diff options
Diffstat (limited to 'playbooks/common')
68 files changed, 1164 insertions, 376 deletions
diff --git a/playbooks/common/README.md b/playbooks/common/README.md index 0b5e26989..968bd99cb 100644 --- a/playbooks/common/README.md +++ b/playbooks/common/README.md @@ -1,9 +1,8 @@ # Common playbooks This directory has a generic set of playbooks that are included by playbooks in -[`byo`](../byo), as well as other playbooks related to the -[`bin/cluster`](../../bin) tool. +[`byo`](../byo). Note: playbooks in this directory use generic group names that do not line up -with the groups used by the `byo` playbooks or `bin/cluster` derived playbooks, -requiring an explicit remapping of groups. +with the groups used by the `byo` playbooks, requiring an explicit remapping of +groups. diff --git a/playbooks/common/openshift-checks/adhoc.yml b/playbooks/common/openshift-checks/adhoc.yml new file mode 100644 index 000000000..dfcef8435 --- /dev/null +++ b/playbooks/common/openshift-checks/adhoc.yml @@ -0,0 +1,12 @@ +--- +- name: OpenShift health checks + hosts: oo_all_hosts + roles: + - openshift_health_checker + vars: + - r_openshift_health_checker_playbook_context: adhoc + post_tasks: + - name: Run health checks + action: openshift_health_check + args: + checks: '{{ openshift_checks | default([]) }}' diff --git a/playbooks/common/openshift-checks/health.yml b/playbooks/common/openshift-checks/health.yml index 1bee460e8..21ea785ef 100644 --- a/playbooks/common/openshift-checks/health.yml +++ b/playbooks/common/openshift-checks/health.yml @@ -1,11 +1,11 @@ --- - name: Run OpenShift health checks - hosts: OSEv3 + hosts: oo_all_hosts roles: - openshift_health_checker vars: - - r_openshift_health_checker_playbook_context: "health" + - r_openshift_health_checker_playbook_context: health post_tasks: - - action: openshift_health_check # https://github.com/ansible/ansible/issues/20513 + - action: openshift_health_check args: checks: ['@health'] diff --git a/playbooks/common/openshift-checks/pre-install.yml b/playbooks/common/openshift-checks/pre-install.yml index e01c6f38d..88e6f9120 100644 --- a/playbooks/common/openshift-checks/pre-install.yml +++ b/playbooks/common/openshift-checks/pre-install.yml @@ -1,11 +1,11 @@ --- -- hosts: OSEv3 - name: run OpenShift pre-install checks +- name: run OpenShift pre-install checks + hosts: oo_all_hosts roles: - openshift_health_checker vars: - - r_openshift_health_checker_playbook_context: "pre-install" + - r_openshift_health_checker_playbook_context: pre-install post_tasks: - - action: openshift_health_check # https://github.com/ansible/ansible/issues/20513 + - action: openshift_health_check args: checks: ['@preflight'] diff --git a/playbooks/common/openshift-cluster/config.yml b/playbooks/common/openshift-cluster/config.yml index 1482b3a3f..26b27ba39 100644 --- a/playbooks/common/openshift-cluster/config.yml +++ b/playbooks/common/openshift-cluster/config.yml @@ -1,4 +1,29 @@ --- +# TODO: refactor this into its own include +# and pass a variable for ctx +- name: Verify Requirements + hosts: oo_all_hosts + roles: + - openshift_health_checker + vars: + - r_openshift_health_checker_playbook_context: install + post_tasks: + - action: openshift_health_check + args: + checks: + - disk_availability + - memory_availability + - package_availability + - package_version + - docker_image_availability + - docker_storage + +- hosts: localhost + tasks: + - fail: + msg: No etcd hosts defined. Running an all-in-one master is deprecated and will no longer be supported in a future upgrade. + when: groups.oo_etcd_to_config | default([]) | length == 0 and not openshift_master_unsupported_all_in_one | default(False) + - include: initialize_oo_option_facts.yml tags: - always @@ -45,6 +70,12 @@ tags: - hosted +- include: service_catalog.yml + when: + - openshift_enable_service_catalog | default(false) | bool + tags: + - servicecatalog + - name: Re-enable excluder if it was previously enabled hosts: oo_masters_to_config:oo_nodes_to_config tags: diff --git a/playbooks/common/openshift-cluster/enable_dnsmasq.yml b/playbooks/common/openshift-cluster/enable_dnsmasq.yml index 5425f448f..50351588f 100644 --- a/playbooks/common/openshift-cluster/enable_dnsmasq.yml +++ b/playbooks/common/openshift-cluster/enable_dnsmasq.yml @@ -37,7 +37,7 @@ dest: "{{ openshift.common.config_base }}/master/master-config.yaml" yaml_key: dnsConfig.bindAddress yaml_value: "{{ openshift.master.bind_addr }}:{{ openshift.master.dns_port }}" - notify: restart master + notify: restart master api - meta: flush_handlers - name: Configure nodes for dnsmasq diff --git a/playbooks/common/openshift-cluster/evaluate_groups.yml b/playbooks/common/openshift-cluster/evaluate_groups.yml index c28ce4c14..c56b07037 100644 --- a/playbooks/common/openshift-cluster/evaluate_groups.yml +++ b/playbooks/common/openshift-cluster/evaluate_groups.yml @@ -5,10 +5,10 @@ become: no gather_facts: no tasks: - - name: Evaluate groups - g_etcd_hosts required + - name: Evaluate groups - g_etcd_hosts or g_new_etcd_hosts required fail: - msg: This playbook requires g_etcd_hosts to be set - when: g_etcd_hosts is not defined + msg: This playbook requires g_etcd_hosts or g_new_etcd_hosts to be set + when: g_etcd_hosts is not defined and g_new_etcd_hosts is not defined - name: Evaluate groups - g_master_hosts or g_new_master_hosts required fail: @@ -67,6 +67,15 @@ when: g_master_hosts|length > 0 changed_when: no + - name: Evaluate oo_new_etcd_to_config + add_host: + name: "{{ item }}" + groups: oo_new_etcd_to_config + ansible_ssh_user: "{{ g_ssh_user | default(omit) }}" + ansible_become: "{{ g_sudo | default(omit) }}" + with_items: "{{ g_new_etcd_hosts | default([]) }}" + changed_when: no + - name: Evaluate oo_masters_to_config add_host: name: "{{ item }}" @@ -157,3 +166,12 @@ ansible_become: "{{ g_sudo | default(omit) }}" with_items: "{{ g_glusterfs_hosts | union(g_glusterfs_registry_hosts | default([])) }}" changed_when: no + + - name: Evaluate oo_etcd_to_migrate + add_host: + name: "{{ item }}" + groups: oo_etcd_to_migrate + ansible_ssh_user: "{{ g_ssh_user | default(omit) }}" + ansible_become: "{{ g_sudo | default(omit) }}" + with_items: "{{ groups.oo_etcd_to_config if groups.oo_etcd_to_config | default([]) | length != 0 else groups.oo_first_master }}" + changed_when: no diff --git a/playbooks/common/openshift-cluster/initialize_facts.yml b/playbooks/common/openshift-cluster/initialize_facts.yml index 9cebecd68..65be436c6 100644 --- a/playbooks/common/openshift-cluster/initialize_facts.yml +++ b/playbooks/common/openshift-cluster/initialize_facts.yml @@ -6,12 +6,155 @@ - name: Initialize host facts hosts: oo_all_hosts - roles: - - openshift_facts tasks: - - openshift_facts: + - name: load openshift_facts module + include_role: + name: openshift_facts + + # TODO: Should this role be refactored into health_checks?? + - name: Run openshift_sanitize_inventory to set variables + include_role: + name: openshift_sanitize_inventory + + - name: Detecting Operating System from ostree_booted + stat: + path: /run/ostree-booted + register: ostree_booted + + # Locally setup containerized facts for now + - name: initialize_facts set fact l_is_atomic + set_fact: + l_is_atomic: "{{ ostree_booted.stat.exists }}" + + - name: initialize_facts set fact for containerized and l_is_*_system_container + set_fact: + l_is_containerized: "{{ (l_is_atomic | bool) or (containerized | default(false) | bool) }}" + l_is_openvswitch_system_container: "{{ (openshift_use_openvswitch_system_container | default(openshift_use_system_containers | default(false)) | bool) }}" + l_is_node_system_container: "{{ (openshift_use_node_system_container | default(openshift_use_system_containers | default(false)) | bool) }}" + l_is_master_system_container: "{{ (openshift_use_master_system_container | default(openshift_use_system_containers | default(false)) | bool) }}" + l_is_etcd_system_container: "{{ (openshift_use_etcd_system_container | default(openshift_use_system_containers | default(false)) | bool) }}" + + - name: initialize_facts set facts for l_any_system_container + set_fact: + l_any_system_container: "{{ l_is_etcd_system_container or l_is_openvswitch_system_container or l_is_node_system_container or l_is_master_system_container }}" + + - name: initialize_facts set fact for l_etcd_runtime + set_fact: + l_etcd_runtime: "{{ 'runc' if l_is_etcd_system_container else 'docker' if l_is_containerized else 'host' }}" + + # TODO: Should this be moved into health checks?? + # Seems as though any check that happens with a corresponding fail should move into health_checks + - name: Validate python version - ans_dist is fedora and python is v3 + fail: + msg: | + openshift-ansible requires Python 3 for {{ ansible_distribution }}; + For information on enabling Python 3 with Ansible, see https://docs.ansible.com/ansible/python_3_support.html + when: + - ansible_distribution == 'Fedora' + - ansible_python['version']['major'] != 3 + + # TODO: Should this be moved into health checks?? + # Seems as though any check that happens with a corresponding fail should move into health_checks + - name: Validate python version - ans_dist not Fedora and python must be v2 + fail: + msg: "openshift-ansible requires Python 2 for {{ ansible_distribution }}" + when: + - ansible_distribution != 'Fedora' + - ansible_python['version']['major'] != 2 + + # TODO: Should this be moved into health checks?? + # Seems as though any check that happens with a corresponding fail should move into health_checks + # Fail as early as possible if Atomic and old version of Docker + - when: + - l_is_atomic | bool + block: + + # See https://access.redhat.com/articles/2317361 + # and https://github.com/ansible/ansible/issues/15892 + # NOTE: the "'s can not be removed at this level else the docker command will fail + # NOTE: When ansible >2.2.1.x is used this can be updated per + # https://github.com/openshift/openshift-ansible/pull/3475#discussion_r103525121 + - name: Determine Atomic Host Docker Version + shell: 'CURLY="{"; docker version --format "$CURLY{json .Server.Version}}"' + register: l_atomic_docker_version + + - name: assert atomic host docker version is 1.12 or later + assert: + that: + - l_atomic_docker_version.stdout | replace('"', '') | version_compare('1.12','>=') + msg: Installation on Atomic Host requires Docker 1.12 or later. Please upgrade and restart the Atomic Host. + + - when: + - not l_is_atomic | bool + block: + - name: Ensure openshift-ansible installer package deps are