C12s/c12s-kubespray
3
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
c12s-kubespray/upgrade-cluster.yml

165 lines
6.1 KiB
YAML
Raw Normal View History

first cut of an upgrade process
2016-12-09 18:44:52 +00:00
---
bump minimal ansible version to 2.8.0 (#5984) * bump minimal ansible version to 2.8.0 * check ansible version in separate playbook
2020-04-22 20:33:44 +00:00
- name: Check ansible version
import_playbook: ansible_version.yml
Assert if ansible 2.7 is used (#3589)
2018-10-26 07:29:21 +00:00
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
- name: Ensure compatibility with old groups
import_playbook: legacy_groups.yml
Replace kube-master with kube_control_plane (#7256) This replaces kube-master with kube_control_plane because of [1]: The Kubernetes project is moving away from wording that is considered offensive. A new working group WG Naming was created to track this work, and the word "master" was declared as offensive. A proposal was formalized for replacing the word "master" with "control plane". This means it should be removed from source code, documentation, and user-facing configuration from Kubernetes and its sub-projects. NOTE: The reason why this changes it to kube_control_plane not kube-control-plane is for valid group names on ansible. [1]: https://github.com/kubernetes/enhancements/blob/master/keps/sig-cluster-lifecycle/kubeadm/2067-rename-master-label-taint/README.md#motivation
2021-03-24 00:26:05 +00:00
Split deploy steps in separate playbooks: part1 (#3451) * Fix bootstrap_os/ubuntu idempotency * Update bastion role * move container_engine in sub-roles * requires ansible 2.5 * ubuntu18 as first CI job
2018-10-10 02:14:33 +00:00
- hosts: bastion[0]
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
gather_facts: False
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
- { role: bastion-ssh-config, tags: ["localhost", "bastion"] }
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
- hosts: k8s_cluster:etcd:calico_rr
add strategy mitogen_linear when installed mitogen (#5985) * add strategy mitogen_linear when installed mitogen * add small docs Rename playbook file The raw action executes as a regular Mitogen connection, which requires Python on the target, so add strategy: linear to bootstrap-os role playbook. * add mitogen to CI test fix typo * enable mitogen test on deploy-part1 tests change version from master to release download tar.gz archive * run all CI tests with mitogen * disable mitogen with upgrade CI tests * enable mitogen on CI tests via env vars * disable mitogen on CI test by default, enable on some different OS * disable mitogen CI test on centos8 (get error /usr/bin/python: No such file or directory)
2020-04-24 12:20:07 +00:00
strategy: linear
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
gather_facts: false
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
vars:
# Need to disable pipelining for bootstrap-os as some systems have requiretty in sudoers set, which makes pipelining
# fail. bootstrap-os fixes this on these systems, so in later plays it can be enabled.
ansible_ssh_pipelining: false
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Add default var role
2017-02-23 11:07:17 +00:00
- { role: bootstrap-os, tags: bootstrap-os}
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
- name: Gather facts
fix use of ansible tags (#6316) tags are not inherited for include_role therefore the change from include to import Co-authored-by: Hans Feldt <hafe@users.noreply.github.com>
2020-06-25 10:00:37 +00:00
tags: always
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
import_playbook: facts.yml
Replace kube-master with kube_control_plane (#7256) This replaces kube-master with kube_control_plane because of [1]: The Kubernetes project is moving away from wording that is considered offensive. A new working group WG Naming was created to track this work, and the word "master" was declared as offensive. A proposal was formalized for replacing the word "master" with "control plane". This means it should be removed from source code, documentation, and user-facing configuration from Kubernetes and its sub-projects. NOTE: The reason why this changes it to kube_control_plane not kube-control-plane is for valid group names on ansible. [1]: https://github.com/kubernetes/enhancements/blob/master/keps/sig-cluster-lifecycle/kubeadm/2067-rename-master-label-taint/README.md#motivation
2021-03-24 00:26:05 +00:00
- name: Download images to ansible host cache via first kube_control_plane node
hosts: kube_control_plane[0]
