diff options
41 files changed, 491 insertions, 702 deletions
diff --git a/README_CONTAINERIZED_INSTALLATION.md b/README_CONTAINERIZED_INSTALLATION.md index c615154ef..0a0ebb836 100644 --- a/README_CONTAINERIZED_INSTALLATION.md +++ b/README_CONTAINERIZED_INSTALLATION.md @@ -31,7 +31,7 @@ native clients. The wrapper scripts mount a limited subset of paths, _~/.kube_, _/etc/origin/_, and _/tmp_. Be mindful of this when passing in files to be processed by `oc` or `oadm`. You may find it easier to redirect input like this : - + `oc create -f - < my_file.json` ## Technical Notes @@ -48,18 +48,18 @@ before attempting to pull any of the following images. openshift/origin openshift/node (node + openshift-sdn + openvswitch rpm for client tools) openshift/openvswitch (centos7 + openvswitch rpm, runs ovsdb ovsctl processes) - registry.access.redhat.com/rhel7/etcd + registry.access.redhat.com/rhel7/etcd3 OpenShift Enterprise openshift3/ose openshift3/node openshift3/openvswitch - registry.access.redhat.com/rhel7/etcd + registry.access.redhat.com/rhel7/etcd3 Atomic Enterprise Platform aep3/aep aep3/node aep3/openvswitch - registry.access.redhat.com/rhel7/etcd - + registry.access.redhat.com/rhel7/etcd3 + * note openshift3/* and aep3/* images come from registry.access.redhat.com and rely on the --additional-repository flag being set appropriately. diff --git a/docs/best_practices_guide.adoc b/docs/best_practices_guide.adoc index cac9645a6..e9d904965 100644 --- a/docs/best_practices_guide.adoc +++ b/docs/best_practices_guide.adoc @@ -52,11 +52,11 @@ If mode lines for other editors are needed, please open a GitHub issue. === Method Signatures ''' -[[When-adding-a-new-paramemter-to-an-existing-method-a-default-value-SHOULD-be-used]] +[[When-adding-a-new-parameter-to-an-existing-method-a-default-value-SHOULD-be-used]] [cols="2v,v"] |=== -| <<When-adding-a-new-paramemter-to-an-existing-method-a-default-value-SHOULD-be-used, Rule>> -| When adding a new paramemter to an existing method, a default value SHOULD be used +| <<When-adding-a-new-parameter-to-an-existing-method-a-default-value-SHOULD-be-used, Rule>> +| When adding a new parameter to an existing method, a default value SHOULD be used |=== The purpose of this rule is to make it so that method signatures are backwards compatible. @@ -194,7 +194,7 @@ The purpose of this rule is to make it easy to include custom modules in our pla | Parameters to Ansible modules SHOULD use the Yaml dictionary format when 3 or more parameters are being passed |=== -When a module has several parameters that are being passed in, it's hard to see exactly what value each parameter is getting. It is preferred to use the Ansible Yaml syntax to pass in parameters so that it's more clear what values are being passed for each paramemter. +When a module has several parameters that are being passed in, it's hard to see exactly what value each parameter is getting. It is preferred to use the Ansible Yaml syntax to pass in parameters so that it's more clear what values are being passed for each parameter. .Bad: [source,yaml] @@ -222,7 +222,7 @@ When a module has several parameters that are being passed in, it's hard to see | Parameters to Ansible modules SHOULD use the Yaml dictionary format when the line length exceeds 120 characters |=== -Lines that are long quickly become a wall of text that isn't easily parsable. It is preferred to use the Ansible Yaml syntax to pass in parameters so that it's more clear what values are being passed for each paramemter. +Lines that are long quickly become a wall of text that isn't easily parsable. It is preferred to use the Ansible Yaml syntax to pass in parameters so that it's more clear what values are being passed for each parameter. .Bad: [source,yaml] @@ -432,7 +432,7 @@ This is very useful when developing and debugging new tasks. It can also signifi [[Ansible-Roles-SHOULD-be-named-like-technology_component_subcomponent]] [cols="2v,v"] |=== -| [[Ansible-Roles-SHOULD-be-named-like-technology_component_subcomponent, Rule]] +| <<Ansible-Roles-SHOULD-be-named-like-technology_component_subcomponent, Rule>> | Ansible Roles SHOULD be named like technology_component[_subcomponent]. |=== diff --git a/filter_plugins/oo_filters.py b/filter_plugins/oo_filters.py index 93fdd5ae4..38bc3ad6b 100644 --- a/filter_plugins/oo_filters.py +++ b/filter_plugins/oo_filters.py @@ -608,8 +608,8 @@ class FilterModule(object): host_type=_get_tag_value(host['group_names'], 'host-type'), sub_host_type=_get_tag_value(host['group_names'], 'sub-host-type'), host={'name': host['inventory_hostname'], - 'public IP': host['ansible_ssh_host'], - 'private IP': host['ansible_default_ipv4']['address']}) + 'public IP': host['oo_public_ipv4'], + 'private IP': host['oo_private_ipv4']}) except KeyError: pass return clusters @@ -889,11 +889,32 @@ class FilterModule(object): 'servers': FilterModule.oo_haproxy_backend_masters(servers_hostvars, nuage_rest_port)}) return loadbalancer_backends + @staticmethod + def oo_chomp_commit_offset(version): + """Chomp any "+git.foo" commit offset string from the given `version` + and return the modified version string. + + Ex: + - chomp_commit_offset(None) => None + - chomp_commit_offset(1337) => "1337" + - chomp_commit_offset("v3.4.0.15+git.derp") => "v3.4.0.15" + - chomp_commit_offset("v3.4.0.15") => "v3.4.0.15" + - chomp_commit_offset("v1.3.0+52492b4") => "v1.3.0" + """ + if version is None: + return version + else: + # Stringify, just in case it's a Number type. Split by '+' and + # return the first split. No concerns about strings without a + # '+', .split() returns an array of the original string. + return str(version).split('+')[0] + def filters(self): """ returns a mapping of filters to methods """ return { "oo_select_keys": self.oo_select_keys, "oo_select_keys_from_list": self.oo_select_keys_from_list, + "oo_chomp_commit_offset": self.oo_chomp_commit_offset, "oo_collect": self.oo_collect, "oo_flatten": self.oo_flatten, "oo_pdb": self.oo_pdb, diff --git a/playbooks/adhoc/uninstall.yml b/playbooks/adhoc/uninstall.yml index 789f66b14..4ea639cbe 100644 --- a/playbooks/adhoc/uninstall.yml +++ b/playbooks/adhoc/uninstall.yml @@ -338,6 +338,7 @@ failed_when: False with_items: - etcd + - etcd3 - firewalld - name: Stop additional atomic services @@ -352,6 +353,7 @@ when: not is_atomic | bool with_items: - etcd + - etcd3 - shell: systemctl reset-failed changed_when: False @@ -365,6 +367,7 @@ - /etc/ansible/facts.d/openshift.fact - /etc/etcd - /etc/systemd/system/etcd_container.service + - /etc/profile.d/etcdctl.sh # Intenationally using rm command over file module because if someone had mounted a filesystem # at /var/lib/etcd then the contents was not removed correctly diff --git a/playbooks/aws/openshift-cluster/list.yml b/playbooks/aws/openshift-cluster/list.yml index 4934ae6d0..ed8aac398 100644 --- a/playbooks/aws/openshift-cluster/list.yml +++ b/playbooks/aws/openshift-cluster/list.yml @@ -16,11 +16,8 @@ groups: oo_list_hosts ansible_ssh_user: "{{ deployment_vars[deployment_type].ssh_user }}" ansible_become: "{{ deployment_vars[deployment_type].become }}" + oo_public_ipv4: "{{ hostvars[item].ec2_ip_address }}" + oo_private_ipv4: "{{ hostvars[item].ec2_private_ip_address }}" with_items: "{{ groups[scratch_group] | default([]) | difference(['localhost']) }}" - -- name: List Hosts - hosts: oo_list_hosts - gather_facts: no - tasks: - debug: - msg: "public ip:{{ hostvars[inventory_hostname].ec2_ip_address }} private ip:{{ hostvars[inventory_hostname].ec2_private_ip_address }}" + msg: "{{ hostvars | oo_select_keys(groups[scratch_group] | default([])) | oo_pretty_print_cluster }}" diff --git a/playbooks/byo/openshift-cluster/upgrades/upgrade_etcd.yml b/playbooks/byo/openshift-cluster/upgrades/upgrade_etcd.yml new