specially monitored hitlist] matches these rules, it gets terminated automatically.
An example of such rules would be:
and so on...