to deal with docker memory and resource leaks that are really hard to
debug, periodically kill docker processes and even reboot the entire
machine ...

---
Example root crontab:

$ sudo crontab -l
0 9 * * * /sbin/shutdown -r now
0 21 * * * /sbin/shutdown -r now
0 4 15 */2 * /home/pgbovine/letsencrypt/certbot-auto renew


Example user crontab to periodically flush away docker containers:

$ crontab -l
@reboot /home/pgbovine/linode-forever-processes.sh
*/5 * * * * docker rm -f $(docker ps -aq)

# apparently this command removes only *stopped* containers:
*/5 * * * * docker rm $(docker ps -a -q)

# very aggressively remove long-running containers ...
@reboot /home/pgbovine/linode-forever-processes.sh
*/1 * * * * /home/pgbovine/OnlinePythonTutor/v4-cokapi/docker-rm-long-processes.py >> /home/pgbovine/docker-rm.out 2>> /home/pgbovine/docker-rm.err