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Refactor download role (#5697) * download file * download containers * fix push image to nodes * pull if none image on host * fix * improve docker image tag checks. do not pull already cached images * rebase fix merge conflict * add support download_run_once when upgrade and scale cluster add some test with download_run_once * set default values to temp flag for every download cycle * add save,load abilty for containerd and crio when download_run_once=true * return redefine image save/load command to set_docker_image_facts.yml * move set command to set_container_facts * ctr in containerd_bin_dir * fix order of ctr image export arguments * temporary disable download_run_once for containerd and crio due https://github.com/containerd/containerd/issues/4075 * remove unused files * fix strict yaml linter warning and errors * refactor logical conditions to pull and cache container images * remove comment due lint check * document role * remove image_load_on_localhost, because cached images are always loaded to docker on remote sites * remove XXX from debug output
2020-03-05 15:31:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Refactor download role (#5697) * download file * download containers * fix push image to nodes * pull if none image on host * fix * improve docker image tag checks. do not pull already cached images * rebase fix merge conflict * add support download_run_once when upgrade and scale cluster add some test with download_run_once * set default values to temp flag for every download cycle * add save,load abilty for containerd and crio when download_run_once=true * return redefine image save/load command to set_docker_image_facts.yml * move set command to set_container_facts * ctr in containerd_bin_dir * fix order of ctr image export arguments * temporary disable download_run_once for containerd and crio due https://github.com/containerd/containerd/issues/4075 * remove unused files * fix strict yaml linter warning and errors * refactor logical conditions to pull and cache container images * remove comment due lint check * document role * remove image_load_on_localhost, because cached images are always loaded to docker on remote sites * remove XXX from debug output
2020-03-05 15:31:39 +00:00
roles:
- { role: kubespray-defaults, when: "not skip_downloads and download_run_once and not download_localhost"}
- { role: kubernetes/preinstall, tags: preinstall, when: "not skip_downloads and download_run_once and not download_localhost" }
- { role: download, tags: download, when: "not skip_downloads and download_run_once and not download_localhost" }
Upgrade container-engine after draining (#5601) * Run 'container-engine' after drain. Move possibly disruptive role 'container-engine' to run after the node is drained. As that role have to be run on non-cluster nodes as well (etcd and calico-rr), and those nodes are not drained, add play for that case. * Check if api is up before upgrade. If container engine is restarted in previous role, api controller can take some time to start. This check ensures api is up before upgrade.
2020-02-27 19:47:28 +00:00
- name: Prepare nodes for upgrade
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
hosts: k8s_cluster:etcd:calico_rr
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
- { role: kubernetes/preinstall, tags: preinstall }
Fix skip_downloads condition.
2018-08-08 10:49:49 +00:00
- { role: download, tags: download, when: "not skip_downloads" }
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
Upgrade container-engine after draining (#5601) * Run 'container-engine' after drain. Move possibly disruptive role 'container-engine' to run after the node is drained. As that role have to be run on non-cluster nodes as well (etcd and calico-rr), and those nodes are not drained, add play for that case. * Check if api is up before upgrade. If container engine is restarted in previous role, api controller can take some time to start. This check ensures api is up before upgrade.
2020-02-27 19:47:28 +00:00
- name: Upgrade container engine on non-cluster nodes
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
hosts: etcd:calico_rr:!k8s_cluster
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Upgrade container-engine after draining (#5601) * Run 'container-engine' after drain. Move possibly disruptive role 'container-engine' to run after the node is drained. As that role have to be run on non-cluster nodes as well (etcd and calico-rr), and those nodes are not drained, add play for that case. * Check if api is up before upgrade. If container engine is restarted in previous role, api controller can take some time to start. This check ensures api is up before upgrade.
2020-02-27 19:47:28 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Upgrade container-engine after draining (#5601) * Run 'container-engine' after drain. Move possibly disruptive role 'container-engine' to run after the node is drained. As that role have to be run on non-cluster nodes as well (etcd and calico-rr), and those nodes are not drained, add play for that case. * Check if api is up before upgrade. If container engine is restarted in previous role, api controller can take some time to start. This check ensures api is up before upgrade.
