Fixes undercloud install failure with setting hostname

There is a new bug when deploying master/rocky where the OS of the
undercloud/overcloud is now upgraded to CentOS 7.6. When the undercloud
install runs it fails to configure the hostname using hostnamectl. This
is because systemd-hostnamed is not running and fails to start. Simply
reloading dbus seems to fix the issue. In the dbus logs there are odd
error messages like:

dbus-daemon[3230]: Unknown username "root" in message bus configuration
file

Disabling selinux seems to fix this. This patch also moves to use
podman instead of docker for container management and invokes a script
in Ansible which updates NIC templates as new variables are added
upstream. Furthermore, with the new patches for routed networks in OOO,
it is now required that the MTU is set in network-data, as well as
adding the External network to the Compute role. Now the External
network is removed by default from the Compute role.

Change-Id: Ie8b86f6f28d69bda11b1f7a430df882970ac3cb9
Signed-off-by: Tim Rozet <trozet@redhat.com>

11 files changed, 79 insertions, 18 deletions

diff --git a/apex/common/utils.py b/apex/common/utils.py
index aae821ef..72a66d10 100644
--- a/apex/common/utils.py
+++ b/apex/common/utils.py
@@ -310,3 +310,15 @@ def fetch_properties(url):
         logging.warning('Unable to fetch properties for: {}'.format(url))
         raise exc.FetchException('Unable determine properties location: '
                                  '{}'.format(url))
+
+
+def find_container_client(os_version):
+    """
+    Determines whether to use docker or podman client
+    :param os_version: openstack version
+    :return: client name as string
+    """
+    if os_version == 'rocky' or os_version == 'queens':
+        return 'docker'
+    else:
+        return 'podman'
diff --git a/apex/deploy.py b/apex/deploy.py
index f5d64820..d0c2b208 100644
--- a/apex/deploy.py
+++ b/apex/deploy.py
@@ -527,6 +527,8 @@ def main():
             container_vars['os_version'] = os_version
             container_vars['aarch64'] = platform.machine() == 'aarch64'
             container_vars['sdn_env_file'] = sdn_env_files
+            container_vars['container_client'] = utils.find_container_client(
+                os_version)
             try:
                 utils.run_ansible(container_vars, docker_playbook,
                                   host=undercloud.ip, user='stack',
@@ -569,6 +571,8 @@ def main():
         deploy_vars['http_proxy'] = net_settings.get('http_proxy', '')
         deploy_vars['https_proxy'] = net_settings.get('https_proxy', '')
         deploy_vars['vim'] = ds_opts['vim']
+        deploy_vars['container_client'] = utils.find_container_client(
+            os_version)
         for dns_server in net_settings['dns_servers']:
             deploy_vars['dns_server_args'] += " --dns-nameserver {}".format(
                 dns_server)
@@ -733,6 +737,9 @@ def main():
         deploy_vars['sriov'] = ds_opts.get('sriov')
         deploy_vars['tacker'] = ds_opts.get('tacker')
         deploy_vars['all_in_one'] = all_in_one
+        # TODO(trozet): need to set container client to docker until OOO
+        # migrates OC to podman. Remove this later.
+        deploy_vars['container_client'] = 'docker'
         # TODO(trozet): pull all logs and store in tmp dir in overcloud
         # playbook
         post_overcloud = os.path.join(args.lib_dir, constants.ANSIBLE_PATH,
diff --git a/apex/network/network_data.py b/apex/network/network_data.py
index 1177af09..6f330c50 100644
--- a/apex/network/network_data.py
+++ b/apex/network/network_data.py
@@ -83,7 +83,7 @@ def create_network_data(ns, target=None):
                           "{}".format(net))
             raise NetworkDataException("cidr is null for network {}".format(
                 net))
-
+        tmp_net['mtu'] = network.get('mtu', 1500)
         network_data.append(copy.deepcopy(tmp_net))
 
     # have to do this due to the aforementioned bug
diff --git a/apex/tests/test_apex_common_utils.py b/apex/tests/test_apex_common_utils.py
index f307990d..1ecb7df6 100644
--- a/apex/tests/test_apex_common_utils.py
+++ b/apex/tests/test_apex_common_utils.py
@@ -155,3 +155,8 @@ class TestCommonUtils:
     def test_unique(self):
         dummy_list = [1, 2, 1, 3, 4, 5, 5]
         assert_equal(utils.unique(dummy_list), [1, 2, 3, 4, 5])
+
+    def test_find_container_client(self):
+        for version in 'rocky', 'queens':
+            assert_equal(utils.find_container_client(version), 'docker')
+        assert_equal(utils.find_container_client('master'), 'podman')
diff --git a/apex/undercloud/undercloud.py b/apex/undercloud/undercloud.py
index ccdcd168..5ee487c2 100644
--- a/apex/undercloud/undercloud.py
+++ b/apex/undercloud/undercloud.py
@@ -155,6 +155,8 @@ class Undercloud:
         ansible_vars['apex_temp_dir'] = apex_temp_dir
 
