Merge pull request #892 from stackhpc/upstream/2026.1-2026-06-29

Synchronise 2026.1 with upstream

Author: Alex-Welsh

ansible/library/kolla_toolbox.py

@@ -382,7 +382,10 @@ def main(self) -> None:
         try:
             self._push_private_data_dir(kolla_toolbox, pdd, user)
 
+            exec_user = user or 'ansible'
             runner_env = {}
+            if exec_user == 'ansible':
+                runner_env['HOME'] = _PDD_BASEDIR
             if self.module._diff:
                 runner_env['ANSIBLE_DIFF_MODE'] = '1'
 
@@ -391,7 +394,7 @@ def main(self) -> None:
                 ['/opt/ansible/bin/ansible-runner', 'run', pdd,
                  '--playbook', 'main.json',
                  '--rotate-artifacts', '1'],
-                user=user,
+                user=exec_user,
                 environment=runner_env
             )
             # exit 2 = task failed/unreachable; handled via event below.

ansible/roles/barbican/defaults/main.yml

@@ -109,7 +109,7 @@ barbican_api_healthcheck_test:
   - >-
     healthcheck_curl {{
     'https' if barbican_enable_tls_backend | bool else 'http' }}://{{
-    api_interface_address | put_address_in_context('url') }}:{{ barbican_api_listen_port }}/healthcheck"
+    api_interface_address | put_address_in_context('url') }}:{{ barbican_api_listen_port }}/healthcheck
 barbican_api_healthcheck_timeout: "{{ default_container_healthcheck_timeout }}"
 barbican_api_healthcheck:
   interval: "{{ barbican_api_healthcheck_interval }}"

ansible/roles/nova-cell/tasks/version-check.yml

@@ -75,3 +75,47 @@
         success_msg: >
           Libvirt version check successful: target {{ new_version }} >= current {{ current_version }}.
       when: libvirt_version_current_results is not skipped
+
+- name: Check nova_libvirt image provides the libvirt start script
+  when: enable_nova_libvirt_container | bool and (groups[service.group] | length) > 0
+  vars:
+    service_name: "nova-libvirt"
+    service: "{{ nova_cell_services[service_name] }}"
+  tags: nova-libvirt-script-check
+  block:
+    - name: Look for kolla_nova_libvirt_start in the nova_libvirt image
+      become: true
+      kolla_container:
+        action: "start_container"
+        command: "ls /usr/local/bin/kolla_nova_libvirt_start"
+        common_options: "{{ docker_common_options }}"
+        container_engine: "{{ kolla_container_engine }}"
+        detach: false
+        environment:
+          KOLLA_CONFIG_STRATEGY: "{{ config_strategy }}"
+        image: "{{ service.image }}"
+        name: "nova_libvirt_script_check"
+        restart_policy: oneshot
+        remove_on_exit: true
+      register: nova_libvirt_script_check
+      failed_when: false
+      check_mode: false
+      run_once: true
+      delegate_to: "{{ groups[service.group] | first }}"
+
+    - name: Assert the nova_libvirt image is recent enough
+      run_once: true
+      delegate_to: "{{ groups[service.group] | first }}"
+      ansible.builtin.assert:
+        that:
+          - "'/usr/local/bin/kolla_nova_libvirt_start' in (nova_libvirt_script_check.stdout | default(''))"
+        fail_msg: >
+          The nova_libvirt image '{{ service.image }}' does not contain
+          /usr/local/bin/kolla_nova_libvirt_start, which this version of
+          Kolla-Ansible needs to launch libvirtd alongside virtlogd. This
+          almost always means the container image has not been updated to
+          match the upgraded Kolla-Ansible code. Pull or build a current
+          nova-libvirt image before deploying. If you intend to keep using
+          an older image, skip the tag 'nova-libvirt-script-check'.
+        success_msg: >
+          nova_libvirt image provides kolla_nova_libvirt_start.

ansible/roles/nova-cell/templates/nova-libvirt.json.j2

@@ -1,5 +1,5 @@
 {
-    "command": "/usr/sbin/libvirtd --listen",
+    "command": "/usr/local/bin/kolla_nova_libvirt_start",
     "config_files": [
         {
             "source": "{{ container_config_directory }}/libvirtd.conf",

ansible/roles/nova-cell/templates/qemu.conf.j2

@@ -1,4 +1,4 @@
-stdio_handler = "file"
+stdio_handler = "logd"
 
 user = "nova"
 group = "nova"

releasenotes/notes/bug-2157639-93e313f168b18133.yaml

@@ -0,0 +1,6 @@
+---
+fixes:
+  - |
+    Fixed a stray trailing quote in the ``barbican_api`` healthcheck
+    command that caused containers to be falsely reported as unhealthy.
+    `LP#2157639 <https://launchpad.net/bugs/2157639>`__

releasenotes/notes/nova-libvirt-script-precheck-7c9f2a1b3d4e5f60.yaml

@@ -0,0 +1,11 @@
+---
+upgrade:
+  - |
+    A deploy-time check now verifies that the ``nova_libvirt`` image contains
+    the ``/usr/local/bin/kolla_nova_libvirt_start`` helper before deploying.
+    This helper launches ``libvirtd`` alongside ``virtlogd`` and is required
+    by current Kolla-Ansible. If you upgrade Kolla-Ansible but continue to
+    use an older ``nova_libvirt`` image that predates the helper, the deploy
+    now fails fast with an explanatory message instead of leaving the
+    container unable to start. Pull or build a matching ``nova_libvirt``
+    image, or skip the ``nova-libvirt-script-check`` tag to bypass the check.

