>_ IzzyLab

Topic breakdown

Containers usually have no systemd, so the role of the supervisor that keeps a daemon alive and brings it back up after a crash is taken on by supervisor. A program is described in a separate .conf file in /etc/supervisor/conf.d/: the [program:NAME] section sets the name the program is visible under in supervisorctl, and inside it you specify what to run and enable auto-restart. After editing the configs, supervisor needs to be brought up (via service) and made to re-read the descriptions — supervisorctl has the reread and update subcommands for this; the status is checked with the status command.

What's checked

  1. supervisor keeps appd RUNNING
  2. appd restarts after a crash

What it looks like