Use a better docker restart policy
Because we use docker run
for some interactive and one-off tasks, the
current restart policy is problematic. Using on-failure
instead of
unless-stopped
makes it so that successful commands don't get run
multiple times.