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

README.md

@@ -95,7 +95,7 @@ Supported Components
     -   [docker](https://www.docker.com/) v17.03 (see note)
     -   [rkt](https://github.com/rkt/rkt) v1.21.0 (see Note 2)
 -   Network Plugin
-    -   [calico](https://github.com/projectcalico/calico) v2.6.8
+    -   [calico](https://github.com/projectcalico/calico) v3.1.3
     -   [canal](https://github.com/projectcalico/canal) (given calico/flannel versions)
     -   [cilium](https://github.com/cilium/cilium) v1.2.0
     -   [contiv](https://github.com/contiv/install) v1.1.7

docs/calico.md

@@ -1,6 +1,13 @@
 Calico
 ===========
 
+---
+ **N.B. Version 2.6.5 upgrade to 3.1.1 is upgrading etcd store to etcdv3**
+ If you create automated backups of etcdv2 please switch for creating etcdv3 backups, as kubernetes and calico now uses etcdv3
+ After migration you can check `/tmp/calico_upgrade/` directory for converted items to etcdv3.
+ **PLEASE TEST upgrade before upgrading production cluster.**
+ ---
+
 Check if the calico-node container is running
 
 ```
@@ -86,7 +93,7 @@ To do so you can deploy BGP route reflectors and peer `calico-node` with them as
 recommended here:
 
 * https://hub.docker.com/r/calico/routereflector/
-* http://docs.projectcalico.org/v2.0/reference/private-cloud/l3-interconnect-fabric
+* https://docs.projectcalico.org/v3.1/reference/private-cloud/l3-interconnect-fabric
 
 You need to edit your inventory and add:
 

inventory/sample/group_vars/k8s-cluster.yml

@@ -83,9 +83,6 @@ kube_network_plugin: calico
 # weave_no_masq_local: true
 # weave_extra_args: ~
 
-# Enable kubernetes network policies
-enable_network_policy: false
-
 # Kubernetes internal network for services, unused block of space.
 kube_service_addresses: 10.233.0.0/18
 

roles/download/defaults/main.yml

@@ -32,11 +32,11 @@ kubeadm_version: "{{ kube_version }}"
 etcd_version: v3.2.18
 # TODO(mattymo): Move calico versions to roles/network_plugins/calico/defaults
 # after migration to container download
-calico_version: "v2.6.8"
+calico_version: "v3.1.3"
-calico_ctl_version: "v1.6.3"
+calico_ctl_version: "v3.1.3"
-calico_cni_version: "v1.11.4"
+calico_cni_version: "v3.1.3"
-calico_policy_version: "v1.0.3"
+calico_policy_version: "v3.1.3"
-calico_rr_version: "v0.4.2"
+calico_rr_version: "v0.6.1"
 flannel_version: "v0.10.0"
 flannel_cni_version: "v0.3.0"
 vault_version: 0.10.1

roles/kubernetes-apps/network_plugin/calico/tasks/main.yml

@@ -7,5 +7,13 @@
     resource: "{{item.item.type}}"
     filename: "{{kube_config_dir}}/{{item.item.file}}"
     state: "latest"
-  with_items: "{{ calico_node_manifests.results }}"
+  with_items:
-  when: inventory_hostname == groups['kube-master'][0] and not item|skipped
+    - "{{ calico_node_manifests.results }}"
+  when:
+    - inventory_hostname == groups['kube-master'][0] and not item|skipped
+
+- name: "calico upgrade complete"
+  shell: "{{ bin_dir }}/calico-upgrade complete --no-prompts --apiconfigv1 /etc/calico/etcdv2.yml --apiconfigv3 /etc/calico/etcdv3.yml"
+  when:
+    - inventory_hostname == groups['kube-master'][0]
+    - calico_upgrade_needed|default(False)

roles/kubernetes-apps/policy_controller/calico/tasks/main.yml

@@ -2,20 +2,12 @@
 - name: Set cert dir
   set_fact:
     calico_cert_dir: "{{ canal_cert_dir }}"
-  when: kube_network_plugin == 'canal'
+  when:
+    - kube_network_plugin == 'canal'
