Journald might keep logs longer than we want to

We have a central Loki log server to which our grid sends all its log data, and it's working well enough. Grid nodes should only have very little, or maybe no, local log storage.

Yet, there's hundreds of MB in the staging grid /var/log/journal directories.

TBD: Investigate if we yet have to apply our options.services.private-storage.monitoring.policy.logRetentionSeconds to local journal storage.