installed + package: + name: "{{ item }}" + state: present + with_items: + - iproute + - "{{ 'python3-dbus' if ansible_distribution == 'Fedora' else 'python-dbus' }}" + - PyYAML + - yum-utils + + - name: Ensure various deps for running system containers are installed + package: + name: "{{ item }}" + state: present + with_items: + - atomic + - ostree + - runc + when: + - l_any_system_container | bool + + - name: Default system_images_registry to a enterprise registry + set_fact: + system_images_registry: "registry.access.redhat.com" + when: + - system_images_registry is not defined + - openshift_deployment_type == "openshift-enterprise" + + - name: Default system_images_registry to community registry + set_fact: + system_images_registry: "docker.io" + when: + - system_images_registry is not defined + - openshift_deployment_type == "origin" + + - name: Gather Cluster facts and set is_containerized if needed + openshift_facts: role: common local_facts: + debug_level: "{{ openshift_debug_level | default(2) }}" + deployment_type: "{{ openshift_deployment_type }}" + deployment_subtype: "{{ openshift_deployment_subtype | default(None) }}" + cli_image: "{{ osm_image | default(None) }}" + cluster_id: "{{ openshift_cluster_id | default('default') }}" hostname: "{{ openshift_hostname | default(None) }}" - - set_fact: + ip: "{{ openshift_ip | default(None) }}" + is_containerized: "{{ l_is_containerized | default(None) }}" + is_openvswitch_system_container: "{{ l_is_openvswitch_system_container | default(false) }}" + is_node_system_container: "{{ l_is_node_system_container | default(false) }}" + is_master_system_container: "{{ l_is_master_system_container | default(false) }}" + is_etcd_system_container: "{{ l_is_etcd_system_container | default(false) }}" + etcd_runtime: "{{ l_etcd_runtime }}" + system_images_registry: "{{ system_images_registry }}" + public_hostname: "{{ openshift_public_hostname | default(None) }}" + public_ip: "{{ openshift_public_ip | default(None) }}" + portal_net: "{{ openshift_portal_net | default(openshift_master_portal_net) | default(None) }}" + http_proxy: "{{ openshift_http_proxy | default(None) }}" + https_proxy: "{{ openshift_https_proxy | default(None) }}" + no_proxy: "{{ openshift_no_proxy | default(None) }}" + generate_no_proxy_hosts: "{{ openshift_generate_no_proxy_hosts | default(True) }}" + no_proxy_internal_hostnames: "{{ openshift_no_proxy_internal_hostnames | default(None) }}" + sdn_network_plugin_name: "{{ os_sdn_network_plugin_name | default(None) }}" + use_openshift_sdn: "{{ openshift_use_openshift_sdn | default(None) }}" + + - name: initialize_facts set_fact repoquery command + set_fact: + repoquery_cmd: "{{ 'dnf repoquery --latest-limit 1 -d 0' if ansible_pkg_mgr == 'dnf' else 'repoquery --plugins' }}" + + - name: initialize_facts set_fact on openshift_docker_hosted_registry_network + set_fact: openshift_docker_hosted_registry_network: "{{ hostvars[groups.oo_first_master.0].openshift.common.portal_net }}" diff --git a/playbooks/common/openshift-cluster/initialize_openshift_repos.yml b/playbooks/common/openshift-cluster/initialize_openshift_repos.yml new file mode 100644 index 000000000..a7114fc80 --- /dev/null +++ b/playbooks/common/openshift-cluster/initialize_openshift_repos.yml @@ -0,0 +1,8 @@ +--- +- name: Setup yum repositories for all hosts + hosts: oo_all_hosts + gather_facts: no + tasks: + - name: initialize openshift repos + include_role: + name: openshift_repos diff --git a/playbooks/common/openshift-cluster/initialize_openshift_version.yml b/playbooks/common/openshift-cluster/initialize_openshift_version.yml index f4e52869e..7112a6084 100644 --- a/playbooks/common/openshift-cluster/initialize_openshift_version.yml +++ b/playbooks/common/openshift-cluster/initialize_openshift_version.yml @@ -1,24 +1,5 @@ --- # NOTE: requires openshift_facts be run -- name: Verify compatible yum/subscription-manager combination - hosts: oo_all_hosts - gather_facts: no - tasks: - # See: - # https://bugzilla.redhat.com/show_bug.cgi?id=1395047 - # https://bugzilla.redhat.com/show_bug.cgi?id=1282961 - # https://github.com/openshift/openshift-ansible/issues/1138 - # Consider the repoquery module for this work - - name: Check for bad combinations of yum and subscription-manager - command: > - {{ repoquery_cmd }} --installed --qf '%{version}' "yum" - register: yum_ver_test - changed_when: false - when: not openshift.common.is_atomic | bool - - fail: - msg: Incompatible versions of yum and subscription-manager found. You may need to update yum and yum-utils. - when: not openshift.common.is_atomic | bool and 'Plugin \"search-disabled-repos\" requires API 2.7. Supported API is 2.6.' in yum_ver_test.stdout - - name: Determine openshift_version to configure on first master hosts: oo_first_master roles: diff --git a/playbooks/common/openshift-cluster/openshift_hosted.yml b/playbooks/common/openshift-cluster/openshift_hosted.yml index 8d94b6509..99a634970 100644 --- a/playbooks/common/openshift-cluster/openshift_hosted.yml +++ b/playbooks/common/openshift-cluster/openshift_hosted.yml @@ -26,7 +26,10 @@ logging_elasticsearch_cluster_size: "{{ openshift_hosted_logging_elasticsearch_cluster_size | default(1) }}" logging_elasticsearch_ops_cluster_size: "{{ openshift_hosted_logging_elasticsearch_ops_cluster_size | default(1) }}" roles: + - role: openshift_default_storage_class + when: openshift_cloudprovider_kind is defined and (openshift_cloudprovider_kind == 'aws' or openshift_cloudprovider_kind == 'gce') - role: openshift_hosted + r_openshift_hosted_use_calico: "{{ openshift.common.use_calico | default(false) | bool }}" - role: openshift_metrics when: openshift_hosted_metrics_deploy | default(false) | bool - role: openshift_logging @@ -45,8 +48,6 @@ - role: cockpit-ui when: ( openshift.common.version_gte_3_3_or_1_3 | bool ) and ( openshift_hosted_manage_registry | default(true) | bool ) and not (openshift.docker.hosted_registry_insecure | default(false) | bool) - - role: openshift_default_storage_class - when: openshift_cloudprovider_kind is defined and (openshift_cloudprovider_kind == 'aws' or openshift_cloudprovider_kind == 'gce') - name: Update master-config for publicLoggingURL hosts: oo_masters_to_config:!oo_first_master diff --git a/playbooks/common/openshift-cluster/openshift_logging.yml b/playbooks/common/openshift-cluster/openshift_logging.yml index 57580406c..c1a5d83cd 100644 --- a/playbooks/common/openshift-cluster/openshift_logging.yml +++ b/playbooks/common/openshift-cluster/openshift_logging.yml @@ -1,6 +1,4 @@ --- -- include: evaluate_groups.yml - - name: OpenShift Aggregated Logging hosts: oo_first_master roles: diff --git a/playbooks/common/openshift-cluster/openshift_metrics.yml b/playbooks/common/openshift-cluster/openshift_metrics.yml index bcff4a1a1..1dc180c26 100644 --- a/playbooks/common/openshift-cluster/openshift_metrics.yml +++ b/playbooks/common/openshift-cluster/openshift_metrics.yml @@ -1,7 +1,14 @@ --- -- include: evaluate_groups.yml - - name: OpenShift Metrics hosts: oo_first_master roles: - openshift_metrics + +- name: OpenShift Metrics + hosts: oo_masters:!oo_first_master + serial: 1 + tasks: + - name: Setup the non-first masters configs + include_role: + name: openshift_metrics + tasks_from: update_master_config.yaml diff --git a/playbooks/common/openshift-cluster/redeploy-certificates/masters.yml b/playbooks/common/openshift-cluster/redeploy-certificates/masters.yml index c30889d64..51b196299 100644 --- a/playbooks/common/openshift-cluster/redeploy-certificates/masters.yml +++ b/playbooks/common/openshift-cluster/redeploy-certificates/masters.yml @@ -51,3 +51,13 @@ | oo_collect('openshift.common.hostname') | default(none, true) }}" openshift_certificates_redeploy: true + - role: lib_utils + post_tasks: + - yedit: + src: "{{ openshift.common.config_base }}/master/master-config.yaml" + key: servingInfo.namedCertificates + value: "{{ openshift.master.named_certificates | default([]) | oo_named_certificates_list }}" + when: + - ('named_certificates' in openshift.master) + - openshift.master.named_certificates | default([]) | length > 0 + - openshift_master_overwrite_named_certificates | default(false) | bool diff --git a/playbooks/common/openshift-cluster/redeploy-certificates/registry.yml b/playbooks/common/openshift-cluster/redeploy-certificates/registry.yml index 8c8062585..afd5463b2 100644 --- a/playbooks/common/openshift-cluster/redeploy-certificates/registry.yml +++ b/playbooks/common/openshift-cluster/redeploy-certificates/registry.yml @@ -66,6 +66,7 @@ --signer-cert={{ openshift.common.config_base }}/master/ca.crt --signer-key={{ openshift.common.config_base }}/master/ca.key --signer-serial={{ openshift.common.config_base }}/master/ca.serial.txt + --config={{ mktemp.stdout }}/admin.kubeconfig --hostnames="{{ docker_registry_service_ip.results.clusterip }},docker-registry.default.svc,docker-registry.default.svc.cluster.local,{{ docker_registry_route_hostname }}" --cert={{ openshift.common.config_base }}/master/registry.crt --key={{ openshift.common.config_base }}/master/registry.key diff --git a/playbooks/common/openshift-cluster/redeploy-certificates/router.yml b/playbooks/common/openshift-cluster/redeploy-certificates/router.yml index 9f14f2d69..748bbbf91 100644 --- a/playbooks/common/openshift-cluster/redeploy-certificates/router.yml +++ b/playbooks/common/openshift-cluster/redeploy-certificates/router.yml @@ -116,8 +116,9 @@ tls.crt="{{ mktemp.stdout }}/openshift-hosted-router-certificate.pem" tls.key="{{ mktemp.stdout }}/openshift-hosted-router-certificate.key" --type=kubernetes.io/tls + --config={{ mktemp.stdout }}/admin.kubeconfig --confirm - -o json | {{ openshift.common.client_binary }} replace -f - + -o json | {{ openshift.common.client_binary }} --config={{ mktemp.stdout }}/admin.kubeconfig replace -f - - name: Remove temporary router certificate and key files file: diff --git a/playbooks/common/openshift-cluster/service_catalog.yml b/playbooks/common/openshift-cluster/service_catalog.yml new file mode 100644 index 000000000..599350258 --- /dev/null +++ b/playbooks/common/openshift-cluster/service_catalog.yml @@ -0,0 +1,20 @@ +--- + +- name: Update Master configs + hosts: oo_masters + serial: 1 + tasks: + - block: + - include_role: + name: openshift_service_catalog + tasks_from: wire_aggregator + vars: + first_master: "{{ groups.oo_first_master[0] }}" + +- name: Service Catalog + hosts: oo_first_master + roles: + - openshift_service_catalog + - ansible_service_broker + vars: + first_master: "{{ groups.oo_first_master[0] }}" diff --git a/playbooks/common/openshift-cluster/std_include.yml b/playbooks/common/openshift-cluster/std_include.yml index 6ed31a644..6cc56889a 100644 --- a/playbooks/common/openshift-cluster/std_include.yml +++ b/playbooks/common/openshift-cluster/std_include.yml @@ -11,6 +11,10 @@ tags: - node +- include: initialize_openshift_repos.yml + tags: + - always + - include: initialize_openshift_version.yml tags: - always diff --git a/playbooks/common/openshift-cluster/upgrades/containerized_node_upgrade.yml b/playbooks/common/openshift-cluster/upgrades/containerized_node_upgrade.yml deleted file mode 100644 index 9f7961614..000000000 --- a/playbooks/common/openshift-cluster/upgrades/containerized_node_upgrade.yml +++ /dev/null @@ -1,14 +0,0 @@ ---- -# This is a hack to allow us to use systemd_units.yml, but skip the handlers which -# restart services. We will unconditionally restart all containerized services -# because we have to unconditionally restart Docker: -- set_fact: - skip_node_svc_handlers: True - -- name: Update systemd units - include: ../../../../roles/openshift_node/tasks/systemd_units.yml openshift_version={{ openshift_image_tag }} - -# This is a no-op because of skip_node_svc_handlers, but lets us trigger it before end of -# play when the node has already been marked schedulable again. (this would look strange -# in logs otherwise) -- meta: flush_handlers diff --git a/playbooks/common/openshift-cluster/upgrades/docker/docker_upgrade.yml b/playbooks/common/openshift-cluster/upgrades/docker/docker_upgrade.yml index 07db071ce..02b8a9d3c 100644 --- a/playbooks/common/openshift-cluster/upgrades/docker/docker_upgrade.yml +++ b/playbooks/common/openshift-cluster/upgrades/docker/docker_upgrade.yml @@ -52,9 +52,13 @@ - name: Drain Node for Kubelet upgrade command: > - {{ openshift.common.admin_binary }} drain {{ openshift.node.nodename }} --force --delete-local-data --ignore-daemonsets + {{ openshift.common.admin_binary }} drain {{ openshift.node.nodename }} --config={{ openshift.common.config_base }}/master/admin.kubeconfig --force --delete-local-data --ignore-daemonsets delegate_to: "{{ groups.oo_first_master.0 }}" when: l_docker_upgrade is defined and l_docker_upgrade | bool and inventory_hostname in groups.oo_nodes_to_upgrade + register: l_docker_upgrade_drain_result + until: not l_docker_upgrade_drain_result | failed + retries: 60 + delay: 60 - include: upgrade.yml when: l_docker_upgrade is defined and l_docker_upgrade | bool diff --git a/playbooks/common/openshift-cluster/upgrades/docker/restart.yml b/playbooks/common/openshift-cluster/upgrades/docker/restart.yml index 1b418920f..83f16ac0d 100644 --- a/playbooks/common/openshift-cluster/upgrades/docker/restart.yml +++ b/playbooks/common/openshift-cluster/upgrades/docker/restart.yml @@ -1,6 +1,10 @@ --- - name: Restart docker service: name=docker state=restarted + register: l_docker_restart_docker_in_upgrade_result + until: not l_docker_restart_docker_in_upgrade_result | failed + retries: 3 + delay: 30 - name: Update docker facts openshift_facts: @@ -11,7 +15,6 @@ with_items: - etcd_container - openvswitch - - "{{ openshift.common.service_type }}-master" - "{{ openshift.common.service_type }}-master-api" - "{{ openshift.common.service_type }}-master-controllers" - "{{ openshift.common.service_type }}-node" @@ -24,4 +27,5 @@ state: started delay: 10 port: "{{ openshift.master.api_port }}" + timeout: 600 when: inventory_hostname in groups.oo_masters_to_config diff --git a/playbooks/common/openshift-cluster/upgrades/docker/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/docker/upgrade.yml index 17f8fc6e9..808cc562c 100644 --- a/playbooks/common/openshift-cluster/upgrades/docker/upgrade.yml +++ b/playbooks/common/openshift-cluster/upgrades/docker/upgrade.yml @@ -4,7 +4,6 @@ - name: Stop containerized services service: name={{ item }} state=stopped with_items: - - "{{ openshift.common.service_type }}-master" - "{{ openshift.common.service_type }}-master-api" - "{{ openshift.common.service_type }}-master-controllers" - "{{ openshift.common.service_type }}-node" @@ -32,7 +31,13 @@ - debug: var=docker_image_count.stdout when: docker_upgrade_nuke_images is defined and docker_upgrade_nuke_images | bool -- service: name=docker state=stopped +- service: + name: docker + state: stopped + register: l_pb_docker_upgrade_stop_result + until: not l_pb_docker_upgrade_stop_result | failed + retries: 3 + delay: 30 - name: Upgrade Docker package: name=docker{{ '-' + docker_version }} state=present diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/etcd/upgrade.yml index 0431c1ce0..39e82498d 100644 --- a/playbooks/common/openshift-cluster/upgrades/etcd/upgrade.yml +++ b/playbooks/common/openshift-cluster/upgrades/etcd/upgrade.yml @@ -36,7 +36,7 @@ - not openshift.common.is_etcd_system_container | bool - name: Record containerized etcd version (runc) - command: runc exec etcd_container rpm -qa --qf '%{version}' etcd\* + command: runc exec etcd rpm -qa --qf '%{version}' etcd\* register: etcd_container_version_runc failed_when: false # AUDIT:changed_when: `false` because we are only inspecting diff --git a/playbooks/common/openshift-cluster/upgrades/master_docker b/playbooks/common/openshift-cluster/upgrades/master_docker deleted file mode 120000 index 6aeca2842..000000000 --- a/playbooks/common/openshift-cluster/upgrades/master_docker +++ /dev/null @@ -1 +0,0 @@ -../../../../roles/openshift_master/templates/master_docker
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/pre/verify_control_plane_running.yml b/playbooks/common/openshift-cluster/upgrades/pre/verify_control_plane_running.yml index 06eb5f936..45022cd61 100644 --- a/playbooks/common/openshift-cluster/upgrades/pre/verify_control_plane_running.yml +++ b/playbooks/common/openshift-cluster/upgrades/pre/verify_control_plane_running.yml @@ -9,23 +9,16 @@ local_facts: ha: "{{ groups.oo_masters_to_config | length > 1 }}" - - name: Ensure Master is running - service: - name: "{{ openshift.common.service_type }}-master" - state: started - enabled: yes - when: openshift.master.ha is defined and not openshift.master.ha | bool and openshift.common.is_containerized | bool - - name: Ensure HA Master is running service: name: "{{ openshift.common.service_type }}-master-api" state: started enabled: yes - when: openshift.master.ha is defined and openshift.master.ha | bool and openshift.common.is_containerized | bool + when: openshift.common.is_containerized | bool - name: Ensure HA Master is running service: name: "{{ openshift.common.service_type }}-master-controllers" state: started enabled: yes - when: openshift.master.ha is defined and openshift.master.ha | bool and openshift.common.is_containerized | bool + when: openshift.common.is_containerized | bool diff --git a/playbooks/common/openshift-cluster/upgrades/pre/verify_health_checks.yml b/playbooks/common/openshift-cluster/upgrades/pre/verify_health_checks.yml new file mode 100644 index 000000000..497709d25 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/pre/verify_health_checks.yml @@ -0,0 +1,13 @@ +--- +- name: Verify Host Requirements + hosts: oo_all_hosts + roles: + - openshift_health_checker + vars: + - r_openshift_health_checker_playbook_context: upgrade + post_tasks: + - action: openshift_health_check + args: + checks: + - disk_availability + - memory_availability diff --git a/playbooks/common/openshift-cluster/upgrades/pre/verify_nodes_running.yml b/playbooks/common/openshift-cluster/upgrades/pre/verify_nodes_running.yml deleted file mode 100644 index 354af3cde..000000000 --- a/playbooks/common/openshift-cluster/upgrades/pre/verify_nodes_running.yml +++ /dev/null @@ -1,13 +0,0 @@ ---- -- name: Verify node processes - hosts: oo_nodes_to_config - roles: - - openshift_facts - - openshift_docker_facts - tasks: - - name: Ensure Node is running - service: - name: "{{ openshift.common.service_type }}-node" - state: started - enabled: yes - when: openshift.common.is_containerized | bool diff --git a/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml index 6738ce11f..abcd21c90 100644 --- a/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml @@ -5,13 +5,19 @@ # oc adm migrate storage should be run prior to etcd v3 upgrade # See: https://github.com/openshift/origin/pull/14625#issuecomment-308467060 -- name: Pre master upgrade - Upgrade job storage +- name: Pre master upgrade - Upgrade all storage hosts: oo_first_master tasks: - - name: Upgrade job storage + - name: Upgrade all storage command: > {{ openshift.common.client_binary }} adm --config={{ openshift.common.config_base }}/master/admin.kubeconfig - migrate storage --include=jobs --confirm + migrate storage --include=* --confirm + register: l_pb_upgrade_control_plane_pre_upgrade_storage + when: openshift_upgrade_pre_storage_migration_enabled | default(true,true) | bool + failed_when: + - openshift_upgrade_pre_storage_migration_enabled | default(true,true) | bool + - l_pb_upgrade_control_plane_pre_upgrade_storage.rc != 0 + - openshift_upgrade_pre_storage_migration_fatal | default(true,true) | bool # If facts cache were for some reason deleted, this fact may not be set, and if not set # it will always default to true. This causes problems for the etcd data dir fact detection @@ -85,6 +91,9 @@ - include_vars: ../../../../roles/openshift_master/vars/main.yml + - name: Remove any legacy systemd units + include: ../../../../roles/openshift_master/tasks/clean_systemd_units.yml + - name: Update systemd units include: ../../../../roles/openshift_master/tasks/systemd_units.yml @@ -140,16 +149,21 @@ - include: "{{ openshift_master_upgrade_post_hook }}" when: openshift_master_upgrade_post_hook is defined - - set_fact: - master_update_complete: True - -- name: Post master upgrade - Upgrade job storage - hosts: oo_first_master - tasks: - - name: Upgrade job storage + - name: Post master upgrade - Upgrade clusterpolicies storage command: > {{ openshift.common.client_binary }} adm --config={{ openshift.common.config_base }}/master/admin.kubeconfig - migrate storage --include=jobs --confirm + migrate storage --include=clusterpolicies --confirm + register: l_pb_upgrade_control_plane_post_upgrade_storage + when: openshift_upgrade_post_storage_migration_enabled | default(true,true) | bool + failed_when: + - openshift_upgrade_post_storage_migration_enabled | default(true,true) | bool + - l_pb_upgrade_control_plane_post_upgrade_storage.rc != 0 + - openshift_upgrade_post_storage_migration_fatal | default(false,true) | bool + run_once: true + delegate_to: "{{ groups.oo_first_master.0 }}" + + - set_fact: + master_update_complete: True ############################################################################## # Gate on master update complete @@ -223,13 +237,25 @@ - name: Reconcile Security Context Constraints command: > - {{ openshift.common.client_binary }} adm policy reconcile-sccs --confirm --additive-only=true -o name + {{ openshift.common.client_binary }} adm policy --config={{ openshift.common.config_base }}/master/admin.kubeconfig reconcile-sccs --confirm --additive-only=true -o name register: reconcile_scc_result changed_when: - reconcile_scc_result.stdout != '' - reconcile_scc_result.rc == 0 run_once: true + - name: Migrate storage post policy reconciliation + command: > + {{ openshift.common.client_binary }} adm --config={{ openshift.common.config_base }}/master/admin.kubeconfig + migrate storage --include=* --confirm + run_once: true + register: l_pb_upgrade_control_plane_post_upgrade_storage + when: openshift_upgrade_post_storage_migration_enabled | default(true,true) | bool + failed_when: + - openshift_upgrade_post_storage_migration_enabled | default(true,true) | bool + - l_pb_upgrade_control_plane_post_upgrade_storage.rc != 0 + - openshift_upgrade_post_storage_migration_fatal | default(false,true) | bool + - set_fact: reconcile_complete: True @@ -288,15 +314,19 @@ - name: Drain Node for Kubelet upgrade command: > - {{ hostvars[groups.oo_first_master.0].openshift.common.admin_binary }} drain {{ openshift.node.nodename | lower }} --force --delete-local-data --ignore-daemonsets + {{ hostvars[groups.oo_first_master.0].openshift.common.admin_binary }} drain {{ openshift.node.nodename | lower }} --config={{ openshift.common.config_base }}/master/admin.kubeconfig --force --delete-local-data --ignore-daemonsets delegate_to: "{{ groups.oo_first_master.0 }}" + register: l_upgrade_control_plane_drain_result + until: not l_upgrade_control_plane_drain_result | failed + retries: 60 + delay: 60 roles: - lib_openshift - openshift_facts - docker - - openshift_node_upgrade - openshift_node_dnsmasq + - openshift_node_upgrade post_tasks: - name: Set node schedulability diff --git a/playbooks/common/openshift-cluster/upgrades/upgrade_nodes.yml b/playbooks/common/openshift-cluster/upgrades/upgrade_nodes.yml index 35a50cf4e..c93a5d89c 100644 --- a/playbooks/common/openshift-cluster/upgrades/upgrade_nodes.yml +++ b/playbooks/common/openshift-cluster/upgrades/upgrade_nodes.yml @@ -26,15 +26,19 @@ - name: Drain Node for Kubelet upgrade command: > - {{ hostvars[groups.oo_first_master.0].openshift.common.admin_binary }} drain {{ openshift.node.nodename | lower }} --force --delete-local-data --ignore-daemonsets + {{ hostvars[groups.oo_first_master.0].openshift.common.admin_binary }} drain {{ openshift.node.nodename | lower }} --config={{ openshift.common.config_base }}/master/admin.kubeconfig --force --delete-local-data --ignore-daemonsets delegate_to: "{{ groups.oo_first_master.0 }}" + register: l_upgrade_nodes_drain_result + until: not l_upgrade_nodes_drain_result | failed + retries: 60 + delay: 60 roles: - lib_openshift - openshift_facts - docker - - openshift_node_upgrade - openshift_node_dnsmasq + - openshift_node_upgrade - role: openshift_excluder r_openshift_excluder_action: enable r_openshift_excluder_service_type: "{{ openshift.common.service_type }}" diff --git a/playbooks/common/openshift-cluster/upgrades/upgrade_scheduler.yml b/playbooks/common/openshift-cluster/upgrades/upgrade_scheduler.yml index 83d2cec81..8558bf3e9 100644 --- a/playbooks/common/openshift-cluster/upgrades/upgrade_scheduler.yml +++ b/playbooks/common/openshift-cluster/upgrades/upgrade_scheduler.yml @@ -74,18 +74,21 @@ - block: - debug: msg: "WARNING: existing scheduler config does not match previous known defaults automated upgrade of scheduler config is disabled.\nexisting scheduler predicates: {{ openshift_master_scheduler_current_predicates }}\ncurrent scheduler default predicates are: {{ openshift_master_scheduler_default_predicates }}" - when: "{{ openshift_master_scheduler_current_predicates != openshift_master_scheduler_default_predicates and - openshift_master_scheduler_current_predicates not in older_predicates + [prev_predicates] }}" + when: + - openshift_master_scheduler_current_predicates != openshift_master_scheduler_default_predicates + - openshift_master_scheduler_current_predicates not in older_predicates + [prev_predicates] - set_fact: openshift_upgrade_scheduler_predicates: "{{ openshift_master_scheduler_default_predicates }}" - when: "{{ openshift_master_scheduler_current_predicates != openshift_master_scheduler_default_predicates and - openshift_master_scheduler_current_predicates in older_predicates + [prev_predicates] }}" + when: + - openshift_master_scheduler_current_predicates != openshift_master_scheduler_default_predicates + - openshift_master_scheduler_current_predicates in older_predicates + [prev_predicates] - set_fact: openshift_upgrade_scheduler_predicates: "{{ default_predicates_no_region }}" - when: "{{ openshift_master_scheduler_current_predicates != default_predicates_no_region and - openshift_master_scheduler_current_predicates in older_predicates_no_region + [prev_predicates_no_region] }}" + when: + - openshift_master_scheduler_current_predicates != default_predicates_no_region + - openshift_master_scheduler_current_predicates in older_predicates_no_region + [prev_predicates_no_region] when: openshift_master_scheduler_predicates | default(none) is none @@ -131,18 +134,21 @@ - block: - debug: msg: "WARNING: existing scheduler config does not match previous known defaults automated upgrade of scheduler config is disabled.\nexisting scheduler priorities: {{ openshift_master_scheduler_current_priorities }}\ncurrent scheduler default priorities are: {{ openshift_master_scheduler_default_priorities }}" - when: "{{ openshift_master_scheduler_current_priorities != openshift_master_scheduler_default_priorities and - openshift_master_scheduler_current_priorities not in older_priorities + [prev_priorities] }}" + when: + - openshift_master_scheduler_current_priorities != openshift_master_scheduler_default_priorities + - openshift_master_scheduler_current_priorities not in older_priorities + [prev_priorities] - set_fact: openshift_upgrade_scheduler_priorities: "{{ openshift_master_scheduler_default_priorities }}" - when: "{{ openshift_master_scheduler_current_priorities != openshift_master_scheduler_default_priorities and - openshift_master_scheduler_current_priorities in older_priorities + [prev_priorities] }}" + when: + - openshift_master_scheduler_current_priorities != openshift_master_scheduler_default_priorities + - openshift_master_scheduler_current_priorities in older_priorities + [prev_priorities] - set_fact: openshift_upgrade_scheduler_priorities: "{{ default_priorities_no_zone }}" - when: "{{ openshift_master_scheduler_current_priorities != default_priorities_no_zone and - openshift_master_scheduler_current_priorities in older_priorities_no_zone + [prev_priorities_no_zone] }}" + when: + - openshift_master_scheduler_current_priorities != default_priorities_no_zone + - openshift_master_scheduler_current_priorities in older_priorities_no_zone + [prev_priorities_no_zone] when: openshift_master_scheduler_priorities | default(none) is none @@ -162,5 +168,6 @@ content: "{{ scheduler_config | to_nice_json }}" dest: "{{ openshift_master_scheduler_conf }}" backup: true - when: "{{ openshift_upgrade_scheduler_predicates is defined or - openshift_upgrade_scheduler_priorities is defined }}" + when: > + openshift_upgrade_scheduler_predicates is defined or + openshift_upgrade_scheduler_priorities is defined diff --git a/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade.yml index f1245aa2e..0f6fb46a4 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade.yml @@ -39,8 +39,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_control_plane.yml index b693ab55c..cfba788a8 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_control_plane.yml @@ -47,8 +47,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_nodes.yml b/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_nodes.yml index 4fd029107..1054f430e 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_nodes.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_3/upgrade_nodes.yml @@ -40,8 +40,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade.yml index 965e39482..783289c87 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade.yml @@ -39,8 +39,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_control_plane.yml index 7830f462c..8aa443c3c 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_control_plane.yml @@ -47,8 +47,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_nodes.yml b/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_nodes.yml index 4364ff8e3..436795694 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_nodes.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_4/upgrade_nodes.yml @@ -40,8 +40,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade.yml index 4e7c14e94..9a000265e 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade.yml @@ -39,8 +39,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_control_plane.yml index 45b664d06..2dd9676c7 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_control_plane.yml @@ -47,8 +47,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_nodes.yml b/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_nodes.yml index 036d3fcf5..d5fe8285e 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_nodes.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_5/upgrade_nodes.yml @@ -40,8 +40,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade.yml index 5b9ac9e8f..8ceab09f4 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade.yml @@ -39,8 +39,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: @@ -70,6 +71,10 @@ # docker is configured and running. skip_docker_role: True +- include: ../pre/verify_health_checks.yml + tags: + - pre_upgrade + - include: ../pre/verify_control_plane_running.yml tags: - pre_upgrade diff --git a/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_control_plane.yml index a470c7595..f765e9064 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_control_plane.yml @@ -47,8 +47,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_nodes.yml b/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_nodes.yml index 25eceaf90..8bed6a8c2 100644 --- a/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_nodes.yml +++ b/playbooks/common/openshift-cluster/upgrades/v3_6/upgrade_nodes.yml @@ -40,8 +40,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool - include: ../pre/verify_inventory_vars.yml tags: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_7/filter_plugins b/playbooks/common/openshift-cluster/upgrades/v3_7/filter_plugins new file mode 120000 index 000000000..7de3c1dd7 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_7/filter_plugins @@ -0,0 +1 @@ +../../../../../filter_plugins/
