This is my latest version of this, updated to work with the
packages in NixOS 21.05.

Note that changing the origin repository in the on-node deployment checkout
is also required.

See https://whetstone.private.storage/privatestorage/privatestorageops/-/merge_requests/197#note_19071

Source: https://whetstone.privatestorage.io/privatestorage/PrivateStorageio/-/merge_requests/237#note_18712

Thanks to @jcalderone for the suggestion!

The current code evaluates our custom packages once for each node, which adds
signifcant amount of time to evaluate a grid. We can reduce this, by adding the
custom package set as an attribute to the nixpkgs set we pass to morph.

This doesn't change how we refer to those packages, as we continue to expose the
custom package set as a module attribute.

These are the times to evaluate all three grids (on a partially loaded system),
when there was nothing new to build:

```
before:
real	2m27.837s
user	3m35.528s
sys	0m3.722s

after:
real	1m12.748s
user	1m34.047s
sys	0m3.346s
```

Fixes privatestorageops#408

This also update ZKAPAuthorizer to a python3 using version.

This is disallowed in nixos 21.11 as it is a potential security issue.

Take what I learned from @tomprince in
https://whetstone.privatestorage.io/privatestorage/PrivateStorageio/-/merge_requests/220#note_17605 :
Assigning a service a `startAt' will make timer.target start it,
no need to also have it `wantedBy' multiuser.target.

Also add megacli2prom to the path even if it should be already because
of megacli2prom.  Better be safe than sorry.

mkfs silently truncates "zkapissuer-database" to "zkapissuer-datab" and
everything falls apart.

The point is this is where PaymentServer's persistent state goes.  That
includes vouchers but might include more stuff too.

The hardware modules can now configure this fileSystem in a way appropriate
for themselves and the issuer module can enable that configuration when it is
enabled itself.

now all nodes load the issuer module so we can always supply issuer
configuration, where it makes sense to do so, even on systems where the issuer
is not going to be enabled.

if that sounds silly read the comment in default.nix

use the new one we're setting up on a new filesystem

Thanks go out to @tomprince!

Tryfix the starting of the timer before tahoe is updated.

This adds a detailed dashboard for tahoe-lafs running on one node.

Thanks to @tomprince for the suggestion.