Otherwise Tahoe might start without the socket unit and not have access to the
socket that's bound to the port and so just fail to start.

Make Systemd open a socket for the Tahoe web server

Closes #122

See merge request !359

Change all Grafana dashboards to fixed UID data sources

See merge request !358

This reverts commit b8a88f35.

The current defaults already do not work.
Go back to set the dependencies manually.

It's safe to assume Systemd defaults are consistent and stay
the same.  Also, I know of a brilliant investment opportunity:
Mars is going to be colonized soon, and prices only can go up...

refs #122

Add UID option to Grafana Service

See merge request !357

Remove obsolete nixpkgs pin

See merge request !356

This comes from nixpgks commit 81291cc793cf88bd6eff3fd8512e5eb9d037066c
and will be included with nixos 22.11.

Refs !222#note_18600

Backup duration alert: Get unit state and trigger times from systemd

See merge request !355

... and use a smarter Prometheus query to combine the two.

Give it TLS certificates

See merge request !354

Otherwise it can't validate TLS certificates when it is trying to download
dependencies.

Put nix in the shell environment

See merge request !353

Previously we got nix from the host env without even realizing it.  As soon as
it upgraded things broke.

We don't want to inherit nix from the host environment because who knows if it
is compatible with our software or not.

Monitoring: Fix backup duration alert

See merge request !349

This should implement my actual intentions: Alert when
backups run for longer than 3h, and the the repo check
for more than 6h.

Add dashboards for Tahoe Incident Report count and rate

Closes privatestorageops#288

See merge request !347

Publish number of Tahoe Incident Reports

Closes privatestorageops#287

See merge request !346

Fixes privatestorageops#287

Also alert when hosts that run ZFS run out of RAM

See merge request !345

Prometheus should keep metrics as long as Loki keeps logs

See merge request !344

refs #122

Forgot to add a second alert when I added the workaround to
not count the ZFS ARC into used memory :/

... which is both governed by our retention policy.

Monitoring: Count Tahoe's corruption advisories

Closes privatestorageops#287

See merge request !341

Monitoring: Add Tahoe-LAFS corruption advisory count + rate + alert on rate > 0

Closes privatestorageops#288

See merge request !342