2020-02-27 19:47:28 +00:00
serial: "{{ serial | default('20%') }}"
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
etcd: deploy container engine if needed (#7532) If the etcd cluster is separate and the etcd_deployment_type is "host", there is no need for a container engine on the etcd nodes Do not rely on a 'default(true)' filter, but define a proper default in kubespray-defaults depending on etcd deployment method and if internal or external etcd is used
2021-10-12 07:31:47 +00:00
- { role: container-engine, tags: "container-engine", when: deploy_container_engine }
Upgrade container-engine after draining (#5601) * Run 'container-engine' after drain. Move possibly disruptive role 'container-engine' to run after the node is drained. As that role have to be run on non-cluster nodes as well (etcd and calico-rr), and those nodes are not drained, add play for that case. * Check if api is up before upgrade. If container engine is restarted in previous role, api controller can take some time to start. This check ensures api is up before upgrade.
2020-02-27 19:47:28 +00:00
Remove standalone etcd specific play, cleanup host mode Now etcd role can optionally disable etcd cluster setup for faster deployment when it is combined with etcd role.
2017-03-03 20:30:37 +00:00
- hosts: etcd
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Enable kubeadm etcd mode (#4818) * Enable kubeadm etcd mode Uses cert commands from kubeadm experimental control plane to enable non-master nodes to obtain etcd certs. Related story: PROD-29434 Change-Id: Idafa1d223e5c6ceadf819b6f9c06adf4c4f74178 * Add validation checks and exclude calico kdd mode Change-Id: Ic234f5e71261d33191376e70d438f9f6d35f358c * Move etcd mode test to ubuntu flannel HA job Change-Id: I9af6fd80a1bbb1692ab10d6da095eb368f6bc732 * rename etcd_mode to etcd_kubeadm_enabled Change-Id: Ib196d6c8a52f48cae370b026f7687ff9ca69c172
2019-06-20 18:12:51 +00:00
- role: etcd
tags: etcd
vars:
etcd_cluster_setup: true
Fix etcd_events not getting upgraded in upgrade-cluster.yml (#8550) Signed-off-by: Mac Chaffee <me@macchaffee.com>
2022-02-17 16:03:38 +00:00
etcd_events_cluster_setup: "{{ etcd_events_cluster_enabled }}"
Enable kubeadm etcd mode (#4818) * Enable kubeadm etcd mode Uses cert commands from kubeadm experimental control plane to enable non-master nodes to obtain etcd certs. Related story: PROD-29434 Change-Id: Idafa1d223e5c6ceadf819b6f9c06adf4c4f74178 * Add validation checks and exclude calico kdd mode Change-Id: Ic234f5e71261d33191376e70d438f9f6d35f358c * Move etcd mode test to ubuntu flannel HA job Change-Id: I9af6fd80a1bbb1692ab10d6da095eb368f6bc732 * rename etcd_mode to etcd_kubeadm_enabled Change-Id: Ib196d6c8a52f48cae370b026f7687ff9ca69c172
2019-06-20 18:12:51 +00:00
when: not etcd_kubeadm_enabled | default(false)
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
- hosts: k8s_cluster
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Enable kubeadm etcd mode (#4818) * Enable kubeadm etcd mode Uses cert commands from kubeadm experimental control plane to enable non-master nodes to obtain etcd certs. Related story: PROD-29434 Change-Id: Idafa1d223e5c6ceadf819b6f9c06adf4c4f74178 * Add validation checks and exclude calico kdd mode Change-Id: Ic234f5e71261d33191376e70d438f9f6d35f358c * Move etcd mode test to ubuntu flannel HA job Change-Id: I9af6fd80a1bbb1692ab10d6da095eb368f6bc732 * rename etcd_mode to etcd_kubeadm_enabled Change-Id: Ib196d6c8a52f48cae370b026f7687ff9ca69c172
2019-06-20 18:12:51 +00:00
- role: etcd
tags: etcd
vars:
etcd_cluster_setup: false
etcd_events_cluster_setup: false
when: not etcd_kubeadm_enabled | default(false)
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
Yamllint fixes (#4410) * Lint everything in the repository with yamllint * yamllint fixes: syntax fixes only * yamllint fixes: move comments to play names * yamllint fixes: indent comments in .gitlab-ci.yml file
2019-04-01 09:38:33 +00:00
- name: Handle upgrades to master components first to maintain backwards compat.