   tags:
     - facts
     - canal
 
-- name: Delete the old calico-policy-controller if it exist
-  kube:
-    name: calico-policy-controller
-    kubectl: "{{bin_dir}}/kubectl"
-    resource: rs
-    namespace: "kube-system"
-    state: absent
-  run_once: true
-
 - name: Create calico-kube-controllers manifests
   template:
     src: "{{item.file}}.j2"
@@ -26,7 +18,9 @@
     - {name: calico-kube-controllers, file: calico-kube-cr.yml, type: clusterrole}
     - {name: calico-kube-controllers, file: calico-kube-crb.yml, type: clusterrolebinding}
   register: calico_kube_manifests
-  when: inventory_hostname == groups['kube-master'][0] and not item|skipped
+  when:
+    - inventory_hostname == groups['kube-master'][0]
+    - rbac_enabled or item.type not in rbac_resources
 
 - name: Start of Calico kube controllers
   kube:
@@ -36,5 +30,8 @@
     resource: "{{item.item.type}}"
     filename: "{{kube_config_dir}}/{{item.item.file}}"
     state: "latest"
-  with_items: "{{ calico_kube_manifests.results }}"
+  with_items:
-  when: inventory_hostname == groups['kube-master'][0] and not item|skipped
+    - "{{ calico_kube_manifests.results }}"
+  when:
+    - inventory_hostname == groups['kube-master'][0]
+    - not item|skipped

roles/kubernetes-apps/policy_controller/calico/templates/calico-kube-controllers.yml.j2

@@ -6,8 +6,12 @@ metadata:
   labels:
     k8s-app: calico-kube-controllers
     kubernetes.io/cluster-service: "true"
+  annotations:
+    scheduler.alpha.kubernetes.io/critical-pod: ''
 spec:
   replicas: 1
+  strategy:
+    type: Recreate
   selector:
     matchLabels:
       kubernetes.io/cluster-service: "true"
@@ -45,17 +49,6 @@ spec:
               value: "{{ calico_cert_dir }}/cert.crt"
             - name: ETCD_KEY_FILE
               value: "{{ calico_cert_dir }}/key.pem"
-            # Location of the Kubernetes API - this shouldn't need to be
-            # changed so long as it is used in conjunction with
-            # CONFIGURE_ETC_HOSTS="true".
-            - name: K8S_API
-              value: "https://kubernetes.default"
-            # Configure /etc/hosts within the container to resolve
-            # the kubernetes.default Service to the correct clusterIP
-            # using the environment provided by the kubelet.
-            # This removes the need for KubeDNS to resolve the Service.
-            - name: CONFIGURE_ETC_HOSTS
-              value: "true"
           volumeMounts:
           - mountPath: {{ calico_cert_dir }}
             name: etcd-certs

roles/kubernetes-apps/policy_controller/calico/templates/calico-kube-cr.yml.j2

@@ -12,6 +12,14 @@ rules:
       - pods
       - namespaces
       - networkpolicies
+      - nodes
+    verbs:
+      - watch
+      - list
+  - apiGroups:
+    - networking.k8s.io
+    resources:
+      - networkpolicies
     verbs:
       - watch
       - list

roles/kubernetes/kubeadm/tasks/main.yml

@@ -88,6 +88,15 @@
     - is_kube_master
     - kubeadm_discovery_address != kube_apiserver_endpoint
 
+# FIXME(mattymo): Reconcile kubelet kubeconfig filename for both deploy modes
+- name: Symlink kubelet kubeconfig for calico/canal
+  file:
+    src: "{{ kube_config_dir }}/kubelet.conf"
+    dest: "{{ kube_config_dir }}/node-kubeconfig.yaml"
+    state: link
+    force: yes
+  when: kube_network_plugin in ['calico','canal']
+
 - name: Restart all kube-proxy pods to ensure that they load the new configmap