         ansible_vars['nat'] = self.detect_nat(net_settings)
+        ansible_vars['container_client'] = utils.find_container_client(
+            self.os_version)
         try:
             utils.run_ansible(ansible_vars, playbook, host=self.ip,
                               user='stack')
@@ -252,7 +254,8 @@ class Undercloud:
             "generate_service_certificate false",
             "undercloud_ntp_servers {}".format(str(ns['ntp'][0])),
             "container_images_file "
-            "/home/stack/containers-prepare-parameter.yaml"
+            "/home/stack/containers-prepare-parameter.yaml",
+            "undercloud_enable_selinux false"
         ]
 
         config['undercloud_network_config'] = [
diff --git a/build/network-environment.yaml b/build/network-environment.yaml
index 3fd22e3d..1397a0c8 100644
--- a/build/network-environment.yaml
+++ b/build/network-environment.yaml
@@ -63,33 +63,53 @@ parameter_defaults:
   NeutronExternalNetworkBridge: 'br-ex'
 
   ServiceNetMap:
+    ApacheNetwork: internal_api
     NeutronTenantNetwork: tenant
     CeilometerApiNetwork: internal_api
     AodhApiNetwork: internal_api
+    PankoApiNetwork: internal_api
+    BarbicanApiNetwork: internal_api
+    GnocchiApiNetwork: internal_api
     OpendaylightApiNetwork: internal_api
     MongoDbNetwork: internal_api
     CinderApiNetwork: internal_api
     CinderIscsiNetwork: storage
     GlanceApiNetwork: internal_api
     GlanceRegistryNetwork: internal_api
+    IronicApiNetwork: ctlplane
+    IronicNetwork: ctlplane
+    IronicInspectorNetwork: ctlplane
     KeystoneAdminApiNetwork: ctlplane
     KeystonePublicApiNetwork: internal_api
     NeutronApiNetwork: internal_api
     HeatApiNetwork: internal_api
+    HeatApiCfnNetwork: internal_api
+    HeatApiCloudwatchNetwork: internal_api
+    ManilaApiNetwork: internal_api
+    MetricsQdrNetwork: internal_api
     NovaApiNetwork: internal_api
     NovaMetadataNetwork: internal_api
+    NovaPlacementNetwork: internal_api
     NovaVncProxyNetwork: internal_api
+    NovaLibvirtNetwork: internal_api
+    NovajoinNetwork: internal_api
+    OctaviaApiNetwork: internal_api
     SwiftMgmtNetwork: storage
     SwiftProxyNetwork: storage
     TackerApiNetwork: internal_api
     CongressApiNetwork: internal_api
     HorizonNetwork: internal_api
+    OsloMessagingRpcNetwork: internal_api
+    OsloMessagingNotifyNetwork: internal_api
     MemcachedNetwork: internal_api
     RabbitMqNetwork: internal_api
     RedisNetwork: internal_api
     MysqlNetwork: internal_api
     CephClusterNetwork: storage
-    CephPublicNetwork: storage
+    CephMonNetwork: storage
+    PublicNetwork: external
+    OvnDbsNetwork: internal_api
+    DockerRegistryNetwork: ctlplane
     # Define which network will be used for hostname resolution
     ControllerHostnameResolveNetwork: internal_api
     ComputeHostnameResolveNetwork: internal_api
diff --git a/lib/ansible/playbooks/configure_undercloud.yml b/lib/ansible/playbooks/configure_undercloud.yml
index 80f3e67e..493cd1a1 100644
--- a/lib/ansible/playbooks/configure_undercloud.yml
+++ b/lib/ansible/playbooks/configure_undercloud.yml
@@ -78,7 +78,7 @@
       become: yes
       with_items: "{{ nova_config }}"
     - name: restart nova services
-      shell: "docker restart {{ item }}"
+      shell: "{{ container_client }} restart {{ item }}"
       with_items:
         - nova_conductor
         - nova_compute
@@ -90,7 +90,7 @@
       become: yes
       with_items: "{{ neutron_config }}"
     - name: restart neutron services
-      shell: "docker restart {{ item }}"
+      shell: "{{ container_client }} restart {{ item }}"
       with_items:
         - neutron_api
         - neutron_dhcp
@@ -100,7 +100,7 @@
       become: yes
       with_items: "{{ ironic_config }}"
     - name: restart ironic services
-      shell: "docker restart {{ item }}"
+      shell: "{{ container_client }} restart {{ item }}"
       with_items:
         - ironic_api
         - ironic_conductor
diff --git a/lib/ansible/playbooks/deploy_overcloud.yml b/lib/ansible/playbooks/deploy_overcloud.yml
index e2e84d18..44470da2 100644
--- a/lib/ansible/playbooks/deploy_overcloud.yml
+++ b/lib/ansible/playbooks/deploy_overcloud.yml
@@ -73,6 +73,16 @@
         owner: root
         group: root
       become: yes
+    - name: Insert External network into Compute role
+      shell: |
+        ruby -e '
+        require "yaml"
+        data = YAML.load(File.read("/usr/share/openstack-tripleo-heat-templates/roles_data.yaml"))
+        data[1]["networks"].push("External")