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Replace kube-master with kube_control_plane (#7256) This replaces kube-master with kube_control_plane because of [1]: The Kubernetes project is moving away from wording that is considered offensive. A new working group WG Naming was created to track this work, and the word "master" was declared as offensive. A proposal was formalized for replacing the word "master" with "control plane". This means it should be removed from source code, documentation, and user-facing configuration from Kubernetes and its sub-projects. NOTE: The reason why this changes it to kube_control_plane not kube-control-plane is for valid group names on ansible. [1]: https://github.com/kubernetes/enhancements/blob/master/keps/sig-cluster-lifecycle/kubeadm/2067-rename-master-label-taint/README.md#motivation
2021-03-24 00:26:05 +00:00
hosts: kube_control_plane
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
serial: 1
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
- { role: upgrade/pre-upgrade, tags: pre-upgrade }
etcd: deploy container engine if needed (#7532) If the etcd cluster is separate and the etcd_deployment_type is "host", there is no need for a container engine on the etcd nodes Do not rely on a 'default(true)' filter, but define a proper default in kubespray-defaults depending on etcd deployment method and if internal or external etcd is used
2021-10-12 07:31:47 +00:00
- { role: container-engine, tags: "container-engine", when: deploy_container_engine }
first cut of an upgrade process
2016-12-09 18:44:52 +00:00
- { role: kubernetes/node, tags: node }
Move kubernetes/master to kubernetes/control-plane (#7218) This is a small step to replace "master" with "control-plane" in Kubespray project.
2021-02-01 15:15:49 +00:00
- { role: kubernetes/control-plane, tags: master, upgrade_cluster_setup: true }
Move cluster roles and system namespace to new role This should be done after kubeconfig is set for admin and before network plugins are up.
2017-10-26 08:10:33 +00:00
- { role: kubernetes/client, tags: client }
tag role kubernetes/node-label in playbooks (#5560)
2020-01-20 19:43:36 +00:00
- { role: kubernetes/node-label, tags: node-label }
Move cluster roles and system namespace to new role This should be done after kubeconfig is set for admin and before network plugins are up.
2017-10-26 08:10:33 +00:00
- { role: kubernetes-apps/cluster_roles, tags: cluster-roles }
install csi-driver not just cinder (#5766)
2020-03-13 12:34:39 +00:00
- { role: kubernetes-apps, tags: csi-driver }
Uncodron Masters which have scheduling Enabled
2017-06-27 20:11:44 +00:00
- { role: upgrade/post-upgrade, tags: post-upgrade }
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
Handle calico-rr nodes as workers so they get upgraded too (#6447) * Handle calico-rr nodes as workers so they get upgraded too * calico-rr nodes run 'calico and external cloud provider' too
2020-09-24 11:38:05 +00:00
- name: Upgrade calico and external cloud provider on all masters, calico-rrs, and nodes
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
hosts: kube_control_plane:calico_rr:kube_node
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
calico upgrade to v3 (#3086) * calico upgrade to v3 * update calico_rr version * add missing file * change contents of main.yml as it was left old version * enable network policy by default * remove unneeded task * Fix kubelet calico settings * fix when statement * switch back to node-kubeconfig.yaml
2018-08-23 14:17:18 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
serial: "{{ serial | default('20%') }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
calico upgrade to v3 (#3086) * calico upgrade to v3 * update calico_rr version * add missing file * change contents of main.yml as it was left old version * enable network policy by default * remove unneeded task * Fix kubelet calico settings * fix when statement * switch back to node-kubeconfig.yaml
2018-08-23 14:17:18 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
[Openstack] Make it possible to apply the new cloud provider during upgrade (#5707) * run external cloud provider setup during upgrade * change name of taskt to better reflect what it does * fix typo
2020-03-11 12:11:36 +00:00
- { role: kubernetes-apps/external_cloud_controller, tags: external-cloud-controller }
calico upgrade to v3 (#3086) * calico upgrade to v3 * update calico_rr version * add missing file * change contents of main.yml as it was left old version * enable network policy by default * remove unneeded task * Fix kubelet calico settings * fix when statement * switch back to node-kubeconfig.yaml
2018-08-23 14:17:18 +00:00
- { role: network_plugin, tags: network }
- { role: kubernetes-apps/network_plugin, tags: network }