   shell: "{{ bin_dir }}/kubectl delete pod -n kube-system -l k8s-app=kube-proxy"
   delegate_to: "{{groups['kube-master']|first}}"

roles/kubernetes/node/templates/kubelet-container.j2

@@ -22,6 +22,7 @@
   -v {{ docker_daemon_graph }}:{{ docker_daemon_graph }}:rw \
   -v /var/log:/var/log:rw \
   -v /var/lib/kubelet:/var/lib/kubelet:shared \
+  -v /var/lib/calico:/var/lib/calico:shared \
   -v /var/lib/cni:/var/lib/cni:shared \
   -v /var/run:/var/run:rw \
   {# we can run into issues with double mounting /var/lib/kubelet #}

roles/kubernetes/node/templates/kubelet.rkt.service.j2

@@ -37,6 +37,10 @@ ExecStart=/usr/bin/rkt run \
         --volume etc-cni,kind=host,source=/etc/cni,readOnly=true \
         --volume opt-cni,kind=host,source=/opt/cni,readOnly=true \
         --volume var-lib-cni,kind=host,source=/var/lib/cni,readOnly=false \
+{% endif %}
+{% if kube_network_plugin in ["calico", "canal"] %}
+        --volume var-lib-calico,kind=host,source=/var/lib/calico,readOnly=false \
+{% endif %}
 {# we can run into issues with double mounting /var/lib/kubelet #}
 {# surely there's a better way to do this #}
 {% if '/var/lib/kubelet' not in kubelet_flexvolumes_plugins_dir %}
@@ -55,6 +59,9 @@ ExecStart=/usr/bin/rkt run \
         --mount volume=etc-cni,target=/etc/cni \
         --mount volume=opt-cni,target=/opt/cni \
         --mount volume=var-lib-cni,target=/var/lib/cni \
+{% if kube_network_plugin in ["calico", "canal"] %}
+        --mount volume=var-lib-calico,target=/var/lib/calico \
+{% endif %}
 {% endif %}
         --mount volume=os-release,target=/etc/os-release \
         --mount volume=dns,target=/etc/resolv.conf \

roles/kubernetes/preinstall/tasks/main.yml

@@ -89,6 +89,7 @@
   with_items:
     - "/etc/cni/net.d"
     - "/opt/cni/bin"
+    - "/var/lib/calico"
   when:
     - kube_network_plugin in ["calico", "weave", "canal", "flannel", "contiv", "cilium"]
     - inventory_hostname in groups['k8s-cluster']

roles/kubernetes/preinstall/tasks/set_resolv_facts.yml

@@ -98,7 +98,7 @@
       {%- elif dns_mode == 'coredns_dual' and not dns_early|bool -%}
         {{ [ skydns_server ] + [ skydns_server_secondary ] + upstream_dns_servers|default([]) }}
       {%- elif dns_mode == 'manual' and not dns_early|bool -%}
-        {{ [ manual_dns_server ] + upstream_dns_servers|default([]) }}
+        {{ ( manual_dns_server.split(',') | list) + upstream_dns_servers|default([]) }}
       {%- elif dns_early|bool -%}
         {{ upstream_dns_servers|default([]) }}
       {%- else -%}

roles/kubespray-defaults/defaults/main.yaml

@@ -187,7 +187,7 @@ dashboard_enabled: true
 efk_enabled: false
 helm_enabled: false
 registry_enabled: false
-enable_network_policy: false
+enable_network_policy: true
 local_volume_provisioner_enabled: "{{ local_volumes_enabled | default('false') }}"
 persistent_volumes_enabled: false
 cephfs_provisioner_enabled: false

roles/network_plugin/calico/defaults/main.yml

@@ -2,9 +2,12 @@
 # Enables Internet connectivity from containers
 nat_outgoing: true
 
+#add default ippool name
+calico_pool_name: "default-pool"
+
 # Use IP-over-IP encapsulation across hosts
 ipip: true
-ipip_mode: always  # change to "cross-subnet" if you only want ipip encapsulation on traffic going across subnets
+ipip_mode: Always  # change to "CrossSubnet" if you only want ipip encapsulation on traffic going across subnets
 