file mode 100644 index 000000000..c25f96212 --- /dev/null +++ b/playbooks/byo/openshift-cluster/upgrades/upgrade_etcd.yml @@ -0,0 +1,26 @@ +--- +- include: ../../../common/openshift-cluster/verify_ansible_version.yml + +- name: Create initial host groups for localhost + hosts: localhost + connection: local + become: no + gather_facts: no + tags: + - always + tasks: + - include_vars: ../cluster_hosts.yml + - add_host: + name: "{{ item }}" + groups: l_oo_all_hosts + with_items: "{{ g_all_hosts | default([]) }}" + +- name: Create initial host groups for all hosts + hosts: l_oo_all_hosts + gather_facts: no + tags: + - always + tasks: + - include_vars: ../cluster_hosts.yml + +- include: ../../../common/openshift-cluster/upgrades/etcd/main.yml diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/backup.yml b/playbooks/common/openshift-cluster/upgrades/etcd/backup.yml new file mode 100644 index 000000000..57b156b1c --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/backup.yml @@ -0,0 +1,73 @@ +- name: Backup etcd + hosts: etcd_hosts_to_backup + vars: + embedded_etcd: "{{ hostvars[groups.oo_first_master.0].openshift.master.embedded_etcd }}" + timestamp: "{{ lookup('pipe', 'date +%Y%m%d%H%M%S') }}" + roles: + - openshift_facts + tasks: + # Ensure we persist the etcd role for this host in openshift_facts + - openshift_facts: + role: etcd + local_facts: {} + when: "'etcd' not in openshift" + + - stat: path=/var/lib/openshift + register: var_lib_openshift + + - stat: path=/var/lib/origin + register: var_lib_origin + + - name: Create origin symlink if necessary + file: src=/var/lib/openshift/ dest=/var/lib/origin state=link + when: var_lib_openshift.stat.exists == True and var_lib_origin.stat.exists == False + + # TODO: replace shell module with command and update later checks + # We assume to be using the data dir for all backups. + - name: Check available disk space for etcd backup + shell: df --output=avail -k {{ openshift.common.data_dir }} | tail -n 1 + register: avail_disk + + # TODO: replace shell module with command and update later checks + - name: Check current embedded etcd disk usage + shell: du -k {{ openshift.etcd.etcd_data_dir }} | tail -n 1 | cut -f1 + register: etcd_disk_usage + when: embedded_etcd | bool + + - name: Abort if insufficient disk space for etcd backup + fail: + msg: > + {{ etcd_disk_usage.stdout }} Kb disk space required for etcd backup, + {{ avail_disk.stdout }} Kb available. + when: (embedded_etcd | bool) and (etcd_disk_usage.stdout|int > avail_disk.stdout|int) + + - name: Install etcd (for etcdctl) + action: "{{ ansible_pkg_mgr }} name=etcd state=present" + when: not openshift.common.is_atomic | bool + + - name: Generate etcd backup + command: > + etcdctl backup --data-dir={{ openshift.etcd.etcd_data_dir }} + --backup-dir={{ openshift.common.data_dir }}/etcd-backup-{{ backup_tag | default('') }}{{ timestamp }} + + - set_fact: + etcd_backup_complete: True + + - name: Display location of etcd backup + debug: + msg: "Etcd backup created in {{ openshift.common.data_dir }}/etcd-backup-{{ backup_tag | default('') }}{{ timestamp }}" + +- name: Gate on etcd backup + hosts: localhost + connection: local + become: no + tasks: + - set_fact: + etcd_backup_completed: "{{ hostvars + | oo_select_keys(groups.etcd_hosts_to_backup) + | oo_collect('inventory_hostname', {'etcd_backup_complete': true}) }}" + - set_fact: + etcd_backup_failed: "{{ groups.etcd_hosts_to_backup | difference(etcd_backup_completed) }}" + - fail: + msg: "Upgrade cannot continue. The following hosts did not complete etcd backup: {{ etcd_backup_failed | join(',') }}" + when: etcd_backup_failed | length > 0 diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/containerized_tasks.yml b/playbooks/common/openshift-cluster/upgrades/etcd/containerized_tasks.yml new file mode 100644 index 000000000..35f391f8c --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/containerized_tasks.yml @@ -0,0 +1,47 @@ +--- +- name: Verify cluster is healthy pre-upgrade + command: "etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://{{ openshift.common.hostname }}:2379 cluster-health" + +- name: Get current image + shell: grep 'ExecStart=' /etc/systemd/system/etcd_container.service | awk '{print $NF}' + register: current_image + +- name: Set new_etcd_image + set_fact: + new_etcd_image: "{{ current_image.stdout | regex_replace('/etcd.*$','/etcd3:' ~ upgrade_version ) if upgrade_version | version_compare('3.0','>=') + else current_image.stdout.split(':')[0] ~ ':' ~ upgrade_version }}" + +- name: Pull new etcd image + command: "docker pull {{ new_etcd_image }}" + +- name: Update to latest etcd image + replace: + dest: /etc/systemd/system/etcd_container.service + regexp: "{{ current_image.stdout }}$" + replace: "{{ new_etcd_image }}" + +- name: Restart etcd_container + systemd: + name: etcd_container + daemon_reload: yes + state: restarted + +## TODO: probably should just move this into the backup playbooks, also this +## will fail on atomic host. We need to revisit how to do etcd backups there as +## the container may be newer than etcdctl on the host. Assumes etcd3 obsoletes etcd (7.3.1) +- name: Upgrade etcd for etcdctl when not atomic + action: "{{ ansible_pkg_mgr }} name=etcd ensure=latest" + when: not openshift.common.is_atomic | bool + +- name: Verify cluster is healthy + command: "etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://{{ openshift.common.hostname }}:2379 cluster-health" + register: etcdctl + until: etcdctl.rc == 0 + retries: 3 + delay: 10 + +- name: Store new etcd_image + openshift_facts: + role: etcd + local_facts: + etcd_image: "{{ new_etcd_image }}" diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/fedora_tasks.yml b/playbooks/common/openshift-cluster/upgrades/etcd/fedora_tasks.yml new file mode 100644 index 000000000..30232110e --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/fedora_tasks.yml @@ -0,0 +1,23 @@ +--- +# F23 GA'd with etcd 2.0, currently has 2.2 in updates +# F24 GA'd with etcd-2.2, currently has 2.2 in updates +# F25 Beta currently has etcd 3.0 +- name: Verify cluster is healthy pre-upgrade + command: "etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://{{ openshift.common.hostname }}:2379 cluster-health" + +- name: Update etcd + package: + name: "etcd" + state: "latest" + +- name: Restart etcd + service: + name: etcd + state: restarted + +- name: Verify cluster is healthy + command: "etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://{{ openshift.common.hostname }}:2379 cluster-health" + register: etcdctl + until: etcdctl.rc == 0 + retries: 3 + delay: 10 diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/files/etcdctl.sh b/playbooks/common/openshift-cluster/upgrades/etcd/files/etcdctl.sh new file mode 120000 index 000000000..641e04e44 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/files/etcdctl.sh @@ -0,0 +1 @@ +../roles/etcd/files/etcdctl.sh
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/filter_plugins b/playbooks/common/openshift-cluster/upgrades/etcd/filter_plugins new file mode 120000 index 000000000..27ddaa18b --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/filter_plugins @@ -0,0 +1 @@ +../../../../../filter_plugins
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/lookup_plugins b/playbooks/common/openshift-cluster/upgrades/etcd/lookup_plugins new file mode 120000 index 000000000..cf407f69b --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/lookup_plugins @@ -0,0 +1 @@ +../../../../../lookup_plugins
