People often report and ask about these "failures".
More-so previously, when the `docker kill/rm` output was collected,
but it still happens now when people do `systemctl status
matrix-something` and notice that it says "FAILURE".

Suppressing to avoid further time being wasted on saying "this is
expected".

Reverts b1b4ba501fdfaa, 90c9801c560b6, a3c84f78ca9c65a, ..

I haven't really traced it (yet), but on some servers, I'm observing
`ansible-playbook ... --tags=start` completing very slowly, waiting
to stop services. I can't reproduce this on all Matrix servers I manage.
I suspect that either the systemd version is to blame or that some
specific service is not responding well to some `docker kill/rm` command.

`ExecStop` seems to work great in all cases and it's what we've been
using for a very long time, so I'm reverting to that.

This is the equivalent of b1b4ba501f for the matrix-dendrite role.

This brings matrix-dendrite.service in line with all the other services.

All the `matrix_s3_*` stuff happens in the `matrix-synapse` role.

If we are to have such S3 support for Dendrite, we should probably
extract it out of the `matrix-synapse` role (into a `matrix-s3` role or
`matrix-goofys`, etc.) and wire `matrix-dendrite` accordingly.

This may or may not be done in the future though. For now, I'm
cleaning things up in the `matrix-dendrite` role.

We refrain from logging to files for all components,
because we rely on systemd-journald anyway.