1. 15

  2. 1

    You can make transient jobs with systemd really quickly at the command line. You can probably get it to save those configs to a file too.

    1. 1

      I use Nomad[0] for job scheduling. It’s distributed, cross platform, and can handle all sorts of runtimes. It’s a pretty simple tool for everything it does. The job config files are declarative, which is great, but they can be a little complicated.

      0: https://www.nomadproject.io/

      1. 1

        I was thinking about this as well. My current solution is to drop scripts into /etc/cron.{hourly,daily,weekly}. Additionally https://healthchecks.io/ monitors it through some curl lines in the scripts.

        For example, here is my Pi-Hole-like script:

        curl --retry 3 https://hc-ping.com/e0997eb0-and-more/start
        set -e
        # Use https://github.com/StevenBlack/hosts
        curl "$SRC" -o /etc/dnsmasq-hostsdir/adblock
        curl --retry 3 https://hc-ping.com/e0997eb0-and-more