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/main.yml b/playbooks/common/openshift-cluster/upgrades/etcd/main.yml new file mode 100644 index 000000000..cce844403 --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/main.yml @@ -0,0 +1,122 @@ +--- +# For 1.4/3.4 we want to upgrade everyone to etcd-3.0. etcd docs say to +# upgrade from 2.0.x to 2.1.x to 2.2.x to 2.3.x to 3.0.x. While this is a tedius +# task for RHEL and CENTOS it's simply not possible in Fedora unless you've +# mirrored packages on your own because only the GA and latest versions are +# available in the repos. So for Fedora we'll simply skip this, sorry. + +- include: ../../evaluate_groups.yml + tags: + - always + +- name: Evaluate additional groups for upgrade + hosts: localhost + connection: local + become: no + tasks: + - name: Evaluate etcd_hosts_to_upgrade + add_host: + name: "{{ item }}" + groups: etcd_hosts_to_upgrade, etcd_hosts_to_backup + with_items: "{{ groups.oo_etcd_to_config if groups.oo_etcd_to_config is defined and groups.oo_etcd_to_config | length > 0 else groups.oo_first_master }}" + +- name: Backup etcd before upgrading anything + include: backup.yml + vars: + backup_tag: "pre-upgrade-" + +- name: Drop etcdctl profiles + hosts: etcd_hosts_to_upgrade + tasks: + - include: roles/etcd/tasks/etcdctl.yml + +- name: Determine etcd version + hosts: etcd_hosts_to_upgrade + tasks: + - name: Record RPM based etcd version + command: rpm -qa --qf '%{version}' etcd\* + register: etcd_installed_version + failed_when: false + when: not openshift.common.is_containerized | bool + - name: Record containerized etcd version + command: docker exec etcd_container rpm -qa --qf '%{version}' etcd\* + register: etcd_installed_version + failed_when: false + when: openshift.common.is_containerized | bool + +# I really dislike this copy/pasta but I wasn't able to find a way to get it to loop +# through hosts, then loop through tasks only when appropriate +- name: Upgrade to 2.1 + hosts: etcd_hosts_to_upgrade + serial: 1 + vars: + upgrade_version: '2.1' + tasks: + - include: rhel_tasks.yml + when: etcd_installed_version.stdout | default('99') | version_compare('2.1','<') and ansible_distribution == 'RedHat' and not openshift.common.is_containerized | bool + +- name: Upgrade RPM hosts to 2.2 + hosts: etcd_hosts_to_upgrade + serial: 1 + vars: + upgrade_version: '2.2' + tasks: + - include: rhel_tasks.yml + when: etcd_installed_version.stdout | default('99') | version_compare('2.2','<') and ansible_distribution == 'RedHat' and not openshift.common.is_containerized | bool + +- name: Upgrade containerized hosts to 2.2.5 + hosts: etcd_hosts_to_upgrade + serial: 1 + vars: + upgrade_version: 2.2.5 + tasks: + - include: containerized_tasks.yml + when: etcd_installed_version.stdout | default('99') | version_compare('2.2','<') and openshift.common.is_containerized | bool + +- name: Upgrade RPM hosts to 2.3 + hosts: etcd_hosts_to_upgrade + serial: 1 + vars: + upgrade_version: '2.3' + tasks: + - include: rhel_tasks.yml + when: etcd_installed_version.stdout | default('99') | version_compare('2.3','<') and ansible_distribution == 'RedHat' and not openshift.common.is_containerized | bool + +- name: Upgrade containerized hosts to 2.3.7 + hosts: etcd_hosts_to_upgrade + serial: 1 + vars: + upgrade_version: 2.3.7 + tasks: + - include: containerized_tasks.yml + when: etcd_installed_version.stdout | default('99') | version_compare('2.3','<') and openshift.common.is_containerized | bool + +- name: Upgrade RPM hosts to 3.0 + hosts: etcd_hosts_to_upgrade + serial: 1 + vars: + upgrade_version: '3.0' + tasks: + - include: rhel_tasks.yml + when: etcd_installed_version.stdout | default('99') | version_compare('3.0','<') and ansible_distribution == 'RedHat' and not openshift.common.is_containerized | bool + +- name: Upgrade containerized hosts to etcd3 image + hosts: etcd_hosts_to_upgrade + serial: 1 + vars: + upgrade_version: 3.0.3 + tasks: + - include: containerized_tasks.yml + when: etcd_installed_version.stdout | default('99') | version_compare('3.0','<') and openshift.common.is_containerized | bool + +- name: Upgrade fedora to latest + hosts: etcd_hosts_to_upgrade + serial: 1 + tasks: + - include: fedora_tasks.yml + when: ansible_distribution == 'Fedora' and not openshift.common.is_containerized | bool + +- name: Backup etcd + include: backup.yml + vars: + backup_tag: "post-3.0-" diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/rhel_tasks.yml b/playbooks/common/openshift-cluster/upgrades/etcd/rhel_tasks.yml new file mode 100644 index 000000000..8e7dc9d9b --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/rhel_tasks.yml @@ -0,0 +1,23 @@ +--- +- name: Verify cluster is healthy pre-upgrade + command: "etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://{{ openshift.common.hostname }}:2379 cluster-health" + +- name: Update etcd package but exclude etcd3 + command: "{{ ansible_pkg_mgr }} install -y etcd-{{ upgrade_version }}\\* --exclude etcd3" + when: upgrade_version | version_compare('3.0','<') + +- name: Update etcd package not excluding etcd3 + command: "{{ ansible_pkg_mgr }} install -y etcd3-{{ upgrade_version }}\\*" + when: not upgrade_version | version_compare('3.0','<') + +- name: Restart etcd + service: + name: etcd + state: restarted + +- name: Verify cluster is healthy + command: "etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://{{ openshift.common.hostname }}:2379 cluster-health" + register: etcdctl + until: etcdctl.rc == 0 + retries: 3 + delay: 10 diff --git a/playbooks/common/openshift-cluster/upgrades/etcd/roles b/playbooks/common/openshift-cluster/upgrades/etcd/roles new file mode 120000 index 000000000..6bc1a7aef --- /dev/null +++ b/playbooks/common/openshift-cluster/upgrades/etcd/roles @@ -0,0 +1 @@ +../../../../../roles
\ No newline at end of file diff --git a/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml b/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml index 764563d28..57c25aa41 100644 --- a/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml +++ b/playbooks/common/openshift-cluster/upgrades/upgrade_control_plane.yml @@ -25,81 +25,10 @@ role: master local_facts: embedded_etcd: "{{ groups.oo_etcd_to_config | default([]) | length == 0 }}" + debug_level: "{{ openshift_master_debug_level | default(openshift.common.debug_level | default(2)) }}" - name: Backup etcd - hosts: etcd_hosts_to_backup - vars: - embedded_etcd: "{{ groups.oo_etcd_to_config | default([]) | length == 0 }}" - timestamp: "{{ lookup('pipe', 'date +%Y%m%d%H%M%S') }}" - roles: - - openshift_facts - tasks: - # Ensure we persist the etcd role for this host in openshift_facts - - openshift_facts: - role: etcd - local_facts: {} - when: "'etcd' not in openshift" - - - stat: path=/var/lib/openshift - register: var_lib_openshift - - - stat: path=/var/lib/origin - register: var_lib_origin - - - name: Create origin symlink if necessary - file: src=/var/lib/openshift/ dest=/var/lib/origin state=link - when: var_lib_openshift.stat.exists == True and var_lib_origin.stat.exists == False - - # TODO: replace shell module with command and update later checks - # We assume to be using the data dir for all backups. - - name: Check available disk space for etcd backup - shell: df --output=avail -k {{ openshift.common.data_dir }} | tail -n 1 - register: avail_disk - - # TODO: replace shell module with command and update later checks - - name: Check current embedded etcd disk usage - shell: du -k {{ openshift.etcd.etcd_data_dir }} | tail -n 1 | cut -f1 - register: etcd_disk_usage - when: embedded_etcd | bool - - - name: Abort if insufficient disk space for etcd backup - fail: - msg: > - {{ etcd_disk_usage.stdout }} Kb disk space required for etcd backup, - {{ avail_disk.stdout }} Kb available. - when: (embedded_etcd | bool) and (etcd_disk_usage.stdout|int > avail_disk.stdout|int) - - - name: Install etcd (for etcdctl) - action: "{{ ansible_pkg_mgr }} name=etcd state=installed" - when: not openshift.common.is_atomic | bool - - - name: Generate etcd backup - command: > - etcdctl backup --data-dir={{ openshift.etcd.etcd_data_dir }} - --backup-dir={{ openshift.common.data_dir }}/etcd-backup-{{ timestamp }} - - - set_fact: - etcd_backup_complete: True - - - name: Display location of etcd backup - debug: - msg: "Etcd backup created in {{ openshift.common.data_dir }}/etcd-backup-{{ timestamp }}" - - -- name: Gate on etcd backup - hosts: localhost - connection: local - become: no - tasks: - - set_fact: - etcd_backup_completed: "{{ hostvars - | oo_select_keys(groups.etcd_hosts_to_backup) - | oo_collect('inventory_hostname', {'etcd_backup_complete': true}) }}" - - set_fact: - etcd_backup_failed: "{{ groups.etcd_hosts_to_backup | difference(etcd_backup_completed) }}" - - fail: - msg: "Upgrade cannot continue. The following hosts did not complete etcd backup: {{ etcd_backup_failed | join(',') }}" - when: etcd_backup_failed | length > 0 + include: ./etcd/backup.yml - name: Upgrade master packages hosts: oo_masters_to_config diff --git a/playbooks/common/openshift-master/config.yml b/playbooks/common/openshift-master/config.yml index a53c55c14..5fcb850a2 100644 --- a/playbooks/common/openshift-master/config.yml +++ b/playbooks/common/openshift-master/config.yml @@ -53,7 +53,7 @@ when: openshift_hosted_metrics_deployer_prefix is not defined - set_fact: openshift_hosted_metrics_deployer_version: "{{ lookup('oo_option', 'openshift_hosted_metrics_deployer_version') | default('latest') }}" - when: openshift_hosted_metrics_deployer_prefix is not defined + when: openshift_hosted_metrics_deployer_version is not defined roles: - openshift_facts post_tasks: diff --git a/playbooks/gce/openshift-cluster/library/gce.py b/playbooks/gce/openshift-cluster/library/gce.py deleted file mode 100644 index fcaa3b850..000000000 --- a/playbooks/gce/openshift-cluster/library/gce.py +++ /dev/null @@ -1,543 +0,0 @@ -#!/usr/bin/python -# Copyright 2013 Google Inc. -# -# This file is part of Ansible -# -# Ansible is free software: you can redistribute it and/or modify -# it under the terms of the GNU General Public License as published by -# the Free Software Foundation, either version 3 of the License, or -# (at your option) any later version. -# -# Ansible is distributed in the hope that it will be useful, -# but WITHOUT ANY WARRANTY; without even the implied warranty of -# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the -# GNU General Public License for more details. -# -# You should have received a copy of the GNU General Public License -# along with Ansible. If not, see <http://www.gnu.org/licenses/>. - -DOCUMENTATION = ''' ---- -module: gce -version_added: "1.4" -short_description: create or terminate GCE instances -description: - - Creates or terminates Google Compute Engine (GCE) instances. See - U(https://cloud.google.com/products/compute-engine) for an overview. - Full install/configuration instructions for the gce* modules can - be found in the comments of ansible/test/gce_tests.py. -options: - image: - description: - - image string to use for the instance - required: false - default: "debian-7" - instance_names: - description: - - a comma-separated list of instance names to create or destroy - required: false - default: null - machine_type: - description: - - machine type to use for the instance, use 'n1-standard-1' by default - required: false - default: "n1-standard-1" - metadata: - description: - - a hash/dictionary of custom data for the instance; - '{"key":"value", ...}' - required: false - default: null - service_account_email: - version_added: "1.5.1" - description: - - service account email - required: false - default: null - service_account_permissions: - version_added: "2.0" - description: - - service account permissions (see - U(https://cloud.google.com/sdk/gcloud/reference/compute/instances/create), - --scopes section for detailed information) - required: false - default: null - choices: [ - "bigquery", "cloud-platform", "compute-ro", "compute-rw", - "computeaccounts-ro", "computeaccounts-rw", "datastore", "logging-write", - "monitoring", "sql", "sql-admin", "storage-full", "storage-ro", - "storage-rw", "taskqueue", "userinfo-email" - ] - pem_file: - version_added: "1.5.1" - description: - - path to the pem file associated with the service account email - required: false - default: null - project_id: - version_added: "1.5.1" - description: - - your GCE project ID - required: false - default: null - name: - description: - - identifier when working with a single instance - required: false - network: - description: - - name of the network, 'default' will be used if not specified - required: false - default: "default" - persistent_boot_disk: - description: - - if set, create the instance with a persistent boot disk - required: false - default: "false" - disks: - description: - - a list of persistent disks to attach to the instance; a string value - gives the name of the disk; alternatively, a dictionary value can - define 'name' and 'mode' ('READ_ONLY' or 'READ_WRITE'). The first entry - will be the boot disk (which must be READ_WRITE). - required: false - default: null - version_added: "1.7" - state: - description: - - desired state of the resource - required: false - default: "present" - choices: ["active", "present", "absent", "deleted"] - tags: - description: - - a comma-separated list of tags to associate with the instance - required: false - default: null - zone: - description: - - the GCE zone to use - required: true - default: "us-central1-a" - ip_forward: - version_added: "1.9" - description: - - set to true if the instance can forward ip packets (useful for - gateways) - required: false - default: "false" - external_ip: - version_added: "1.9" - description: - - type of external ip, ephemeral by default - required: false - default: "ephemeral" - disk_auto_delete: - version_added: "1.9" - description: - - if set boot disk will be removed after instance destruction - required: false - default: "true" - -requirements: - - "python >= 2.6" - - "apache-libcloud >= 0.13.3" -notes: - - Either I(name) or I(instance_names) is required. -author: "Eric Johnson (@erjohnso) <erjohnso@google.com>" -''' - -EXAMPLES = ''' -# Basic provisioning example. Create a single Debian 7 instance in the -# us-central1-a Zone of n1-standard-1 machine type. -- local_action: - module: gce - name: test-instance - zone: us-central1-a - machine_type: n1-standard-1 - image: debian-7 - -# Example using defaults and with metadata to create a single 'foo' instance -- local_action: - module: gce - name: foo - metadata: '{"db":"postgres", "group":"qa", "id":500}' - - -# Launch instances from a control node, runs some tasks on the new instances, -# and then terminate them -- name: Create a sandbox instance - hosts: localhost - vars: - names: foo,bar - machine_type: n1-standard-1 - image: debian-6 - zone: us-central1-a - service_account_email: unique-email@developer.gserviceaccount.com - pem_file: /path/to/pem_file - project_id: project-id - tasks: - - name: Launch instances - local_action: gce instance_names={{names}} machine_type={{machine_type}} - image={{image}} zone={{zone}} - service_account_email={{ service_account_email }} - pem_file={{ pem_file }} project_id={{ project_id }} - register: gce - - name: Wait for SSH to come up - local_action: wait_for host={{item.public_ip}} port=22 delay=10 - timeout=60 state=started - with_items: {{gce.instance_data}} - -- name: Configure instance(s) - hosts: launched - sudo: True - roles: - - my_awesome_role - - my_awesome_tasks - -- name: Terminate instances - hosts: localhost - connection: local - tasks: - - name: Terminate instances that were previously launched - local_action: - module: gce - state: 'absent' - instance_names: {{gce.instance_names}} - -''' - -try: - import libcloud - from libcloud.compute.types