 # Set to true if you want your calico cni binaries to overwrite the
 # ones from hyperkube while leaving other cni plugins intact.
@@ -37,6 +40,9 @@ calico_felix_prometheusmetricsport: 9091
 calico_felix_prometheusgometricsenabled: "true"
 calico_felix_prometheusprocessmetricsenabled: "true"
 
+### check latest version https://github.com/projectcalico/calico-upgrade/releases
+calico_upgrade_version: v1.0.5
+
 # Should calico ignore kernel's RPF check setting,
 # see https://github.com/projectcalico/felix/blob/ab8799eaea66627e5db7717e62fca61fd9c08646/python/calico/felix/config.py#L198
 calico_node_ignorelooserpf: false
@@ -48,3 +54,6 @@ calico_node_ignorelooserpf: false
 # calico_ip_auto_method: "interface=eth.*"
 
 calico_baremetal_nodename: "{{ inventory_hostname }}"
+
+### do not enable this, this is detected in scope of tasks, this is just a default value
+calico_upgrade_needed: false

roles/network_plugin/calico/tasks/check.yml

@@ -0,0 +1,37 @@
+---
+- name: "Check vars defined correctly"
+  assert:
+    that:
+      - "calico_pool_name is defined"
+      - "calico_pool_name | match('^[a-zA-Z0-9-_\\\\.]{2,63}$')"
+      - "ipip_mode is defined"
+      - "ipip_mode in ['Always', 'CrossSubnet', 'Never']"
+    msg: "Check variable definitions seems something is wrong"
+  run_once: yes
+
+- name: "Get current version of calico cluster version"
+  shell: "{{ bin_dir }}/calicoctl version  | grep 'Cluster Version' | awk '{ print $3}'"
+  register: calico_version_on_server
+  run_once: yes
+  delegate_to: "{{ groups['kube-master'][0] }}"
+
+- name: "Determine if calico upgrade is needed"
+  block:
+    - name: "Check that calico version is enought for upgrade"
+      assert:
+        that:
+          - calico_version_on_server.stdout|version_compare('v2.6.5', '>=')
+        msg: "Your version of calico is not fresh enough for upgrade"
+
+    - name: "Set upgrade flag when version needs to be updated"
+      set_fact:
+        calico_upgrade_needed: True
+      when:
+        - calico_version_on_server.stdout|version_compare('v2.6.5', '>=')
+        - calico_version_on_server.stdout|version_compare('v3.0.0', '<')
+
+  when:
+    - 'calico_version_on_server.stdout is defined'
+    - 'calico_version_on_server.stdout != ""'
+    - inventory_hostname == groups['kube-master'][0]
+  run_once: yes

roles/network_plugin/calico/tasks/install.yml

@@ -0,0 +1,189 @@
+---
+
+- name: Calico | Write Calico cni config
+  template:
+    src: "cni-calico.conflist.j2"
+    dest: "/etc/cni/net.d/10-calico.conflist"
+    owner: kube
+
+- name: Calico | Create calico certs directory
+  file:
+    dest: "{{ calico_cert_dir }}"
+    state: directory
+    mode: 0750
+    owner: root
+    group: root
+
+- name: Calico | Link etcd certificates for calico-node
+  file:
+    src: "{{ etcd_cert_dir }}/{{ item.s }}"
+    dest: "{{ calico_cert_dir }}/{{ item.d }}"
+    state: hard
+    force: yes
+  with_items:
+    - {s: "ca.pem", d: "ca_cert.crt"}
+    - {s: "node-{{ inventory_hostname }}.pem", d: "cert.crt"}
+    - {s: "node-{{ inventory_hostname }}-key.pem", d: "key.pem"}
+
+- name: Calico | Install calicoctl container script
+  template:
+    src: calicoctl-container.j2
+    dest: "{{ bin_dir }}/calicoctl"
+    mode: 0755
+    owner: root
+    group: root
+  changed_when: false
+
+- name: Calico | Copy cni plugins from hyperkube
+  command: "{{ docker_bin_dir }}/docker run --rm -v /opt/cni/bin:/cnibindir {{ hyperkube_image_repo }}:{{ hyperkube_image_tag }} /bin/cp -r /opt/cni/bin/. /cnibindir/"
+  register: cni_task_result
+  until: cni_task_result.rc == 0
+  retries: 4
+  delay: "{{ retry_stagger | random + 3 }}"
+  changed_when: false
+  tags:
+    - hyperkube
+    - upgrade
+
+- name: Calico | Copy cni plugins from calico/cni container
+  command: "{{ docker_bin_dir }}/docker run --rm -v /opt/cni/bin:/cnibindir {{ calico_cni_image_repo }}:{{ calico_cni_image_tag }} sh -c 'cp /opt/cni/bin/* /cnibindir/'"
+  register: cni_task_result