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/v3_7/master_config_upgrade.yml b/playbooks/common/openshift-cluster/upgrades/v3_7/master_config_upgrade.yml new file mode 100644 index 000000000..ed89dbe8d --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_7/master_config_upgrade.yml @@ -0,0 +1,16 @@ +--- +- modify_yaml: + dest: "{{ openshift.common.config_base}}/master/master-config.yaml" + yaml_key: 'admissionConfig.pluginConfig' + yaml_value: "{{ openshift.master.admission_plugin_config }}" + when: "'admission_plugin_config' in openshift.master" + +- modify_yaml: + dest: "{{ openshift.common.config_base}}/master/master-config.yaml" + yaml_key: 'admissionConfig.pluginOrderOverride' + yaml_value: + +- modify_yaml: + dest: "{{ openshift.common.config_base}}/master/master-config.yaml" + yaml_key: 'kubernetesMasterConfig.admissionConfig' + yaml_value: diff --git a/playbooks/common/openshift-cluster/upgrades/v3_7/roles b/playbooks/common/openshift-cluster/upgrades/v3_7/roles new file mode 120000 index 000000000..415645be6 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_7/roles @@ -0,0 +1 @@ +../../../../../roles/
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade.yml b/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade.yml new file mode 100644 index 000000000..4f05d0c64 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade.yml @@ -0,0 +1,122 @@ +--- +# +# Full Control Plane + Nodes Upgrade +# +- include: ../init.yml + tags: + - pre_upgrade + +- name: Configure the upgrade target for the common upgrade tasks + hosts: oo_all_hosts + tags: + - pre_upgrade + tasks: + - set_fact: + openshift_upgrade_target: '3.7' + openshift_upgrade_min: '3.6' + +# Pre-upgrade + +- include: ../initialize_nodes_to_upgrade.yml + tags: + - pre_upgrade + +- name: Update repos and initialize facts on all hosts + hosts: oo_masters_to_config:oo_nodes_to_upgrade:oo_etcd_to_config:oo_lb_to_config + tags: + - pre_upgrade + roles: + - openshift_repos + +- name: Set openshift_no_proxy_internal_hostnames + hosts: oo_masters_to_config:oo_nodes_to_upgrade + tags: + - pre_upgrade + tasks: + - set_fact: + openshift_no_proxy_internal_hostnames: "{{ hostvars | oo_select_keys(groups['oo_nodes_to_config'] + | union(groups['oo_masters_to_config']) + | union(groups['oo_etcd_to_config'] | default([]))) + | oo_collect('openshift.common.hostname') | default([]) | join (',') + }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool + +- include: ../pre/verify_inventory_vars.yml + tags: + - pre_upgrade + +- include: ../disable_master_excluders.yml + tags: + - pre_upgrade + +- include: ../disable_node_excluders.yml + tags: + - pre_upgrade + +- include: ../../initialize_openshift_version.yml + tags: + - pre_upgrade + vars: + # Request specific openshift_release and let the openshift_version role handle converting this + # to a more specific version, respecting openshift_image_tag and openshift_pkg_version if + # defined, and overriding the normal behavior of protecting the installed version + openshift_release: "{{ openshift_upgrade_target }}" + openshift_protect_installed_version: False + + # We skip the docker role at this point in upgrade to prevent + # unintended package, container, or config upgrades which trigger + # docker restarts. At this early stage of upgrade we can assume + # docker is configured and running. + skip_docker_role: True + +- include: ../pre/verify_health_checks.yml + tags: + - pre_upgrade + +- include: ../pre/verify_control_plane_running.yml + tags: + - pre_upgrade + +- include: ../../../openshift-master/validate_restart.yml + tags: + - pre_upgrade + +- name: Verify upgrade targets + hosts: oo_masters_to_config:oo_nodes_to_upgrade + tasks: + - include: ../pre/verify_upgrade_targets.yml + tags: + - pre_upgrade + +- name: Verify docker upgrade targets + hosts: oo_masters_to_config:oo_nodes_to_upgrade:oo_etcd_to_config + tasks: + - include: ../pre/verify_docker_upgrade_targets.yml + tags: + - pre_upgrade + +- include: validator.yml + tags: + - pre_upgrade + +- include: ../pre/gate_checks.yml + tags: + - pre_upgrade + +# Pre-upgrade completed, nothing after this should be tagged pre_upgrade. + +# Separate step so we can execute in parallel and clear out anything unused +# before we get into the serialized upgrade process which will then remove +# remaining images if possible. +- name: Cleanup unused Docker images + hosts: oo_masters_to_config:oo_nodes_to_upgrade:oo_etcd_to_config + tasks: + - include: ../cleanup_unused_images.yml + +- include: ../upgrade_control_plane.yml + +- include: ../upgrade_nodes.yml + +- include: ../post_control_plane.yml diff --git a/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade_control_plane.yml new file mode 100644 index 000000000..2ef95e778 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade_control_plane.yml @@ -0,0 +1,122 @@ +--- +# +# Control Plane Upgrade Playbook +# +# Upgrades masters and Docker (only on standalone etcd hosts) +# +# This upgrade does not include: +# - node service running on masters +# - docker running on masters +# - node service running on dedicated nodes +# +# You can run the upgrade_nodes.yml playbook after this to upgrade these components separately. +# +- include: ../init.yml + tags: + - pre_upgrade + +- name: Configure the upgrade target for the common upgrade tasks + hosts: oo_all_hosts + tags: + - pre_upgrade + tasks: + - set_fact: + openshift_upgrade_target: '3.7' + openshift_upgrade_min: '3.6' + +# Pre-upgrade +- include: ../initialize_nodes_to_upgrade.yml + tags: + - pre_upgrade + +- name: Update repos on control plane hosts + hosts: oo_masters_to_config:oo_etcd_to_config:oo_lb_to_config + tags: + - pre_upgrade + roles: + - openshift_repos + +- name: Set openshift_no_proxy_internal_hostnames + hosts: oo_masters_to_config:oo_nodes_to_upgrade + tags: + - pre_upgrade + tasks: + - set_fact: + openshift_no_proxy_internal_hostnames: "{{ hostvars | oo_select_keys(groups['oo_nodes_to_config'] + | union(groups['oo_masters_to_config']) + | union(groups['oo_etcd_to_config'] | default([]))) + | oo_collect('openshift.common.hostname') | default([]) | join (',') + }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool + +- include: ../pre/verify_inventory_vars.yml + tags: + - pre_upgrade + +- include: ../disable_master_excluders.yml + tags: + - pre_upgrade + +- include: ../../initialize_openshift_version.yml + tags: + - pre_upgrade + vars: + # Request specific openshift_release and let the openshift_version role handle converting this + # to a more specific version, respecting openshift_image_tag and openshift_pkg_version if + # defined, and overriding the normal behavior of protecting the installed version + openshift_release: "{{ openshift_upgrade_target }}" + openshift_protect_installed_version: False + + # We skip the docker role at this point in upgrade to prevent + # unintended package, container, or config upgrades which trigger + # docker restarts. At this early stage of upgrade we can assume + # docker is configured and running. + skip_docker_role: True + +- include: ../pre/verify_control_plane_running.yml + tags: + - pre_upgrade + +- include: ../../../openshift-master/validate_restart.yml + tags: + - pre_upgrade + +- name: Verify upgrade targets + hosts: oo_masters_to_config + tasks: + - include: ../pre/verify_upgrade_targets.yml + tags: + - pre_upgrade + +- name: Verify docker upgrade targets + hosts: oo_masters_to_config:oo_etcd_to_config + tasks: + - include: ../pre/verify_docker_upgrade_targets.yml + tags: + - pre_upgrade + +- include: validator.yml + tags: + - pre_upgrade + +- include: ../pre/gate_checks.yml + tags: + - pre_upgrade + +# Pre-upgrade completed, nothing after this should be tagged pre_upgrade. + +# Separate step so we can execute in parallel and clear out anything unused +# before we get into the serialized upgrade process which will then remove +# remaining images if possible. +- name: Cleanup unused Docker images + hosts: oo_masters_to_config:oo_etcd_to_config + tasks: + - include: ../cleanup_unused_images.yml + +- include: ../upgrade_control_plane.yml + vars: + master_config_hook: "v3_7/master_config_upgrade.yml" + +- include: ../post_control_plane.yml diff --git a/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade_nodes.yml b/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade_nodes.yml new file mode 100644 index 000000000..abc4c245b --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_7/upgrade_nodes.yml @@ -0,0 +1,111 @@ +--- +# +# Node Upgrade Playbook +# +# Upgrades nodes only, but requires the control plane to have already been upgraded. +# +- include: ../init.yml + tags: + - pre_upgrade + +- name: Configure the upgrade target for the common upgrade tasks + hosts: oo_all_hosts + tags: + - pre_upgrade + tasks: + - set_fact: + openshift_upgrade_target: '3.7' + openshift_upgrade_min: '3.6' + +# Pre-upgrade +- include: ../initialize_nodes_to_upgrade.yml + tags: + - pre_upgrade + +- name: Update repos on nodes + hosts: oo_masters_to_config:oo_nodes_to_upgrade:oo_etcd_to_config:oo_lb_to_config + roles: + - openshift_repos + tags: + - pre_upgrade + +- name: Set openshift_no_proxy_internal_hostnames + hosts: oo_masters_to_config:oo_nodes_to_upgrade + tags: + - pre_upgrade + tasks: + - set_fact: + openshift_no_proxy_internal_hostnames: "{{ hostvars | oo_select_keys(groups['oo_nodes_to_upgrade'] + | union(groups['oo_masters_to_config']) + | union(groups['oo_etcd_to_config'] | default([]))) + | oo_collect('openshift.common.hostname') | default([]) | join (',') + }}" + when: + - openshift_http_proxy is defined or openshift_https_proxy is defined + - openshift_generate_no_proxy_hosts | default(True) | bool + +- include: ../pre/verify_inventory_vars.yml + tags: + - pre_upgrade + +- include: ../disable_node_excluders.yml + tags: + - pre_upgrade + +- include: ../../initialize_openshift_version.yml + tags: + - pre_upgrade + vars: + # Request specific openshift_release and let the openshift_version role handle