import Provider - from libcloud.compute.providers import get_driver - from libcloud.common.google import GoogleBaseError, QuotaExceededError, \ - ResourceExistsError, ResourceInUseError, ResourceNotFoundError - _ = Provider.GCE - HAS_LIBCLOUD = True -except ImportError: - HAS_LIBCLOUD = False - -try: - from ast import literal_eval - HAS_PYTHON26 = True -except ImportError: - HAS_PYTHON26 = False - - -def get_instance_info(inst): - """Retrieves instance information from an instance object and returns it - as a dictionary. - - """ - metadata = {} - if 'metadata' in inst.extra and 'items' in inst.extra['metadata']: - for md in inst.extra['metadata']['items']: - metadata[md['key']] = md['value'] - - try: - netname = inst.extra['networkInterfaces'][0]['network'].split('/')[-1] - except: - netname = None - if 'disks' in inst.extra: - disk_names = [disk_info['source'].split('/')[-1] - for disk_info - in sorted(inst.extra['disks'], - key=lambda disk_info: disk_info['index'])] - else: - disk_names = [] - - if len(inst.public_ips) == 0: - public_ip = None - else: - public_ip = inst.public_ips[0] - - return({ - 'image': inst.image is not None and inst.image.split('/')[-1] or None, - 'disks': disk_names, - 'machine_type': inst.size, - 'metadata': metadata, - 'name': inst.name, - 'network': netname, - 'private_ip': inst.private_ips[0], - 'public_ip': public_ip, - 'status': ('status' in inst.extra) and inst.extra['status'] or None, - 'tags': ('tags' in inst.extra) and inst.extra['tags'] or [], - 'zone': ('zone' in inst.extra) and inst.extra['zone'].name or None, - }) - - -def create_instances(module, gce, instance_names): - """Creates new instances. Attributes other than instance_names are picked - up from 'module' - - module : AnsibleModule object - gce: authenticated GCE libcloud driver - instance_names: python list of instance names to create - - Returns: - A list of dictionaries with instance information - about the instances that were launched. - - """ - image = module.params.get('image') - machine_type = module.params.get('machine_type') - metadata = module.params.get('metadata') - network = module.params.get('network') - persistent_boot_disk = module.params.get('persistent_boot_disk') - disks = module.params.get('disks') - state = module.params.get('state') - tags = module.params.get('tags') - zone = module.params.get('zone') - ip_forward = module.params.get('ip_forward') - external_ip = module.params.get('external_ip') - disk_auto_delete = module.params.get('disk_auto_delete') - service_account_permissions = module.params.get('service_account_permissions') - service_account_email = module.params.get('service_account_email') - - if external_ip == "none": - external_ip = None - - new_instances = [] - changed = False - - lc_image = gce.ex_get_image(image) - lc_disks = [] - disk_modes = [] - for i, disk in enumerate(disks or []): - if isinstance(disk, dict): - lc_disks.append(gce.ex_get_volume(disk['name'])) - disk_modes.append(disk['mode']) - else: - lc_disks.append(gce.ex_get_volume(disk)) - # boot disk is implicitly READ_WRITE - disk_modes.append('READ_ONLY' if i > 0 else 'READ_WRITE') - lc_network = gce.ex_get_network(network) - lc_machine_type = gce.ex_get_size(machine_type) - lc_zone = gce.ex_get_zone(zone) - - # Try to convert the user's metadata value into the format expected - # by GCE. First try to ensure user has proper quoting of a - # dictionary-like syntax using 'literal_eval', then convert the python - # dict into a python list of 'key' / 'value' dicts. Should end up - # with: - # [ {'key': key1, 'value': value1}, {'key': key2, 'value': value2}, ...] - if metadata: - if isinstance(metadata, dict): - md = metadata - else: - try: - md = literal_eval(str(metadata)) - if not isinstance(md, dict): - raise ValueError('metadata must be a dict') - except ValueError as e: - module.fail_json(msg='bad metadata: %s' % str(e)) - except SyntaxError as e: - module.fail_json(msg='bad metadata syntax') - - if hasattr(libcloud, '__version__') and libcloud.__version__ < '0.15': - items = [] - for k, v in md.items(): - items.append({"key": k, "value": v}) - metadata = {'items': items} - else: - metadata = md - - ex_sa_perms = [] - bad_perms = [] - if service_account_permissions: - for perm in service_account_permissions: - if perm not in gce.SA_SCOPES_MAP.keys(): - bad_perms.append(perm) - if len(bad_perms) > 0: - module.fail_json(msg='bad permissions: %s' % str(bad_perms)) - if service_account_email: - ex_sa_perms.append({'email': service_account_email}) - else: - ex_sa_perms.append({'email': "default"}) - ex_sa_perms[0]['scopes'] = service_account_permissions - - # These variables all have default values but check just in case - if not lc_image or not lc_network or not lc_machine_type or not lc_zone: - module.fail_json(msg='Missing required create instance variable', - changed=False) - - for name in instance_names: - pd = None - if lc_disks: - pd = lc_disks[0] - elif persistent_boot_disk: - try: - pd = gce.create_volume(None, "%s" % name, image=lc_image) - except ResourceExistsError: - pd = gce.ex_get_volume("%s" % name, lc_zone) - inst = None - try: - inst = gce.create_node( - name, lc_machine_type, lc_image, location=lc_zone, - ex_network=network, ex_tags=tags, ex_metadata=metadata, - ex_boot_disk=pd, ex_can_ip_forward=ip_forward, - external_ip=external_ip, ex_disk_auto_delete=disk_auto_delete, - ex_service_accounts=ex_sa_perms - ) - changed = True - except ResourceExistsError: - inst = gce.ex_get_node(name, lc_zone) - except GoogleBaseError as e: - module.fail_json(msg='Unexpected error attempting to create ' + - 'instance %s, error: %s' % (name, e.value)) - - for i, lc_disk in enumerate(lc_disks): - # Check whether the disk is already attached - if (len(inst.extra['disks']) > i): - attached_disk = inst.extra['disks'][i] - if attached_disk['source'] != lc_disk.extra['selfLink']: - module.fail_json( - msg=("Disk at index %d does not match: requested=%s found=%s" % ( - i, lc_disk.extra['selfLink'], attached_disk['source']))) - elif attached_disk['mode'] != disk_modes[i]: - module.fail_json( - msg=("Disk at index %d is in the wrong mode: requested=%s found=%s" % ( - i, disk_modes[i], attached_disk['mode']))) - else: - continue - gce.attach_volume(inst, lc_disk, ex_mode=disk_modes[i]) - # Work around libcloud bug: attached volumes don't get added - # to the instance metadata. get_instance_info() only cares about - # source and index. - if len(inst.extra['disks']) != i+1: - inst.extra['disks'].append( - {'source': lc_disk.extra['selfLink'], 'index': i}) - - if inst: - new_instances.append(inst) - - instance_names = [] - instance_json_data = [] - for inst in new_instances: - d = get_instance_info(inst) - instance_names.append(d['name']) - instance_json_data.append(d) - - return (changed, instance_json_data, instance_names) - - -def terminate_instances(module, gce, instance_names, zone_name): - """Terminates a list of instances. - - module: Ansible module object - gce: authenticated GCE connection object - instance_names: a list of instance names to terminate - zone_name: the zone where the instances reside prior to termination - - Returns a dictionary of instance names that were terminated. - - """ - changed = False - terminated_instance_names = [] - for name in instance_names: - inst = None - try: - inst = gce.ex_get_node(name, zone_name) - except ResourceNotFoundError: - pass - except Exception as e: - module.fail_json(msg=unexpected_error_msg(e), changed=False) - if inst: - gce.destroy_node(inst) - terminated_instance_names.append(inst.name) - changed = True - - return (changed, terminated_instance_names) - - -def main(): - module = AnsibleModule( - argument_spec=dict( - image=dict(default='debian-7'), - instance_names=dict(), - machine_type=dict(default='n1-standard-1'), - metadata=dict(), - name=dict(), - network=dict(default='default'), - persistent_boot_disk=dict(type='bool', default=False), - disks=dict(type='list'), - state=dict(choices=['active', 'present', 'absent', 'deleted'], - default='present'), - tags=dict(type='list'), - zone=dict(default='us-central1-a'), - service_account_email=dict(), - service_account_permissions=dict(type='list'), - pem_file=dict(), - project_id=dict(), - ip_forward=dict(type='bool', default=False), - external_ip=dict(choices=['ephemeral', 'none'], - default='ephemeral'), - disk_auto_delete=dict(type='bool', default=True), - ) - ) - - if not HAS_PYTHON26: - module.fail_json(msg="GCE module requires python's 'ast' module, python v2.6+") - if not HAS_LIBCLOUD: - module.fail_json(msg='libcloud with GCE support (0.13.3+) required for this module') - - gce = gce_connect(module) - - image = module.params.get('image') - instance_names = module.params.get('instance_names') - machine_type = module.params.get('machine_type') - metadata = module.params.get('metadata') - name = module.params.get('name') - network = module.params.get('network') - persistent_boot_disk = module.params.get('persistent_boot_disk') - state = module.params.get('state') - tags = module.params.get('tags') - zone = module.params.get('zone') - ip_forward = module.params.get('ip_forward') - changed = False - - inames = [] - if isinstance(instance_names, list): - inames = instance_names - elif isinstance(instance_names, str): - inames = instance_names.split(',') - if name: - inames.append(name) - if not inames: - module.fail_json(msg='Must specify a "name" or "instance_names"', - changed=False) - if not zone: - module.fail_json(msg='Must specify a "zone"', changed=False) - - json_output = {'zone': zone} - if state in ['absent', 'deleted']: - json_output['state'] = 'absent' - (changed, terminated_instance_names) = terminate_instances( - module, gce, inames, zone) - - # based on what user specified, return the same variable, although - # value could be different if an instance could not be destroyed - if instance_names: - json_output['instance_names'] = terminated_instance_names - elif name: - json_output['name'] = name - - elif state in ['active', 'present']: - json_output['state'] = 'present' - (changed, instance_data, instance_name_list) = create_instances( - module, gce, inames) - json_output['instance_data'] = instance_data - if instance_names: - json_output['instance_names'] = instance_name_list - elif name: - json_output['name'] = name - - json_output['changed'] = changed - module.exit_json(**json_output) - -# import module snippets -from ansible.module_utils.basic import * -from ansible.module_utils.gce import * -if __name__ == '__main__': - main() diff --git a/playbooks/gce/openshift-cluster/list.yml b/playbooks/gce/openshift-cluster/list.yml index 34dcd2496..34ab09533 100644 --- a/playbooks/gce/openshift-cluster/list.yml +++ b/playbooks/gce/openshift-cluster/list.yml @@ -16,18 +16,8 @@ groups: oo_list_hosts ansible_ssh_user: "{{ deployment_vars[deployment_type].ssh_user }}" ansible_become: "{{ deployment_vars[deployment_type].become }}" + oo_public_ipv4: "{{ hostvars[item].gce_public_ip }}" + oo_private_ipv4: "{{ hostvars[item].gce_private_ip }}" with_items: "{{ groups[scratch_group] | default([], true) | difference(['localhost']) | difference(groups.status_terminated | default([], true)) }}" - -- name: List Hosts - hosts: oo_list_hosts - -- name: List Hosts - hosts: localhost - become: no - connection: local - gather_facts: no - vars_files: - - vars.yml - tasks: - debug: msg: "{{ hostvars | oo_select_keys(groups[scratch_group] | default([])) | oo_pretty_print_cluster }}" diff --git a/playbooks/gce/openshift-cluster/tasks/launch_instances.yml b/playbooks/gce/openshift-cluster/tasks/launch_instances.yml index 7c8189224..b7604580c 100644 --- a/playbooks/gce/openshift-cluster/tasks/launch_instances.yml +++ b/playbooks/gce/openshift-cluster/tasks/launch_instances.yml @@ -9,6 +9,7 @@ project_id: "{{ lookup('env', 'gce_project_id') }}" zone: "{{ lookup('env', 'zone') }}" network: "{{ lookup('env', 'network') }}" + subnetwork: "{{ lookup('env', 'subnetwork') | default(omit, True) }}" # unsupported in 1.9.+ #service_account_permissions: "datastore,logging-write" tags: diff --git a/playbooks/libvirt/openshift-cluster/list.yml b/playbooks/libvirt/openshift-cluster/list.yml index 86d5d0aad..579cd7ac6 100644 --- a/playbooks/libvirt/openshift-cluster/list.yml +++ b/playbooks/libvirt/openshift-cluster/list.yml @@ -16,18 +16,8 @@ groups: oo_list_hosts ansible_ssh_user: "{{ deployment_vars[deployment_type].ssh_user }}" ansible_become: "{{ deployment_vars[deployment_type].become }}" + oo_public_ipv4: "" + oo_private_ipv4: "{{ hostvars[item].libvirt_ip_address }}" with_items: "{{ groups[scratch_group] | default([]) | difference(['localhost']) }}" - -- name: List Hosts - hosts: oo_list_hosts - -- name: List Hosts - hosts: localhost - become: no - connection: local - gather_facts: no - vars_files: - - vars.yml - tasks: - debug: msg: "{{ hostvars | oo_select_keys(groups[scratch_group] | default([])) | oo_pretty_print_cluster }}" diff --git a/playbooks/openstack/openshift-cluster/list.yml b/playbooks/openstack/openshift-cluster/list.yml index de68f5207..6c6f671be 100644 --- a/playbooks/openstack/openshift-cluster/list.yml +++ b/playbooks/openstack/openshift-cluster/list.yml @@ -17,18 +17,8 @@ ansible_ssh_user: "{{ deployment_vars[deployment_type].ssh_user }}" ansible_ssh_host: "{{ hostvars[item].ansible_ssh_host | default(item) }}" ansible_become: "{{ deployment_vars[deployment_type].become }}" + oo_public_ipv4: "{{ hostvars[item].openstack.public_v4 }}" + oo_private_ipv4: "{{ hostvars[item].openstack.private_v4 }}" with_items: "{{ groups[scratch_group] | default([]) | difference(['localhost']) }}" - -- name: List Hosts - hosts: oo_list_hosts - -- name: List Hosts - hosts: localhost - become: no - connection: local - gather_facts: no - vars_files: - - vars.yml - tasks: - debug: msg: "{{ hostvars | oo_select_keys(groups[scratch_group] | default([])) | oo_pretty_print_cluster('meta-') }}" diff --git a/roles/docker/README.md b/roles/docker/README.md index 6b5ee4421..1f0d94da0 100644 --- a/roles/docker/README.md +++ b/roles/docker/README.md @@ -17,7 +17,7 @@ docker_udev_workaround: raises udevd timeout to 5 minutes (https://bugzilla.redh Dependencies ------------ -None +Depends on the os_firewall role. Example Playbook ---------------- diff --git a/roles/docker/meta/main.yml b/roles/docker/meta/main.yml index 6e2c98601..3d362158d 100644 --- a/roles/docker/meta/main.yml +++ b/roles/docker/meta/main.yml @@ -9,4 +9,6 @@ galaxy_info: - name: EL versions: - 7 -dependencies: [] +dependencies: + - role: os_firewall + os_firewall_use_firewalld: False diff --git a/roles/docker/tasks/main.yml b/roles/docker/tasks/main.yml index 240673b22..9b7ef0830 100644 --- a/roles/docker/tasks/main.yml +++ b/roles/docker/tasks/main.yml @@ -1,7 +1,4 @@ --- -- stat: path=/etc/sysconfig/docker-storage - register: docker_storage_check - - name: Get current installed Docker version command: "{{ repoquery_cmd }} --installed --qf '%{version}' docker" when: not openshift.common.is_atomic | bool @@ -57,17 +54,6 @@ dest: "{{ docker_systemd_dir }}/custom.conf" src: custom.conf.j2 -- name: Start the Docker service - systemd: - name: docker - enabled: yes - state: started - daemon_reload: yes - register: start_result - -- set_fact: - docker_service_status_changed: start_result | changed - - include: udev_workaround.yml when: docker_udev_workaround | default(False) | bool @@ -125,4 +111,15 @@ notify: - restart docker +- name: Start the Docker service + systemd: + name: docker + enabled: yes + state: started + daemon_reload: yes + register: start_result + +- set_fact: + docker_service_status_changed: start_result | changed + - meta: flush_handlers diff --git a/roles/etcd/etcdctl.sh b/roles/etcd/etcdctl.sh new file mode 100644 index 000000000..0e324a8a9 --- /dev/null +++ b/roles/etcd/etcdctl.sh @@ -0,0 +1,11 @@ +#!