Prevent shutdown while creating a time-based APT snapshot

Time-based APT snapshots for the debian archive were broken since 2 days because apparently apt.lizard was shut down while it was creating a snapshot. Fixing this required manual intervention by someone familiar with this piece of infrastructure, aka. myself. Let's make this more robust.

I see 2 options:

• A. Wrap the commands run in the cronjobs with systemd-inhibit.
  • advantage: should be cheap and simple
  • Option A is ruled out, see the discussion below.
• B. Migrate the cronjobs to systemd templated timers+services, with appropriate configuration to get the "stop" behavior we want; ensure we're aware of failure of such services
  • advantage: also allows anyone with access to easily (re-)run a specific failed snapshot attempt; currently I'm the only one who knows how to do that; this is needed every few weeks

I've been wanting to do B for the aforementioned unrelated reason for years, so perhaps it's a good time to do that. But A is probably a good first step anyway, and then B can be tracked elsewhere.

Note: For S11, this fits in:

• D.1 - Make Tails development process more efficient: Broken time-based APT snapshots impacts the creation of ISO images on our build system and thus the whole development process.

From FT's perspective

Looks like it’s not as much of a LHF as what we initially thought. The problem happened only once, 2 years ago. So low priority.