Fix ansible handler timeouts for alloy and loki restarts

Use async with poll: 0 to fire-and-forget service restarts.
These services have graceful shutdown periods that can exceed
ansible's default command timeout.

Also update transmission_metrics and devpi_metrics to depend on
alloy instead of node_exporter, since alloy now handles the
textfile collector.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

ansible/roles/alloy/handlers/main.yml

@@ -1,3 +1,5 @@
 ---
 - name: restart alloy
   ansible.builtin.command: brew services restart grafana-alloy
+  async: 120
+  poll: 0

ansible/roles/devpi_metrics/meta/main.yml

@@ -1,4 +1,4 @@
 ---
 dependencies:
-  - role: node_exporter
+  - role: alloy
   - role: devpi

ansible/roles/loki/handlers/main.yml

@@ -1,3 +1,5 @@
 ---
 - name: restart loki
   ansible.builtin.command: brew services restart loki
+  async: 120
+  poll: 0

ansible/roles/transmission_metrics/meta/main.yml

@@ -1,4 +1,4 @@
 ---
 dependencies:
-  - role: node_exporter
+  - role: alloy
   - role: transmission