/bin/bash +# Sets up handy aliases for etcd, need etcdctl2 and etcdctl3 because +# command flags are different between the two. Should work on stand +# alone etcd hosts and master + etcd hosts too because we use the peer keys. +etcdctl2() { + /usr/bin/etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://`hostname`:2379 ${@} +} + +etcdctl3() { + ETCDCTL_API=3 /usr/bin/etcdctl --cert /etc/etcd/peer.crt --key /etc/etcd/peer.key --cacert /etc/etcd/ca.crt --endpoints https://`hostname`:2379 ${@} +} diff --git a/roles/etcd/files/etcdctl.sh b/roles/etcd/files/etcdctl.sh new file mode 100644 index 000000000..0e324a8a9 --- /dev/null +++ b/roles/etcd/files/etcdctl.sh @@ -0,0 +1,11 @@ +#!/bin/bash +# Sets up handy aliases for etcd, need etcdctl2 and etcdctl3 because +# command flags are different between the two. Should work on stand +# alone etcd hosts and master + etcd hosts too because we use the peer keys. +etcdctl2() { + /usr/bin/etcdctl --cert-file /etc/etcd/peer.crt --key-file /etc/etcd/peer.key --ca-file /etc/etcd/ca.crt -C https://`hostname`:2379 ${@} +} + +etcdctl3() { + ETCDCTL_API=3 /usr/bin/etcdctl --cert /etc/etcd/peer.crt --key /etc/etcd/peer.key --cacert /etc/etcd/ca.crt --endpoints https://`hostname`:2379 ${@} +} diff --git a/roles/etcd/tasks/etcdctl.yml b/roles/etcd/tasks/etcdctl.yml new file mode 100644 index 000000000..32c176449 --- /dev/null +++ b/roles/etcd/tasks/etcdctl.yml @@ -0,0 +1,11 @@ +- name: Install etcd for etcdctl + action: "{{ ansible_pkg_mgr }} name=etcd state=present" + when: not openshift.common.is_atomic | bool + +- name: Configure etcd profile.d alises + copy: + src: etcdctl.sh + dest: /etc/profile.d/etcdctl.sh + mode: 0755 + owner: root + group: root diff --git a/roles/etcd/tasks/main.yml b/roles/etcd/tasks/main.yml index 2bc6a8678..790eb3c5a 100644 --- a/roles/etcd/tasks/main.yml +++ b/roles/etcd/tasks/main.yml @@ -74,5 +74,8 @@ enabled: yes register: start_result +- include: etcdctl.yml + when: openshift_etcd_etcdctl_profile | default(true) | bool + - set_fact: etcd_service_status_changed: "{{ start_result | changed }}" diff --git a/roles/openshift_examples/files/examples/v1.4/image-streams/dotnet_imagestreams.json b/roles/openshift_examples/files/examples/v1.4/image-streams/dotnet_imagestreams.json index 00b363c42..a65d35c2e 100644 --- a/roles/openshift_examples/files/examples/v1.4/image-streams/dotnet_imagestreams.json +++ b/roles/openshift_examples/files/examples/v1.4/image-streams/dotnet_imagestreams.json @@ -12,14 +12,18 @@ "kind": "ImageStream", "apiVersion": "v1", "metadata": { - "name": "dotnet" + "name": "dotnet", + "annotations": { + "openshift.io/display-name": ".NET Core" + } }, "spec": { "tags": [ { "name": "latest", "annotations": { - "description": "Build and run .NET Core 1.0 applications", + "openshift.io/display-name": ".NET Core (Latest)", + "description": "Build and run .NET Core applications on RHEL 7. For more information about using this builder image, including OpenShift considerations, see https://github.com/redhat-developer/s2i-dotnetcore/tree/master/1.0/README.md.\n\nWARNING: By selecting this tag, your application will automatically update to use the latest version of .NET Core available on OpenShift, including major versions updates.", "iconClass": "icon-dotnet", "tags": "builder,.net,dotnet,dotnetcore", "supports":"dotnet", @@ -34,7 +38,8 @@ { "name": "1.0", "annotations": { - "description": "Build and run .NET Core 1.0 applications", + "openshift.io/display-name": ".NET Core 1.0", + "description": "Build and run .NET Core 1.0 applications on RHEL 7. For more information about using this builder image, including OpenShift considerations, see https://github.com/redhat-developer/s2i-dotnetcore/tree/master/1.0/README.md.", "iconClass": "icon-dotnet", "tags": "builder,.net,dotnet,dotnetcore,rh-dotnetcore10", "supports":"dotnet:1.0,dotnet", diff --git a/roles/openshift_facts/library/openshift_facts.py b/roles/openshift_facts/library/openshift_facts.py index a28b58e85..d797eb4d3 100755 --- a/roles/openshift_facts/library/openshift_facts.py +++ b/roles/openshift_facts/library/openshift_facts.py @@ -1595,7 +1595,7 @@ def set_container_facts_if_unset(facts): cli_image = master_image node_image = 'openshift3/node' ovs_image = 'openshift3/openvswitch' - etcd_image = 'registry.access.redhat.com/rhel7/etcd' + etcd_image = 'registry.access.redhat.com/rhel7/etcd3' pod_image = 'openshift3/ose-pod' router_image = 'openshift3/ose-haproxy-router' registry_image = 'openshift3/ose-docker-registry' @@ -1605,7 +1605,7 @@ def set_container_facts_if_unset(facts): cli_image = master_image node_image = 'aep3_beta/node' ovs_image = 'aep3_beta/openvswitch' - etcd_image = 'registry.access.redhat.com/rhel7/etcd' + etcd_image = 'registry.access.redhat.com/rhel7/etcd3' pod_image = 'aep3_beta/aep-pod' router_image = 'aep3_beta/aep-haproxy-router' registry_image = 'aep3_beta/aep-docker-registry' @@ -1615,7 +1615,7 @@ def set_container_facts_if_unset(facts): cli_image = master_image node_image = 'openshift/node' ovs_image = 'openshift/openvswitch' - etcd_image = 'registry.access.redhat.com/rhel7/etcd' + etcd_image = 'registry.access.redhat.com/rhel7/etcd3' pod_image = 'openshift/origin-pod' router_image = 'openshift/origin-haproxy-router' registry_image = 'openshift/origin-docker-registry' diff --git a/roles/openshift_facts/tasks/main.yml b/roles/openshift_facts/tasks/main.yml index b0785a9e4..4d4a232cc 100644 --- a/roles/openshift_facts/tasks/main.yml +++ b/roles/openshift_facts/tasks/main.yml @@ -38,6 +38,8 @@ 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: Set repoquery command set_fact: diff --git a/roles/openshift_hosted_logging/tasks/deploy_logging.yaml b/roles/openshift_hosted_logging/tasks/deploy_logging.yaml index 0162d1fb0..625af9acd 100644 --- a/roles/openshift_hosted_logging/tasks/deploy_logging.yaml +++ b/roles/openshift_hosted_logging/tasks/deploy_logging.yaml @@ -40,7 +40,7 @@ - name: "Create templates for logging accounts and the deployer" command: > - {{ openshift.common.client_binary }} create + {{ openshift.common.client_binary }} create --config={{ mktemp.stdout }}/admin.kubeconfig -f {{ hosted_base }}/logging-deployer.yaml --config={{ mktemp.stdout }}/admin.kubeconfig -n logging @@ -49,28 +49,41 @@ changed_when: "'created' in logging_import_template.stdout" - name: "Process the logging accounts template" - shell: "{{ openshift.common.client_binary }} --config={{ mktemp.stdout }}/admin.kubeconfig process logging-deployer-account-template | {{ openshift.common.client_binary }} --config={{ mktemp.stdout }}/admin.kubeconfig create -f -" + shell: > + {{ openshift.common.client_binary }} --config={{ mktemp.stdout }}/admin.kubeconfig + process logging-deployer-account-template | {{ openshift.common.client_binary }} --config={{ mktemp.stdout }}/admin.kubeconfig create -f - register: process_deployer_accounts failed_when: process_deployer_accounts.rc == 1 and 'already exists' not in process_deployer_accounts.stderr - name: "Set permissions for logging-deployer service account" command: > - {{ openshift.common.client_binary }} adm --config={{ mktemp.stdout }}/admin.kubeconfig policy add-cluster-role-to-user oauth-editor system:serviceaccount:logging:logging-deployer + {{ openshift.common.client_binary }} adm --config={{ mktemp.stdout }}/admin.kubeconfig + policy add-cluster-role-to-user oauth-editor system:serviceaccount:logging:logging-deployer register: permiss_output failed_when: "permiss_output.rc == 1 and 'exists' not in