releasenotes/notes/nova-libvirt-virtlogd-0b77f24764dcecfc.yaml

@@ -0,0 +1,13 @@
+---
+features:
+  - |
+    The ``nova_libvirt`` container now runs ``virtlogd`` as a sidecar
+    alongside ``libvirtd``, and ``qemu.conf`` is configured with
+    ``stdio_handler = "logd"`` so that libvirt routes instance console
+    chardevs through ``virtlogd``. This allows guest console logs to be
+    rotated according to ``/etc/libvirt/virtlogd.conf`` (by default 2MB
+    per file with 3 backups), instead of growing without bound as they
+    did when ``stdio_handler`` was set to ``"file"``.
+
+    Note that existing instances must be hard rebooted to pick up the
+    new console handler, as the chardev mode is fixed at qemu launch.

tests/test_kolla_toolbox.py

@@ -649,6 +649,66 @@ def test_fail_json_when_ansible_runner_not_found(self):
                                   self.fake_ktbw.main)
         self.assertIn('ansible-runner', error.result['msg'])
 
+    def test_runner_exec_uses_ansible_user_when_no_user_param(self):
+        """user=None must exec ansible-runner as 'ansible', not as root.
+
+        Podman serialises user=None as null in ExecCreate; some daemon
+        versions interpret null as UID 0 (root), which breaks openstacksdk
+        lookups because HOME becomes /root instead of /var/lib/ansible.
+        """
+        self.mock_ansible_module.params['user'] = None
+
+        with mock.patch.object(self.fake_ktbw, '_push_private_data_dir'), \
+             mock.patch.object(self.fake_ktbw, '_parse_runner_result',
+                               return_value={'changed': False}):
+            self.fake_ktbw.main()
+
+        # call[0]=check_runner, call[1]=ansible-runner, call[2]=cleanup
+        runner_call = self.mock_container.exec_run.call_args_list[1]
+        self.assertEqual('ansible', runner_call.kwargs.get('user'))
+
+    def test_runner_exec_sets_home_to_pdd_basedir_for_ansible_user(self):
+        """HOME must be set to _PDD_BASEDIR to finds clouds.yaml."""
+        self.mock_ansible_module.params['user'] = None
+
+        with mock.patch.object(self.fake_ktbw, '_push_private_data_dir'), \
+             mock.patch.object(self.fake_ktbw, '_parse_runner_result',
+                               return_value={'changed': False}):
+            self.fake_ktbw.main()
+
+        runner_call = self.mock_container.exec_run.call_args_list[1]
+        env = runner_call.kwargs.get('environment', {})
+        self.assertEqual(kolla_toolbox._PDD_BASEDIR, env.get('HOME'))
+
+    def test_runner_exec_uses_explicit_user_param(self):
+        """When user='rabbitmq', ansible-runner must exec as 'rabbitmq'."""
+        self.mock_ansible_module.params['user'] = 'rabbitmq'
+
+        with mock.patch.object(self.fake_ktbw, '_push_private_data_dir'), \
+             mock.patch.object(self.fake_ktbw, '_parse_runner_result',
+                               return_value={'changed': False}):
+            self.fake_ktbw.main()
+
+        runner_call = self.mock_container.exec_run.call_args_list[1]
+        self.assertEqual('rabbitmq', runner_call.kwargs.get('user'))
+
+    def test_runner_exec_no_home_override_for_non_ansible_user(self):
+        """HOME must not be overridden for non-ansible users.
+
+        Forcing HOME=/var/lib/ansible when running as rabbitmq causes
+        PermissionError because rabbitmq cannot write to ansible's home.
+        """
+        self.mock_ansible_module.params['user'] = 'rabbitmq'
+
+        with mock.patch.object(self.fake_ktbw, '_push_private_data_dir'), \
+             mock.patch.object(self.fake_ktbw, '_parse_runner_result',
+                               return_value={'changed': False}):
+            self.fake_ktbw.main()
+
+        runner_call = self.mock_container.exec_run.call_args_list[1]
+        env = runner_call.kwargs.get('environment', {})
+        self.assertNotIn('HOME', env)
+
 
 class TestModuleInteraction(TestKollaToolboxModule):
     """Class focused on testing user input data from playbook."""