+        data[1]["default_route_networks"] = Array.new(["External"])
+        File.open("/usr/share/openstack-tripleo-heat-templates/roles_data.yaml", "w") { |f| f.write(data.to_yaml) }
+        '
+      become: yes
     - name: Upload glance images
       shell: "{{ stackrc }} && openstack overcloud image upload"
       become: yes
@@ -92,11 +102,6 @@
         - baremetal
         - control
         - compute
-    - name: Downgrade ceph
-      yum:
-        allow_downgrade: yes
-        name: ceph-ansible-3.1.6
-      become: yes
     - name: Re-enable ceph config for aarch64
       replace:
         path: "/usr/share/ceph-ansible/roles/ceph-client/tasks/create_users_keys.yml"
@@ -106,6 +111,15 @@
       when: aarch64
     - name: Configure DNS server for ctlplane network
       shell: "{{ stackrc }} && openstack subnet set ctlplane-subnet {{ dns_server_args }}"
+    - name: Update NIC templates before deployment
+      shell: >
+        /usr/share/openstack-tripleo-heat-templates/tools/merge-new-params-nic-config-script.py
+        -n /home/stack/network_data.yaml -t /home/stack/nics/{{ item }}.yaml --discard-comments True
+      become: yes
+      become_user: stack
+      with_items:
+        - controller
+        - compute
     - block:
         - name: Execute Overcloud Deployment
           shell: "{{ stackrc }} && bash deploy_command"
diff --git a/lib/ansible/playbooks/patch_containers.yml b/lib/ansible/playbooks/patch_containers.yml
index bc4899ba..1ef05810 100644
--- a/lib/ansible/playbooks/patch_containers.yml
+++ b/lib/ansible/playbooks/patch_containers.yml
@@ -1,13 +1,13 @@
 ---
   - name: "Pull docker image to ensure it exists locally: {{ item }}"
-    shell: docker pull {{ undercloud_ip }}:8787/tripleo{{ os_version }}/centos-binary-{{ item }}:current-tripleo
+    shell: "{{ container_client }} pull {{ undercloud_ip }}:8787/tripleo{{ os_version }}/centos-binary-{{ item }}:current-tripleo"
   - name: "Find docker image user {{ item }}"
     shell: >
-      docker inspect --format='{{ '{{' }}.ContainerConfig.User{{ '}}' }}'
+      {{ container_client }} inspect --format='{{ '{{' }}.ContainerConfig.User{{ '}}' }}'
       {{ undercloud_ip }}:8787/tripleo{{ os_version }}/centos-binary-{{ item }}:current-tripleo
     register: user_result
   - name: "Patch docker image {{ item }}"
     shell: >
-      cd /home/stack/containers/{{ item }} && docker build
+      cd /home/stack/containers/{{ item }} && {{ container_client }} build
       --build-arg REAL_USER={{ user_result.stdout }}
       -t {{ undercloud_ip }}:8787/tripleo{{ os_version }}/centos-binary-{{ item }}:apex .
diff --git a/lib/ansible/playbooks/post_deploy_overcloud.yml b/lib/ansible/playbooks/post_deploy_overcloud.yml
index f3dbfbd2..2b90ab1f 100644
--- a/lib/ansible/playbooks/post_deploy_overcloud.yml
+++ b/lib/ansible/playbooks/post_deploy_overcloud.yml
@@ -55,11 +55,11 @@
         - openstack-nova-scheduler
         - openstack-nova-conductor
     - name: Restart Compute Nova Compute (workaround for NFS)
-      shell: "docker restart nova_compute"
+      shell: "{{ container_client }} restart nova_compute"
       become: yes
       when: "'compute' in ansible_hostname or all_in_one"
     - name: Update ODL container restart policy to always
-      shell: "docker update --restart=always opendaylight_api"
+      shell: "{{ container_client }} update --restart=always opendaylight_api"
       become: yes
       when:
         - sdn == 'opendaylight'
@@ -88,7 +88,7 @@
         - "'controller' in ansible_hostname"
         - sdn != 'ovn'
     - name: Restart metadata service
-      shell: "docker restart neutron_metadata_agent"
+      shell: "{{ container_client }} restart neutron_metadata_agent"
       become: yes
       when:
         - "'controller' in ansible_hostname"
diff --git a/lib/ansible/playbooks/prepare_overcloud_containers.yml b/lib/ansible/playbooks/prepare_overcloud_containers.yml
index db1bff89..ebf081dc 100644
--- a/lib/ansible/playbooks/prepare_overcloud_containers.yml
+++ b/lib/ansible/playbooks/prepare_overcloud_containers.yml
@@ -36,7 +36,7 @@
         - patched_docker_services|length > 0
         - item in (response.json)['repositories']|join(" ")
     - name: Push patched docker images to local registry
-      shell: docker push {{ undercloud_ip }}:8787/tripleo{{ os_version }}/centos-binary-{{ item }}:apex
+      shell: "{{ container_client }} push {{ undercloud_ip }}:8787/tripleo{{ os_version }}/centos-binary-{{ item }}:apex"
       when:
         - patched_docker_services|length > 0
         - item in (response.json)['repositories']|join(" ")