- { role: kubernetes-apps/policy_controller, tags: policy-controller }
Yamllint fixes (#4410) * Lint everything in the repository with yamllint * yamllint fixes: syntax fixes only * yamllint fixes: move comments to play names * yamllint fixes: indent comments in .gitlab-ci.yml file
2019-04-01 09:38:33 +00:00
- name: Finally handle worker upgrades, based on given batch size
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
hosts: kube_node:calico_rr:!kube_control_plane
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
serial: "{{ serial | default('20%') }}"
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
- { role: upgrade/pre-upgrade, tags: pre-upgrade }
etcd: deploy container engine if needed (#7532) If the etcd cluster is separate and the etcd_deployment_type is "host", there is no need for a container engine on the etcd nodes Do not rely on a 'default(true)' filter, but define a proper default in kubespray-defaults depending on etcd deployment method and if internal or external etcd is used
2021-10-12 07:31:47 +00:00
- { role: container-engine, tags: "container-engine", when: deploy_container_engine }
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
- { role: kubernetes/node, tags: node }
Remove non-kubeadm deployment (#3811) * Remove non-kubeadm deployment * More cleanup * More cleanup * More cleanup * More cleanup * Fix gitlab * Try stop gce first before absent to make the delete process work * More cleanup * Fix bug with checking if kubeadm has already run * Fix bug with checking if kubeadm has already run * More fixes * Fix test * fix * Fix gitlab checkout untill kubespray 2.8 is on quay * Fixed * Add upgrade path from non-kubeadm to kubeadm. Revert ssl path * Readd secret checking * Do gitlab checks from v2.7.0 test upgrade path to 2.8.0 * fix typo * Fix CI jobs to kubeadm again. Fix broken hyperkube path * Fix gitlab * Fix rotate tokens * More fixes * More fixes * Fix tokens
2018-12-06 10:33:38 +00:00
- { role: kubernetes/kubeadm, tags: kubeadm }
tag role kubernetes/node-label in playbooks (#5560)
2020-01-20 19:43:36 +00:00
- { role: kubernetes/node-label, tags: node-label }
Fix upgrade k8s
2018-09-12 05:24:31 +00:00
- { role: upgrade/post-upgrade, tags: post-upgrade }
move network plugins out of grouped upgrades
2017-03-23 14:18:51 +00:00
Replace kube-master with kube_control_plane (#7256) This replaces kube-master with kube_control_plane because of [1]: The Kubernetes project is moving away from wording that is considered offensive. A new working group WG Naming was created to track this work, and the word "master" was declared as offensive. A proposal was formalized for replacing the word "master" with "control plane". This means it should be removed from source code, documentation, and user-facing configuration from Kubernetes and its sub-projects. NOTE: The reason why this changes it to kube_control_plane not kube-control-plane is for valid group names on ansible. [1]: https://github.com/kubernetes/enhancements/blob/master/keps/sig-cluster-lifecycle/kubeadm/2067-rename-master-label-taint/README.md#motivation
2021-03-24 00:26:05 +00:00
- hosts: kube_control_plane[0]
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
move network plugins out of grouped upgrades
2017-03-23 14:18:51 +00:00
any_errors_fatal: true
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
move network plugins out of grouped upgrades
2017-03-23 14:18:51 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
- { role: win_nodes/kubernetes_patch, tags: ["master", "win_nodes"] }
Run rotate_tokens role only once (#1970)
2017-11-15 18:50:23 +00:00
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
- hosts: calico_rr
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
- { role: network_plugin/calico/rr, tags: network }
Replace kube-master with kube_control_plane (#7256) This replaces kube-master with kube_control_plane because of [1]: The Kubernetes project is moving away from wording that is considered offensive. A new working group WG Naming was created to track this work, and the word "master" was declared as offensive. A proposal was formalized for replacing the word "master" with "control plane". This means it should be removed from source code, documentation, and user-facing configuration from Kubernetes and its sub-projects. NOTE: The reason why this changes it to kube_control_plane not kube-control-plane is for valid group names on ansible. [1]: https://github.com/kubernetes/enhancements/blob/master/keps/sig-cluster-lifecycle/kubeadm/2067-rename-master-label-taint/README.md#motivation
2021-03-24 00:26:05 +00:00
- hosts: kube_control_plane
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Make any_errors_fatal configurable Useful at scale when 1 or 2 noes my fail and you can proceed with the majority and go back and fix the others later.
