Add initial support for synapse workers

· needs documentation; no checks yet for port clashes or typos in worker name
· according to https://github.com/matrix-org/synapse/wiki/Workers-setup-with-nginx#results
  about 90% of requests go to the synchrotron endpoint
· thus, the synchrotron worker is especially suited to be load-balanced
· most of the other workers are documented to support only a single instance
· https://github.com/matrix-org/synapse/blob/master/docs/workers.md

roles/matrix-synapse/templates/synapse/systemd/matrix-synapse-worker@.service.j2

@@ -0,0 +1,29 @@
+#jinja2: lstrip_blocks: "True"
+# c.f. https://github.com/matrix-org/synapse/pull/4662
+[Unit]
+Description=Synapse Matrix Worker
+After=matrix-synapse.service
+BindsTo=matrix-synapse.service
+
+[Service]
+Type=simple
+
+# Intentional delay, so that the homeserver (we likely depend on) can manage to start.
+ExecStartPre=/bin/sleep 5
+
+# systemd ftw 🤦‍♂️
+# https://github.com/systemd/systemd/issues/14895#issuecomment-594123923
+ExecStart=/bin/sh -c "WORKER=%i; WORKER=$${WORKER%%:*}; \
+			exec /usr/bin/docker exec \
+			--user={{ matrix_user_uid }}:{{ matrix_user_gid }} \
+			matrix-synapse \
+			python -m synapse.app.$${WORKER} -c /data/homeserver.yaml -c /data/worker.%i.yaml"
+
+ExecStop=/usr/bin/docker exec matrix-synapse pkill -f %i
+KillMode=process
+Restart=always
+RestartSec=10
+SyslogIdentifier=matrix-synapse-%i
+
+[Install]
+WantedBy=matrix-synapse.service