fix: retry transient first boot ssh disconnects

ansible/site.yml

@@ -21,6 +21,23 @@
 
     - name: Gather facts after SSH is reachable
       setup:
+      register: initial_setup
+      ignore_errors: true
+      ignore_unreachable: true
+
+    - name: Clear transient SSH unreachable state after first fact gather
+      meta: clear_host_errors
+      when: initial_setup.unreachable | default(false)
+
+    - name: Wait for SSH after transient first-boot disconnect
+      wait_for_connection:
+        delay: 10
+        timeout: 300
+      when: initial_setup.unreachable | default(false) or initial_setup.failed | default(false)
+
+    - name: Gather facts after transient first-boot disconnect
+      setup:
+      when: initial_setup.unreachable | default(false) or initial_setup.failed | default(false)
 
   roles:
     - common