Users are often scared to leave cloud instances running which can hamper cloud adoption. Froster has an experimental idle detection which lets ec2 instances self-terminate after a few days (72 hours by default). It has a monitoring cron job in systemd that is run hourly and if this cron job has detected an idle state 72 consecutive times the machine will be terminated.
Is this a reasonable approach for Amazon Linux 2003?
Users are often scared to leave cloud instances running which can hamper cloud adoption. Froster has an experimental idle detection which lets ec2 instances self-terminate after a few days (72 hours by default). It has a monitoring cron job in systemd that is run hourly and if this cron job has detected an idle state 72 consecutive times the machine will be terminated.
Is this a reasonable approach for Amazon Linux 2003?
function AWSBoto._monitor_is_idle: https://github.com/dirkpetersen/froster/blob/main/froster.py#L3915