converting this + # to a more specific version, respecting openshift_image_tag and openshift_pkg_version if + # defined, and overriding the normal behavior of protecting the installed version + openshift_release: "{{ openshift_upgrade_target }}" + openshift_protect_installed_version: False + + # We skip the docker role at this point in upgrade to prevent + # unintended package, container, or config upgrades which trigger + # docker restarts. At this early stage of upgrade we can assume + # docker is configured and running. + skip_docker_role: True + +- name: Verify masters are already upgraded + hosts: oo_masters_to_config + tags: + - pre_upgrade + tasks: + - fail: msg="Master running {{ openshift.common.version }} must be upgraded to {{ openshift_version }} before node upgrade can be run." + when: openshift.common.version != openshift_version + +- include: ../pre/verify_control_plane_running.yml + tags: + - pre_upgrade + +- name: Verify upgrade targets + hosts: oo_nodes_to_upgrade + tasks: + - include: ../pre/verify_upgrade_targets.yml + tags: + - pre_upgrade + +- name: Verify docker upgrade targets + hosts: oo_nodes_to_upgrade + tasks: + - include: ../pre/verify_docker_upgrade_targets.yml + tags: + - pre_upgrade + +- include: ../pre/gate_checks.yml + tags: + - pre_upgrade + +# Pre-upgrade completed, nothing after this should be tagged pre_upgrade. + +# Separate step so we can execute in parallel and clear out anything unused +# before we get into the serialized upgrade process which will then remove +# remaining images if possible. +- name: Cleanup unused Docker images + hosts: oo_nodes_to_upgrade + tasks: + - include: ../cleanup_unused_images.yml + +- include: ../upgrade_nodes.yml diff --git a/playbooks/common/openshift-cluster/upgrades/v3_7/validator.yml b/playbooks/common/openshift-cluster/upgrades/v3_7/validator.yml new file mode 100644 index 000000000..136ad5362 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/v3_7/validator.yml @@ -0,0 +1,22 @@ +--- +############################################################################### +# Pre upgrade checks for known data problems, if this playbook fails you should +# contact support. If you're not supported contact users@lists.openshift.com +############################################################################### +- name: Verify 3.7 specific upgrade checks + hosts: oo_first_master + roles: + - { role: lib_openshift } + + tasks: + - name: Check for invalid namespaces and SDN errors + oc_objectvalidator: + + - name: Confirm OpenShift authorization objects are in sync + command: > + {{ openshift.common.client_binary }} adm migrate authorization + changed_when: false + register: l_oc_result + until: l_oc_result.rc == 0 + retries: 4 + delay: 15 diff --git a/playbooks/common/openshift-cluster/validate_hostnames.yml b/playbooks/common/openshift-cluster/validate_hostnames.yml index 33fc5630f..be2e6a15a 100644 --- a/playbooks/common/openshift-cluster/validate_hostnames.yml +++ b/playbooks/common/openshift-cluster/validate_hostnames.yml @@ -1,17 +1,22 @@ --- -- name: Gather and set facts for node hosts +- name: Validate node hostnames hosts: oo_nodes_to_config - roles: - - openshift_facts tasks: - - shell: + - name: Query DNS for IP address of {{ openshift.common.hostname }} + shell: getent ahostsv4 {{ openshift.common.hostname }} | head -n 1 | awk '{ print $1 }' register: lookupip changed_when: false failed_when: false - name: Warn user about bad openshift_hostname values pause: - prompt: "The hostname \"{{ openshift.common.hostname }}\" for \"{{ ansible_nodename }}\" doesn't resolve to an ip address owned by this host. Please set openshift_hostname variable to a hostname that when resolved on the host in question resolves to an IP address matching an interface on this host. This host will fail liveness checks for pods utilizing hostPorts, press ENTER to continue or CTRL-C to abort." + prompt: + The hostname {{ openshift.common.hostname }} for {{ ansible_nodename }} + doesn't resolve to an IP address owned by this host. Please set + openshift_hostname variable to a hostname that when resolved on the host + in question resolves to an IP address matching an interface on this + host. This host will fail liveness checks for pods utilizing hostPorts, + press ENTER to continue or CTRL-C to abort. seconds: "{{ 10 if openshift_override_hostname_check | default(false) | bool else omit }}" when: - lookupip.stdout != '127.0.0.1' diff --git a/playbooks/common/openshift-etcd/config.yml b/playbooks/common/openshift-etcd/config.yml index 2cb6197d1..f2b85eea1 100644 --- a/playbooks/common/openshift-etcd/config.yml +++ b/playbooks/common/openshift-etcd/config.yml @@ -3,6 +3,7 @@ hosts: oo_etcd_to_config any_errors_fatal: true roles: + - role: os_firewall - role: openshift_etcd etcd_peers: "{{ groups.oo_etcd_to_config | default([], true) }}" etcd_ca_host: "{{ groups.oo_etcd_to_config.0 }}" diff --git a/playbooks/common/openshift-etcd/migrate.yml b/playbooks/common/openshift-etcd/migrate.yml new file mode 100644 index 000000000..311ff84b6 --- /dev/null +++ b/playbooks/common/openshift-etcd/migrate.yml @@ -0,0 +1,165 @@ +--- +- include: ../openshift-cluster/evaluate_groups.yml + tags: + - always + +- name: Run pre-checks + hosts: oo_etcd_to_migrate + tags: + - always + roles: + - role: etcd_migrate + r_etcd_migrate_action: check + r_etcd_common_embedded_etcd: "{{ groups.oo_etcd_to_config | default([]) | length == 0 }}" + etcd_peer: "{{ ansible_default_ipv4.address }}" + +- include: ../openshift-cluster/initialize_facts.yml + tags: + - always + +# TODO: This will be different for release-3.6 branch +- name: Prepare masters for etcd data migration + hosts: oo_masters_to_config + tasks: + - set_fact: + master_services: + - "{{ openshift.common.service_type + '-master-controllers' }}" + - "{{ openshift.common.service_type + '-master-api' }}" + - debug: + msg: "master service name: {{ master_services }}" + - name: Stop masters + service: + name: "{{ item }}" + state: stopped + with_items: "{{ master_services }}" + +- name: Backup v2 data + hosts: oo_etcd_to_migrate + gather_facts: no + tags: + - always + roles: + - role: openshift_facts + - role: etcd_common + r_etcd_common_action: backup + r_etcd_common_etcd_runtime: "{{ openshift.common.etcd_runtime }}" + r_etcd_common_backup_tag: pre-migration + r_etcd_common_embedded_etcd: "{{ groups.oo_etcd_to_config | default([]) | length == 0 }}" + r_etcd_common_backup_sufix_name: "{{ lookup('pipe', 'date +%Y%m%d%H%M%S') }}" + +- name: Gate on etcd backup + hosts: localhost + connection: local + become: no + tasks: + - set_fact: + etcd_backup_completed: "{{ hostvars + | oo_select_keys(groups.oo_etcd_to_migrate) + | oo_collect('inventory_hostname', {'r_etcd_common_backup_complete': true}) }}" + - set_fact: + etcd_backup_failed: "{{ groups.oo_etcd_to_migrate | difference(etcd_backup_completed) }}" + - fail: + msg: "Migration cannot continue. The following hosts did not complete etcd backup: {{ etcd_backup_failed | join(',') }}" + when: + - etcd_backup_failed | length > 0 + +- name: Stop etcd + hosts: oo_etcd_to_migrate + gather_facts: no + tags: + - always + pre_tasks: + - set_fact: + l_etcd_service: "{{ 'etcd_container' if openshift.common.is_containerized else 'etcd' }}" + - name: Disable etcd members + service: + name: "{{ l_etcd_service }}" + state: stopped + +- name: Migrate data on first etcd + hosts: oo_etcd_to_migrate[0] + gather_facts: no + tags: + - always + roles: + - role: etcd_migrate + r_etcd_migrate_action: migrate + r_etcd_common_embedded_etcd: "{{ groups.oo_etcd_to_config | default([]) | length == 0 }}" + etcd_peer: "{{ ansible_default_ipv4.address }}" + etcd_url_scheme: "https" + etcd_peer_url_scheme: "https" + +- name: Clean data stores on remaining etcd hosts + hosts: oo_etcd_to_migrate[1:] + gather_facts: no + tags: + - always + roles: + - role: etcd_migrate + r_etcd_migrate_action: clean_data + r_etcd_common_embedded_etcd: "{{ groups.oo_etcd_to_config | default([]) | length == 0 }}" + etcd_peer: "{{ ansible_default_ipv4.address }}" + etcd_url_scheme: "https" + etcd_peer_url_scheme: "https" + post_tasks: + - name: Add etcd hosts + delegate_to: localhost + add_host: + name: "{{ item }}" + groups: oo_new_etcd_to_config + ansible_ssh_user: "{{ g_ssh_user | default(omit) }}" + ansible_become: "{{ g_sudo | default(omit) }}" + with_items: "{{ groups.oo_etcd_to_migrate[1:] | default([]) }}" + changed_when: no + - name: Set success + set_fact: + r_etcd_migrate_success: true + +- include: ./scaleup.yml + +- name: Gate on etcd migration + hosts: oo_masters_to_config + gather_facts: no + tasks: + - set_fact: + etcd_migration_completed: "{{ hostvars + | oo_select_keys(groups.oo_etcd_to_migrate) + | oo_collect('inventory_hostname', {'r_etcd_migrate_success': true}) }}" + - set_fact: + etcd_migration_failed: "{{ groups.oo_etcd_to_migrate | difference(etcd_migration_completed) }}" + +- name: Add TTLs on the first master + hosts: oo_first_master[0] + roles: + - role: etcd_migrate + r_etcd_migrate_action: add_ttls + etcd_peer: "{{ hostvars[groups.oo_etcd_to_migrate.0].ansible_default_ipv4.address }}" + etcd_url_scheme: "https" + etcd_peer_url_scheme: "https" + when: etcd_migration_failed | length == 0 + +- name: Configure masters if etcd data migration is succesfull + hosts: oo_masters_to_config + roles: + - role: etcd_migrate + r_etcd_migrate_action: configure + when: etcd_migration_failed | length == 0 + tasks: + - debug: + msg: "Skipping master re-configuration since migration failed." + when: + - etcd_migration_failed | length > 0 + - name: Start master services + service: + name: "{{ item }}" + state: started + register: service_status + # Sometimes the master-api, resp. master-controllers fails to start for the first time + until: service_status.state is defined and service_status.state == "started" + retries: 5 + delay: 10 + with_items: "{{ master_services[::-1] }}" + - fail: + msg: "Migration failed. The following hosts were not properly migrated: {{ etcd_migration_failed | join(',') }}" + when: + - etcd_migration_failed | length > 0 diff --git a/playbooks/common/openshift-etcd/scaleup.yml b/playbooks/common/openshift-etcd/scaleup.yml new file mode 100644 index 000000000..5f8bb1c7a --- /dev/null +++ b/playbooks/common/openshift-etcd/scaleup.yml @@ -0,0 +1,55 @@ +--- +- name: Gather facts + hosts: oo_etcd_to_config:oo_new_etcd_to_config + roles: + - openshift_etcd_facts + post_tasks: + - set_fact: + etcd_hostname: "{{ etcd_hostname }}" + etcd_ip: "{{ etcd_ip }}" + +- name: Configure etcd + hosts: oo_new_etcd_to_config + serial: 1 + any_errors_fatal: true + vars: + etcd_ca_host: "{{ groups.oo_etcd_to_config.0 }}" + pre_tasks: + - name: Add new etcd members to cluster + command: > + /usr/bin/etcdctl --cert-file {{ etcd_peer_cert_file }} + --key-file {{ etcd_peer_key_file }} + --ca-file {{ etcd_peer_ca_file }} + -C {{ etcd_peer_url_scheme }}://{{ hostvars[etcd_ca_host].etcd_hostname }}:{{ etcd_client_port }} + member add {{ etcd_hostname }} {{ etcd_peer_url_scheme }}://{{ etcd_ip }}:{{ etcd_peer_port }} + delegate_to: "{{ etcd_ca_host }}" + register: etcd_add_check + retries: 3 + delay: 10 + until: etcd_add_check.rc == 0 + roles: + - role: os_firewall + when: etcd_add_check.rc == 0 + - role: openshift_etcd + when: etcd_add_check.rc == 0 + etcd_peers: "{{ groups.oo_etcd_to_config | union(groups.oo_new_etcd_to_config)| default([], true) }}" + etcd_ca_host: "{{ groups.oo_etcd_to_config.0 }}" + etcd_certificates_etcd_hosts: "{{ groups.oo_etcd_to_config | default([], true) }}" + etcd_initial_cluster_state: "existing" + initial_etcd_cluster: "{{ etcd_add_check.stdout_lines[3] | regex_replace('ETCD_INITIAL_CLUSTER=','') | regex_replace('\"','') }}" + etcd_ca_setup: False + r_etcd_common_etcd_runtime: "{{ openshift.common.etcd_runtime }}" + - role: nickhammond.logrotate + when: etcd_add_check.rc == 0 + post_tasks: + - name: Verify cluster is stable + command: > + /usr/bin/etcdctl --cert-file {{ etcd_peer_cert_file }} + --key-file {{ etcd_peer_key_file }} + --ca-file {{ etcd_peer_ca_file }} + -C {{ etcd_peer_url_scheme }}://{{ hostvars[etcd_ca_host].etcd_hostname }}:{{ etcd_client_port }} + cluster-health + register: scaleup_health + retries: 3 + delay: 30 + until: scaleup_health.rc == 0 diff --git a/playbooks/common/openshift-etcd/service.yml b/playbooks/common/openshift-etcd/service.yml deleted file mode 100644 index ced4bddc5..000000000 --- a/playbooks/common/openshift-etcd/service.yml +++ /dev/null @@ -1,23 +0,0 @@ ---- -- name: Populate g_service_masters host group if needed - hosts: localhost - connection: local - become: no - gather_facts: no - tasks: - - fail: msg="new_cluster_state is required to be injected in this playbook" - when: new_cluster_state is not defined - - - name: Evaluate g_service_etcd - add_host: - name: "{{ item }}" - groups: g_service_etcd - with_items: "{{ oo_host_group_exp | default([]) }}" - changed_when: False - -- name: Change etcd state on etcd instance(s) - hosts: g_service_etcd - connection: ssh - gather_facts: no - tasks: - - service: name=etcd state="{{ new_cluster_state }}" diff --git a/playbooks/common/openshift-glusterfs/config.yml b/playbooks/common/openshift-glusterfs/config.yml index 1efdfb336..d9de578f3 100644 --- a/playbooks/common/openshift-glusterfs/config.yml +++ b/playbooks/common/openshift-glusterfs/config.yml @@ -1,23 +1,26 @@ --- -- name: Open firewall ports for GlusterFS - hosts: oo_glusterfs_to_config - vars: - os_firewall_allow: - - service: glusterfs_sshd - port: "2222/tcp" - - service: glusterfs_daemon - port: "24007/tcp" - - service: glusterfs_management - port: "24008/tcp" - - service: glusterfs_bricks - port: "49152-49251/tcp" - roles: - - role: os_firewall +- name: Open firewall ports for GlusterFS nodes + hosts: glusterfs + tasks: + - include_role: + name: openshift_storage_glusterfs + tasks_from: firewall.yml when: - - openshift_storage_glusterfs_is_native | default(True) + - openshift_storage_glusterfs_is_native | default(True) | bool + +- name: Open firewall ports for GlusterFS registry nodes + hosts: glusterfs_registry + tasks: + - include_role: + name: openshift_storage_glusterfs + tasks_from: firewall.yml + when: + - openshift_storage_glusterfs_registry_is_native | default(True) | bool - name: Configure GlusterFS hosts: oo_first_master - roles: - - role: openshift_storage_glusterfs + tasks: + - name: setup glusterfs + include_role: + name: openshift_storage_glusterfs when: groups.oo_glusterfs_to_config | default([]) | count > 0 diff --git a/playbooks/common/openshift-loadbalancer/config.yml b/playbooks/common/openshift-loadbalancer/config.yml index c414913bf..09ed81a83 100644 --- a/playbooks/common/openshift-loadbalancer/config.yml +++ b/playbooks/common/openshift-loadbalancer/config.yml @@ -12,5 +12,7 @@ openshift_use_nuage | default(false), nuage_mon_rest_server_port | default(none))) + openshift_loadbalancer_additional_backends | default([]) }}" + openshift_image_tag: "{{ hostvars[groups.oo_first_master.0].openshift_image_tag }}" roles: + - role: os_firewall - role: openshift_loadbalancer diff --git a/playbooks/common/openshift-loadbalancer/service.yml b/playbooks/common/openshift-loadbalancer/service.yml deleted file mode 100644 index d3762c961..000000000 --- a/playbooks/common/openshift-loadbalancer/service.yml +++ /dev/null @@ -1,23 +0,0 @@ ---- -- name: Populate g_service_nodes host group if needed - hosts: localhost - connection: local - become: no - gather_facts: no - tasks: - - fail: msg="new_cluster_state is required to be injected in this playbook" - when: new_cluster_state is not defined - - - name: Evaluate g_service_lb - add_host: - name: "{{ item }}" - groups: g_service_lb - with_items: "{{ oo_host_group_exp | default([]) }}" - changed_when: False - -- name: Change state on lb instance(s) - hosts: g_service_lb - connection: ssh - gather_facts: no - tasks: - - service: name=haproxy state="{{ new_cluster_state }}" diff --git a/playbooks/common/openshift-master/config.yml b/playbooks/common/openshift-master/config.yml index 70108fb7a..cd25dd211 100644 --- a/playbooks/common/openshift-master/config.yml +++ b/playbooks/common/openshift-master/config.yml @@ -5,6 +5,19 @@ t_oo_option_master_debug_level: "{{ lookup('oo_option', 'openshift_master_debug_level') }}" pre_tasks: + # Per https://bugzilla.redhat.com/show_bug.cgi?id=1469336 + # + # When scaling up a cluster upgraded from OCP <= 3.5, ensure that + # OPENSHIFT_DEFAULT_REGISTRY is present as defined on the existing + # masters, or absent if such is the case. + - name: Detect if this host is a new master in a scale up + set_fact: + g_openshift_master_is_scaleup: "{{ openshift.common.hostname in ( groups['new_masters'] | default([]) ) }}" + + - name: Scaleup Detection + debug: + var: g_openshift_master_is_scaleup + - name: Check for RPM generated config marker file .config_managed stat: path: /etc/origin/.config_managed @@ -20,25 +33,6 @@ - node - .config_managed - - name: Check for existing configuration - stat: - path: /etc/origin/master/master-config.yaml - register: master_config_stat - - - name: Set clean install fact - set_fact: - l_clean_install: "{{ not master_config_stat.stat.exists | bool }}" - - - name: Determine if etcd3 storage is in use - command: grep -Pzo "storage-backend:\n.*etcd3" /etc/origin/master/master-config.yaml -q - register: etcd3_grep - failed_when: false - changed_when: false - - - name: Set etcd3 fact - set_fact: - l_etcd3_enabled: "{{ etcd3_grep.rc == 0 | bool }}" - - set_fact: openshift_master_pod_eviction_timeout: "{{ lookup('oo_option', 'openshift_master_pod_eviction_timeout') | default(none, true) }}" when: openshift_master_pod_eviction_timeout is not defined @@ -88,7 +82,7 @@ ha: "{{ openshift_master_ha | default(groups.oo_masters | length > 1) }}" master_count: "{{ openshift_master_count | default(groups.oo_masters | length) }}" -- name: Determine if session secrets must be generated +- name: Inspect state of first master config settings hosts: oo_first_master roles: - role: openshift_facts @@ -98,6 +92,60 @@ local_facts: session_auth_secrets: "{{ openshift_master_session_auth_secrets | default(openshift.master.session_auth_secrets | default(None)) }}" session_encryption_secrets: "{{ openshift_master_session_encryption_secrets | default(openshift.master.session_encryption_secrets | default(None)) }}" + - name: Check for existing configuration + stat: + path: /etc/origin/master/master-config.yaml + register: master_config_stat + + - name: Set clean install fact + set_fact: + l_clean_install: "{{ not master_config_stat.stat.exists | bool }}" + + - name: Determine if etcd3 storage is in use + command: grep -Pzo "storage-backend:\n.*etcd3" /etc/origin/master/master-config.yaml -q + register: etcd3_grep + failed_when: false + changed_when: false + + - name: Set etcd3 fact + set_fact: + l_etcd3_enabled: "{{ etcd3_grep.rc == 0 | bool }}" + + - name: Check if atomic-openshift-master sysconfig exists yet + stat: + path: /etc/sysconfig/atomic-openshift-master + register: l_aom_exists + + - name: Preserve OPENSHIFT_DEFAULT_REGISTRY master parameter if present + command: awk '/^OPENSHIFT_DEFAULT_REGISTRY/' /etc/sysconfig/atomic-openshift-master + register: l_default_registry_defined + when: l_aom_exists.stat.exists | bool + + - name: Check if atomic-openshift-master-api sysconfig exists yet + stat: + path: /etc/sysconfig/atomic-openshift-master-api + register: l_aom_api_exists + + - name: Preserve OPENSHIFT_DEFAULT_REGISTRY master-api parameter if present + command: awk '/^OPENSHIFT_DEFAULT_REGISTRY/' /etc/sysconfig/atomic-openshift-master-api + register: l_default_registry_defined_api + when: l_aom_api_exists.stat.exists | bool + + - name: Check if atomic-openshift-master-controllers sysconfig exists yet + stat: + path: /etc/sysconfig/atomic-openshift-master-controllers + register: l_aom_controllers_exists + + - name: Preserve OPENSHIFT_DEFAULT_REGISTRY master-controllers parameter if present + command: awk '/^OPENSHIFT_DEFAULT_REGISTRY/' /etc/sysconfig/atomic-openshift-master-controllers + register: l_default_registry_defined_controllers + when: l_aom_controllers_exists.stat.exists | bool + + - name: Update facts with OPENSHIFT_DEFAULT_REGISTRY value + set_fact: + l_default_registry_value: "{{ l_default_registry_defined.stdout | default('') }}" + l_default_registry_value_api: "{{ l_default_registry_defined_api.stdout | default('') }}" + l_default_registry_value_controllers: "{{ l_default_registry_defined_controllers.stdout | default('') }}" - name: Generate master session secrets hosts: oo_first_master @@ -128,7 +176,11 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" + openshift_no_proxy_etcd_host_ips: "{{ hostvars | oo_select_keys(groups['oo_etcd_to_config'] | default([])) + | oo_collect('openshift.common.ip') | default([]) | join(',') + }}" roles: + - role: os_firewall - role: openshift_master openshift_ca_host: "{{ groups.oo_first_master.0 }}" openshift_master_etcd_hosts: "{{ hostvars @@ -143,6 +195,10 @@ etcd_cert_prefix: "master.etcd-" r_openshift_master_clean_install: "{{ hostvars[groups.oo_first_master.0].l_clean_install }}" r_openshift_master_etcd3_storage: "{{ hostvars[groups.oo_first_master.0].l_etcd3_enabled }}" + openshift_master_is_scaleup_host: "{{ g_openshift_master_is_scaleup | default(false) }}" + openshift_master_default_registry_value: "{{ hostvars[groups.oo_first_master.0].l_default_registry_value }}" + openshift_master_default_registry_value_api: "{{ hostvars[groups.oo_first_master.0].l_default_registry_value_api }}" + openshift_master_default_registry_value_controllers: "{{ hostvars[groups.oo_first_master.0].l_default_registry_value_controllers }}" - role: nuage_master when: openshift.common.use_nuage | bool - role: calico_master diff --git a/playbooks/common/openshift-master/restart_hosts.yml b/playbooks/common/openshift-master/restart_hosts.yml index 67ba0aa2e..a5dbe0590 100644 --- a/playbooks/common/openshift-master/restart_hosts.yml +++ b/playbooks/common/openshift-master/restart_hosts.yml @@ -37,3 +37,4 @@ state: started delay: 10 port: "{{ openshift.master.api_port }}" + timeout: 600 diff --git a/playbooks/common/openshift-master/restart_services.yml b/playbooks/common/openshift-master/restart_services.yml index 508b5a3ac..4f8b758fd 100644 --- a/playbooks/common/openshift-master/restart_services.yml +++ b/playbooks/common/openshift-master/restart_services.yml @@ -1,9 +1,4 @@ --- -- name: Restart master - service: - name: "{{ openshift.common.service_type }}-master" - state: restarted - when: not openshift_master_ha | bool - name: Restart master API service: name: "{{ openshift.common.service_type }}-master-api" @@ -15,6 +10,7 @@ state: started delay: 10 port: "{{ openshift.master.api_port }}" + timeout: 600 when: openshift_master_ha | bool - name: Restart master controllers service: diff --git a/playbooks/common/openshift-master/scaleup.yml b/playbooks/common/openshift-master/scaleup.yml index bc61ee9bb..6ad4cde65 100644 --- a/playbooks/common/openshift-master/scaleup.yml +++ b/playbooks/common/openshift-master/scaleup.yml @@ -1,11 +1,4 @@ --- -- include: ../openshift-cluster/evaluate_groups.yml - -- name: Gather facts - hosts: oo_etcd_to_config:oo_masters_to_config:oo_nodes_to_config - roles: - - openshift_facts - - name: Update master count hosts: oo_masters:!oo_masters_to_config serial: 1 @@ -50,16 +43,6 @@ delay: 1 changed_when: false -- name: Configure docker hosts - hosts: oo_masters_to_config:oo_nodes_to_config - vars: - docker_additional_registries: "{{ lookup('oo_option', 'docker_additional_registries') | oo_split }}" - docker_insecure_registries: "{{ lookup('oo_option', 'docker_insecure_registries') | oo_split }}" - docker_blocked_registries: "{{ lookup('oo_option', 'docker_blocked_registries') | oo_split }}" - roles: - - openshift_facts - - openshift_docker - - name: Disable excluders hosts: oo_masters_to_config tags: diff --git a/playbooks/common/openshift-master/service.yml b/playbooks/common/openshift-master/service.yml deleted file mode 100644 index 48a2731aa..000000000 --- a/playbooks/common/openshift-master/service.yml +++ /dev/null @@ -1,23 +0,0 @@ ---- -- name: Populate g_service_masters host group if needed - hosts: localhost - gather_facts: no - connection: local - become: no - tasks: - - fail: msg="new_cluster_state is required to be injected in this playbook" - when: new_cluster_state is not defined - - - name: Evaluate g_service_masters - add_host: - name: "{{ item }}" - groups: g_service_masters - with_items: "{{ oo_host_group_exp | default([]) }}" - changed_when: False - -- name: Change state on master instance(s) - hosts: g_service_masters - connection: ssh - gather_facts: no - tasks: - - service: name={{ openshift.common.service_type }}-master state="{{ new_cluster_state }}" diff --git a/playbooks/common/openshift-nfs/service.yml b/playbooks/common/openshift-nfs/service.yml deleted file mode 100644 index b1e35e4b1..000000000 --- a/playbooks/common/openshift-nfs/service.yml +++ /dev/null @@ -1,21 +0,0 @@ ---- -- name: Populate g_service_nfs host group if needed - hosts: localhost - gather_facts: no - tasks: - - fail: msg="new_cluster_state is required to be injected in this playbook" - when: new_cluster_state is not defined - - - name: Evaluate g_service_nfs - add_host: - name: "{{ item }}" - groups: g_service_nfs - with_items: "{{ oo_host_group_exp | default([]) }}" - changed_when: False - -- name: Change state on nfs instance(s) - hosts: g_service_nfs - connection: ssh - gather_facts: no - tasks: - - service: name=nfs-server state="{{ new_cluster_state }}" diff --git a/playbooks/common/openshift-node/config.yml b/playbooks/common/openshift-node/config.yml index acebabc91..04c811c22 100644 --- a/playbooks/common/openshift-node/config.yml +++ b/playbooks/common/openshift-node/config.yml @@ -1,25 +1,4 @@ --- -- name: Gather and set facts for node hosts - hosts: oo_nodes_to_config - vars: - t_oo_option_node_debug_level: "{{ lookup('oo_option', 'openshift_node_debug_level') }}" - pre_tasks: - - set_fact: - openshift_node_debug_level: "{{ t_oo_option_node_debug_level }}" - when: openshift_node_debug_level is not defined and t_oo_option_node_debug_level != "" - roles: - - openshift_facts - tasks: - # Since the master is generating the node certificates before they are - # configured, we need to make sure to set the node properties beforehand if - # we do not want the defaults - - openshift_facts: - role: node - local_facts: - labels: "{{ openshift_node_labels | default(None) }}" - annotations: "{{ openshift_node_annotations | default(None) }}" - schedulable: "{{ openshift_schedulable | default(openshift_scheduleable) | default(None) }}" - - name: Evaluate node groups hosts: localhost become: no @@ -32,7 +11,11 @@ ansible_ssh_user: "{{ g_ssh_user | default(omit) }}" ansible_become: "{{ g_sudo | default(omit) }}" with_items: "{{ groups.oo_nodes_to_config | default([]) }}" - when: hostvars[item].openshift is defined and hostvars[item].openshift.common is defined and hostvars[item].openshift.common.is_containerized | bool and (item in groups.oo_nodes_to_config and item in groups.oo_masters_to_config) + when: + - hostvars[item].openshift is defined + - hostvars[item].openshift.common is defined + - hostvars[item].openshift.common.is_containerized | bool + - (item in groups.oo_nodes_to_config and item in groups.oo_masters_to_config) changed_when: False - name: Configure containerized nodes @@ -47,9 +30,9 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" + roles: + - role: os_firewall - role: openshift_node openshift_ca_host: "{{ groups.oo_first_master.0 }}" @@ -64,9 +47,8 @@ | union(groups['oo_etcd_to_config'] | default([]))) | oo_collect('openshift.common.hostname') | default([]) | join (',') }}" - when: "{{ (openshift_http_proxy is defined or openshift_https_proxy is defined) and - openshift_generate_no_proxy_hosts | default(True) | bool }}" roles: + - role: os_firewall - role: openshift_node openshift_ca_host: "{{ groups.oo_first_master.0 }}" diff --git a/playbooks/common/openshift-node/network_manager.yml b/playbooks/common/openshift-node/network_manager.yml index 0014a5dbd..b3a7399dc 100644 --- a/playbooks/common/openshift-node/network_manager.yml +++ b/playbooks/common/openshift-node/network_manager.yml @@ -1,4 +1,6 @@ --- +- include: ../openshift-cluster/evaluate_groups.yml + - name: Install and configure NetworkManager hosts: oo_all_hosts become: yes diff --git a/playbooks/common/openshift-node/restart.yml b/playbooks/common/openshift-node/restart.yml index 01cf948e0..c3beb59b7 100644 --- a/playbooks/common/openshift-node/restart.yml +++ b/playbooks/common/openshift-node/restart.yml @@ -11,6 +11,10 @@ service: name: docker state: restarted + register: l_docker_restart_docker_in_node_result + until: not l_docker_restart_docker_in_node_result | failed + retries: 3 + delay: 30 - name: Update docker facts openshift_facts: @@ -23,7 +27,6 @@ with_items: - etcd_container - openvswitch - - "{{ openshift.common.service_type }}-master" - "{{ openshift.common.service_type }}-master-api" - "{{ openshift.common.service_type }}-master-controllers" - "{{ openshift.common.service_type }}-node" @@ -36,6 +39,7 @@ state: started delay: 10 port: "{{ openshift.master.api_port }}" + timeout: 600 when: inventory_hostname in groups.oo_masters_to_config - name: restart node diff --git a/playbooks/common/openshift-node/scaleup.yml b/playbooks/common/openshift-node/scaleup.yml index 40da8990d..b1bbbb14c 100644 --- a/playbooks/common/openshift-node/scaleup.yml +++ b/playbooks/common/openshift-node/scaleup.yml @@ -1,32 +1,4 @@ --- -- include: ../openshift-cluster/evaluate_groups.yml - -- name: Gather facts - hosts: oo_etcd_to_config:oo_masters_to_config:oo_nodes_to_config - roles: - - openshift_facts - -- name: Gather and set facts for first master - hosts: oo_first_master - vars: - openshift_master_count: "{{ groups.oo_masters | length }}" - pre_tasks: - - set_fact: - openshift_master_default_subdomain: "{{ lookup('oo_option', 'openshift_master_default_subdomain') | default(None, true) }}" - when: openshift_master_default_subdomain is not defined - roles: - - openshift_master_facts - -- name: Configure docker hosts - hosts: oo_nodes_to_config - vars: - docker_additional_registries: "{{ lookup('oo_option', 'docker_additional_registries') | oo_split }}" - docker_insecure_registries: "{{ lookup('oo_option', 'docker_insecure_registries') | oo_split }}" - docker_blocked_registries: "{{ lookup('oo_option', 'docker_blocked_registries') | oo_split }}" - roles: - - openshift_facts - - openshift_docker - - name: Disable excluders hosts: oo_nodes_to_config tags: diff --git a/playbooks/common/openshift-node/service.yml b/playbooks/common/openshift-node/service.yml deleted file mode 100644 index 130a5416f..000000000 --- a/playbooks/common/openshift-node/service.yml +++ /dev/null @@ -1,26 +0,0 @@ ---- -- name: Populate g_service_nodes host group if needed - hosts: localhost - connection: local - become: no - gather_facts: no - tasks: - - fail: msg="new_cluster_state is required to be injected in this playbook" - when: new_cluster_state is not defined - - - name: Evaluate g_service_nodes - add_host: - name: "{{ item }}" - groups: g_service_nodes - with_items: "{{ oo_host_group_exp | default([]) }}" - changed_when: False - -- name: Change state on node instance(s) - hosts: g_service_nodes - connection: ssh - gather_facts: no - tasks: - - name: Change state on node instance(s) - service: - name: "{{ service_type }}-node" - state: "{{ new_cluster_state }}" |