2017-04-04 09:52:39 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Add graceful upgrade process Based on #718 introduced by rsmitty. Includes all roles and all options to support deployment of new hosts in case they were added to inventory. Main difference here is that master role is evaluated first so that master components get upgraded first. Fixes #694
2017-02-14 16:08:44 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Install Helm on all masters
2018-09-10 09:39:26 +00:00
- { role: kubernetes-apps, tags: apps }
Run dns_late preinstall tasks on all k8s nodes (#4672) * Run dns_late preinstall tasks on all k8s nodes Related issue: #4656 Change-Id: I63f8559ef1a497b7580ab084561e6603fe647834 * Fix ansible-lint Change-Id: Ia5b33fa63dbc36d8c3e9557ef3f2ea02af2325a5 * Fix recover_control_plane lint issues Change-Id: I16643a3193c11b6ba704e9698812cac7e4fd19a8
2019-04-29 12:12:22 +00:00
Rename ansible groups to use _ instead of - (#7552) * rename ansible groups to use _ instead of - k8s-cluster -> k8s_cluster k8s-node -> k8s_node calico-rr -> calico_rr no-floating -> no_floating Note: kube-node,k8s-cluster groups in upgrade CI need clean-up after v2.16 is tagged * ensure old groups are mapped to the new ones
2021-04-29 12:20:50 +00:00
- hosts: k8s_cluster
Gather just the necessary facts (#5955) * Gather just the necessary facts * Move fact gathering to separate playbook.
2020-04-17 23:23:36 +00:00
gather_facts: False
Run dns_late preinstall tasks on all k8s nodes (#4672) * Run dns_late preinstall tasks on all k8s nodes Related issue: #4656 Change-Id: I63f8559ef1a497b7580ab084561e6603fe647834 * Fix ansible-lint Change-Id: Ia5b33fa63dbc36d8c3e9557ef3f2ea02af2325a5 * Fix recover_control_plane lint issues Change-Id: I16643a3193c11b6ba704e9698812cac7e4fd19a8
2019-04-29 12:12:22 +00:00
any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
Fix proxy usage when *_PROXY are present in environment (#7309) Since a790935d02dc2787f6de41695ec955dc49fc93b1 all proxy users should be properly configured Now when you have *_PROXY vars in your environment it can leads to failure if NO_PROXY is not correct, or to persistent configuration changes as seen with kubeadm in 1c5391dda78b43fc629320dd59f1234e81afb2ad Instead of playing constant whack-a-bug, inject empty *_PROXY vars everywhere at the play level, and override at the task level when needed Signed-off-by: Etienne Champetier <e.champetier@ateme.com>
2021-02-23 17:44:02 +00:00
environment: "{{ proxy_disable_env }}"
Run dns_late preinstall tasks on all k8s nodes (#4672) * Run dns_late preinstall tasks on all k8s nodes Related issue: #4656 Change-Id: I63f8559ef1a497b7580ab084561e6603fe647834 * Fix ansible-lint Change-Id: Ia5b33fa63dbc36d8c3e9557ef3f2ea02af2325a5 * Fix recover_control_plane lint issues Change-Id: I16643a3193c11b6ba704e9698812cac7e4fd19a8
2019-04-29 12:12:22 +00:00
roles:
Fix chicken and egg problem with proxy_env not defined on the first … (#5896) * Fix chicken and egg problem with proxy_env not defined on the first envinronment usage. * Disable fact gathering for the first proxy_env evaluation. * Move proxy_env var set up from the role defaults to the root playbooks as fact.
2020-04-08 07:53:43 +00:00
- { role: kubespray-defaults }
Remove kubedns and dnsmasq. Move dns_late phase after apps (#4406) Both kubedns and dnsmasq modes are long not maintained. We should run dns_late steps at the end because sshd makes DNS lookups during Ansible run and has 2s timeouts for each failed lookup trying to connect to coredns before it is ready.
2019-04-01 19:32:34 +00:00
- { role: kubernetes/preinstall, when: "dns_mode != 'none' and resolvconf_mode == 'host_resolvconf'", tags: resolvconf }
Reference in a new issue Copy permalink