sushy_emulator: healthcheck + restart on failure

In adoption jobs the sushy_emulator stops working somewhere in
the job run. We see errors raised from python libvirt library:

- `libvirt: XML-RPC error : Cannot write data: Broken pipe`
- `libvirt: XML-RPC error : internal error: client socket is closed`

This change adds a healthcheck to the podman pod, probe the service
every 30 seconds, and trigger a container restart on two failures.

Jira: OSPRH-15686

Author: hjensas

roles/sushy_emulator/tasks/create_container.yml

@@ -41,3 +41,12 @@
       - "{{ dest_dir }}/known_hosts:/root/.ssh/known_hosts:ro,Z"
       - "{{ cifmw_sushy_emulator_sshkey_path }}:/root/.ssh/id_rsa:ro,Z"
       - "{{ cifmw_sushy_emulator_sshkey_path }}.pub:/root/.ssh/id_rsa.pub:ro,Z"
+    healthcheck: >-
+      python3 -c "import urllib.request, base64;
+      req = urllib.request.Request('http://localhost:8000/redfish/v1/Systems/{{ _cifmw_sushy_emulator_instances[0] }}');
+      req.add_header('Authorization', 'Basic ' + base64.b64encode(b'{{ cifmw_sushy_emulator_redfish_username }}:{{ cifmw_sushy_emulator_redfish_password }}').decode());
+      urllib.request.urlopen(req, timeout=5).read()"
+    healthcheck_interval: "30s"
+    healthcheck_timeout: "30s"
+    healthcheck_retries: 2
+    healthcheck_failure_action: restart