permiss_output.stderr" - name: "Set permissions for fluentd" command: > - {{ openshift.common.client_binary }} adm policy add-scc-to-user privileged system:serviceaccount:logging:aggregated-logging-fluentd + {{ openshift.common.client_binary }} adm --config={{ mktemp.stdout }}/admin.kubeconfig + policy add-scc-to-user privileged system:serviceaccount:logging:aggregated-logging-fluentd register: fluentd_output failed_when: "fluentd_output.rc == 1 and 'exists' not in fluentd_output.stderr" - name: "Set additional permissions for fluentd" command: > - {{ openshift.common.client_binary }} adm policy add-cluster-role-to-user cluster-reader system:serviceaccount:logging:aggregated-logging-fluentd + {{ openshift.common.client_binary }} adm policy --config={{ mktemp.stdout }}/admin.kubeconfig + add-cluster-role-to-user cluster-reader system:serviceaccount:logging:aggregated-logging-fluentd register: fluentd2_output failed_when: "fluentd2_output.rc == 1 and 'exists' not in fluentd2_output.stderr" + - name: "Add rolebinding-reader to aggregated-logging-elasticsearch" + command: > + {{ openshift.common.client_binary }} adm --config={{ mktemp.stdout }}/admin.kubeconfig + policy add-cluster-role-to-user rolebinding-reader \ + system:serviceaccount:logging:aggregated-logging-elasticsearch + register: rolebinding_reader_output + failed_when: "rolebinding_reader_output == 1 and 'exists' not in rolebinding_reader_output.stderr" + - name: "Create ConfigMap for deployer parameters" command: > {{ openshift.common.client_binary}} --config={{ mktemp.stdout }}/admin.kubeconfig create configmap logging-deployer {{ deployer_cmap_params }} diff --git a/roles/openshift_hosted_templates/files/v1.4/enterprise/logging-deployer.yaml b/roles/openshift_hosted_templates/files/v1.4/enterprise/logging-deployer.yaml index 9cff9daca..ddfda1272 100644 --- a/roles/openshift_hosted_templates/files/v1.4/enterprise/logging-deployer.yaml +++ b/roles/openshift_hosted_templates/files/v1.4/enterprise/logging-deployer.yaml @@ -66,6 +66,15 @@ items: - watch - delete - update + - apiVersion: v1 + kind: ClusterRole + metadata: + name: rolebinding-reader + rules: + - resources: + - clusterrolebindings + verbs: + - get - apiVersion: v1 kind: RoleBinding @@ -88,6 +97,17 @@ items: subjects: - kind: ServiceAccount name: logging-deployer + - + apiVersion: v1 + kind: RoleBinding + metadata: + name: logging-elasticsearch-view-role + roleRef: + kind: ClusterRole + name: view + subjects: + - kind: ServiceAccount + name: aggregated-logging-elasticsearch - apiVersion: "v1" kind: "Template" diff --git a/roles/openshift_hosted_templates/files/v1.4/enterprise/metrics-deployer.yaml b/roles/openshift_hosted_templates/files/v1.4/enterprise/metrics-deployer.yaml index 97a86c11d..66051755c 100644 --- a/roles/openshift_hosted_templates/files/v1.4/enterprise/metrics-deployer.yaml +++ b/roles/openshift_hosted_templates/files/v1.4/enterprise/metrics-deployer.yaml @@ -86,6 +86,8 @@ objects: value: ${HEAPSTER_NODE_ID} - name: METRIC_RESOLUTION value: ${METRIC_RESOLUTION} + - name: STARTUP_TIMEOUT + value: ${STARTUP_TIMEOUT} dnsPolicy: ClusterFirst restartPolicy: Never serviceAccount: metrics-deployer @@ -160,3 +162,7 @@ parameters: description: "How often metrics should be gathered. Defaults value of '15s' for 15 seconds" name: METRIC_RESOLUTION value: "15s" +- + description: "How long in seconds we should wait until Hawkular Metrics and Heapster starts up before attempting a restart" + name: STARTUP_TIMEOUT + value: "500" diff --git a/roles/openshift_hosted_templates/files/v1.4/origin/metrics-deployer.yaml b/roles/openshift_hosted_templates/files/v1.4/origin/metrics-deployer.yaml index 5f2290419..54691572a 100644 --- a/roles/openshift_hosted_templates/files/v1.4/origin/metrics-deployer.yaml +++ b/roles/openshift_hosted_templates/files/v1.4/origin/metrics-deployer.yaml @@ -86,6 +86,8 @@ objects: value: ${HEAPSTER_NODE_ID} - name: METRIC_RESOLUTION value: ${METRIC_RESOLUTION} + - name: STARTUP_TIMEOUT + value: ${STARTUP_TIMEOUT} dnsPolicy: ClusterFirst restartPolicy: Never serviceAccount: metrics-deployer @@ -160,3 +162,7 @@ parameters: description: "How often metrics should be gathered. Defaults value of '15s' for 15 seconds" name: METRIC_RESOLUTION value: "15s" +- + description: "How long in seconds we should wait until Hawkular Metrics and Heapster starts up before attempting a restart" + name: STARTUP_TIMEOUT + value: "500" diff --git a/roles/openshift_master/templates/master.yaml.v1.j2 b/roles/openshift_master/templates/master.yaml.v1.j2 index 4d45e8591..169d7f2e3 100644 --- a/roles/openshift_master/templates/master.yaml.v1.j2 +++ b/roles/openshift_master/templates/master.yaml.v1.j2 @@ -158,7 +158,7 @@ masterPublicURL: {{ openshift.master.public_api_url }} networkConfig: clusterNetworkCIDR: {{ openshift.master.sdn_cluster_network_cidr }} hostSubnetLength: {{ openshift.master.sdn_host_subnet_length }} -{% if openshift.common.use_openshift_sdn or openshift.common.use_nuage %} +{% if openshift.common.use_openshift_sdn or openshift.common.use_nuage or openshift.common.sdn_network_plugin_name == 'cni' %} networkPluginName: {{ openshift.common.sdn_network_plugin_name }} {% endif %} # serviceNetworkCIDR must match kubernetesMasterConfig.servicesSubnet diff --git a/roles/openshift_node/templates/node.yaml.v1.j2 b/roles/openshift_node/templates/node.yaml.v1.j2 index 9bcaf4d84..55ae4bf54 100644 --- a/roles/openshift_node/templates/node.yaml.v1.j2 +++ b/roles/openshift_node/templates/node.yaml.v1.j2 @@ -27,7 +27,7 @@ networkPluginName: {{ openshift.common.sdn_network_plugin_name }} # deprecates networkPluginName above. The two should match. networkConfig: mtu: {{ openshift.node.sdn_mtu }} -{% if openshift.common.use_openshift_sdn | bool or openshift.common.use_nuage | bool %} +{% if openshift.common.use_openshift_sdn | bool or openshift.common.use_nuage | bool or openshift.common.sdn_network_plugin_name == 'cni' %} networkPluginName: {{ openshift.common.sdn_network_plugin_name }} {% endif %} {% if openshift.node.set_node_ip | bool %} diff --git a/roles/openshift_version/tasks/set_version_containerized.yml b/roles/openshift_version/tasks/set_version_containerized.yml index 8e2702391..718537287 100644 --- a/roles/openshift_version/tasks/set_version_containerized.yml +++ b/roles/openshift_version/tasks/set_version_containerized.yml @@ -37,3 +37,8 @@ openshift_version: "{{ cli_image_version.stdout_lines[0].split(' ')[1].split('-')[0:2][1:] | join('-') if openshift.common.deployment_type == 'origin' else cli_image_version.stdout_lines[0].split(' ')[1].split('-')[0][1:] }}" when: openshift_version is defined and openshift_version.split('.') | length == 2 +# We finally have the specific version. Now we clean up any strange +# dangly +c0mm1t-offset tags in the version. See also, +# openshift_facts.py +- set_fact: + openshift_version: "{{ openshift_version | oo_chomp_commit_offset }}" diff --git a/roles/os_firewall/README.md b/roles/os_firewall/README.md index 187d74b06..c6c70b81d 100644 --- a/roles/os_firewall/README.md +++ b/roles/os_firewall/README.md @@ -14,7 +14,7 @@ Role Variables | Name | Default | | |---------------------------|---------|----------------------------------------| -| os_firewall_use_firewalld | True | If false, use iptables | +| os_firewall_use_firewalld | False | If false, use iptables | | os_firewall_allow | [] | List of service,port mappings to allow | | os_firewall_deny | [] | List of service, port mappings to deny | diff --git a/roles/os_firewall/meta/main.yml b/roles/os_firewall/meta/main.yml index c93335b7b..6df7c9f2b 100644 --- a/roles/os_firewall/meta/main.yml +++ b/roles/os_firewall/meta/main.yml @@ -11,5 +11,6 @@ galaxy_info: - 7 categories: - system +allow_duplicates: yes dependencies: - { role: openshift_facts } |