+  until: cni_task_result.rc == 0
+  retries: 4
+  delay: "{{ retry_stagger | random + 3 }}"
+  changed_when: false
+  when:
+    - "overwrite_hyperkube_cni|bool"
+  tags:
+    - hyperkube
+    - upgrade
+
+- name: Calico | Set cni directory permissions
+  file:
+    path: /opt/cni/bin
+    state: directory
+    owner: kube
+    recurse: true
+    mode: 0755
+
+- name: Calico | wait for etcd
+  uri:
+    url: "{{ etcd_access_addresses.split(',') | first }}/health"
+    validate_certs: no
+    client_cert: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}.pem"
+    client_key: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}-key.pem"
+  register: result
+  until: result.status == 200 or result.status == 401
+  retries: 10
+  delay: 5
+  run_once: true
+
+- name: Calico | Check if calico network pool has already been configured
+  shell: >
+    {{ bin_dir }}/calicoctl get ippool | grep -w "{{ kube_pods_subnet }}" | wc -l
+  register: calico_conf
+  retries: 4
+  delay: "{{ retry_stagger | random + 3 }}"
+  delegate_to: "{{ groups['kube-master'][0] }}"
+  run_once: true
+
+- name: Calico | Configure calico network pool
+  shell: >
+    echo "
+      { "kind": "IPPool",
+        "apiVersion": "projectcalico.org/v3",
+        "metadata": {
+          "name": "{{ calico_pool_name }}",
+        },
+        "spec": {
+          "cidr": "{{ kube_pods_subnet }}",
+          "ipipMode": "{{ ipip_mode|capitalize }}",
+          "natOutgoing": {{ nat_outgoing|default(false) and not peer_with_router|default(false) }} }} " | {{ bin_dir }}/calicoctl create -f -
+  run_once: true
+  delegate_to: "{{ groups['kube-master'][0] }}"
+  when:
+    - 'calico_conf.stdout == "0"'
+
+- name: "Determine nodeToNodeMesh needed state"
+  set_fact:
+    nodeToNodeMeshEnabled: "false"
+  when:
+    - peer_with_router|default(false) or peer_with_calico_rr|default(false)
+    - inventory_hostname in groups['k8s-cluster']
+  run_once: yes
+
+
+- name: Calico | Set global as_num
+  shell: >
+    echo '
+    { "kind": "BGPConfiguration",
+    "apiVersion": "projectcalico.org/v3",
+    "metadata": {
+        "name": "default",
+    },
+    "spec": {
+        "logSeverityScreen": "Info",
+        "nodeToNodeMeshEnabled": {{ nodeToNodeMeshEnabled|default('true') }} ,
+        "asNumber": {{ global_as_num }} }} ' | {{ bin_dir }}/calicoctl --skip-exists create -f -
+  run_once: true
+  delegate_to: "{{ groups['kube-master'][0] }}"
+
+- name: Calico | Configure peering with router(s)
+  shell: >
+   echo '{
+   "apiVersion": "projectcalico.org/v3",
+   "kind": "bgpPeer",
+   "metadata": {
+      "name": "{{ inventory_hostname }}-bgp"
+   },
+   "spec": {
+      "asNumber": "{{ item.as }}",
+      "node": "{{ inventory_hostname }}",
+      "scope": "node",
+      "peerIP": "{{ item.router_id }}"
+   }}' | {{ bin_dir }}/calicoctl create --skip-exists -f -
+  retries: 4
+  delay: "{{ retry_stagger | random + 3 }}"
+  with_items:
+    - "{{ peers|default([]) }}"
+  when:
+    - peer_with_router|default(false)
+    - inventory_hostname in groups['k8s-cluster']
+
+- name: Calico | Configure peering with route reflectors
+  shell: >
+   echo '{
+   "apiVersion": "projectcalico.org/v3",
+   "kind": "bgpPeer",
+   "metadata": {
+      "name": "{{ inventory_hostname }}"
+   },
+   "spec": {
+      "asNumber": "{{ local_as | default(global_as_num)}}",
+      "scope": "node",
+      "node": "{{ inventory_hostname }}",
+      "peerIP": "{{ hostvars[item]["calico_rr_ip"]|default(hostvars[item]["ip"])|default(hostvars[item]["ansible_default_ipv4"]["address"]) }}"
+   }}' | {{ bin_dir }}/calicoctl create --skip-exists -f -
+  retries: 4
+  delay: "{{ retry_stagger | random + 3 }}"
+  with_items:
+    - "{{ groups['calico-rr'] | default([]) }}"
+  when:
+    - (peer_with_calico_rr|default(false)
+    - inventory_hostname in groups['k8s-cluster']
+    - hostvars[item]['cluster_id'] == cluster_id)
+
+- name: Calico | Create calico manifests
+  template:
+    src: "{{item.file}}.j2"
+    dest: "{{kube_config_dir}}/{{item.file}}"
+  with_items:
+    - {name: calico-config, file: calico-config.yml, type: cm}
+    - {name: calico-node, file: calico-node.yml, type: ds}
+    - {name: calico, file: calico-node-sa.yml, type: sa}
+    - {name: calico, file: calico-cr.yml, type: clusterrole}
+    - {name: calico, file: calico-crb.yml, type: clusterrolebinding}
+  register: calico_node_manifests
+  when:
+    - inventory_hostname in groups['kube-master']
+    - rbac_enabled or item.type not in rbac_resources

roles/network_plugin/calico/tasks/main.yml

@@ -1,193 +1,12 @@
 ---
-- name: Calico | Disable calico-node service if it exists
+- include_tasks: check.yml
-  service:
-    name: calico-node
-    state: stopped
-    enabled: yes
-  failed_when: false
 
-- name: Calico | Get kubelet hostname
+- include_tasks: pre.yml
-  shell: >-
-    {{ bin_dir }}/kubectl get node -o custom-columns='NAME:.metadata.name,INTERNAL-IP:.status.addresses[?(@.type=="InternalIP")].address'
-    | egrep "{{ ansible_all_ipv4_addresses | join('$|') }}$" | cut -d" " -f1
-  register: calico_kubelet_name
-  delegate_to: "{{ groups['kube-master'][0] }}"
-  when: cloud_provider is defined
 
-- name: Calico | Write Calico cni config
+- include_tasks: upgrade.yml
-  template:
-    src: "cni-calico.conflist.j2"
-    dest: "/etc/cni/net.d/10-calico.conflist"
-    owner: kube
-
-- name: Calico | Create calico certs directory
-  file:
-    dest: "{{ calico_cert_dir }}"
-    state: directory
-    mode: 0750
-    owner: root
-    group: root
-
-- name: Calico | Link etcd certificates for calico-node
-  file:
-    src: "{{ etcd_cert_dir }}/{{ item.s }}"
-    dest: "{{ calico_cert_dir }}/{{ item.d }}"
-    state: hard
-    force: yes
-  with_items:
-    - {s: "ca.pem", d: "ca_cert.crt"}
-    - {s: "node-{{ inventory_hostname }}.pem", d: "cert.crt"}
-    - {s: "node-{{ inventory_hostname }}-key.pem", d: "key.pem"}
-
-- name: Calico | Install calicoctl container script
-  template:
-    src: calicoctl-container.j2
-    dest: "{{ bin_dir }}/calicoctl"
-    mode: 0755
-    owner: root
-    group: root
-  changed_when: false
-
-- name: Calico | Copy cni plugins from hyperkube
-  command: "{{ docker_bin_dir }}/docker run --rm -v /opt/cni/bin:/cnibindir {{ hyperkube_image_repo }}:{{ hyperkube_image_tag }} /bin/cp -rf /opt/cni/bin/. /cnibindir/"
-  register: cni_task_result
-  until: cni_task_result.rc == 0
-  retries: 4
-  delay: "{{ retry_stagger | random + 3 }}"
-  changed_when: false
-  tags:
-    - hyperkube
-    - upgrade
-
-- name: Calico | Copy cni plugins from calico/cni container
-  command: "{{ docker_bin_dir }}/docker run --rm -v /opt/cni/bin:/cnibindir {{ calico_cni_image_repo }}:{{ calico_cni_image_tag }} sh -c 'cp /opt/cni/bin/* /cnibindir/'"
-  register: cni_task_result
-  until: cni_task_result.rc == 0
-  retries: 4
-  delay: "{{ retry_stagger | random + 3 }}"
-  changed_when: false
-  when: overwrite_hyperkube_cni|bool
-  tags:
-    - hyperkube
-    - upgrade
-
-- name: Calico | Set cni directory permissions
-  file:
-    path: /opt/cni/bin
-    state: directory
-    owner: kube
-    recurse: true
-    mode: 0755
-
-- name: Calico | wait for etcd
-  uri:
-    url: "{{ etcd_access_addresses.split(',') | first }}/health"
-    validate_certs: no
-    client_cert: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}.pem"
-    client_key: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}-key.pem"
-  register: result
-  until: result.status == 200 or result.status == 401
-  retries: 10
-  delay: 5
-  run_once: true
-
-- name: Calico | Check if calico network pool has already been configured
-  command: |-
-    curl \
-      --cacert {{ etcd_cert_dir }}/ca.pem \
-      --cert {{ etcd_cert_dir}}/node-{{ inventory_hostname }}.pem \
-      --key {{ etcd_cert_dir }}/node-{{ inventory_hostname }}-key.pem \
-      {{ etcd_access_addresses.split(',') | first }}/v2/keys/calico/v1/ipam/v4/pool
-  register: calico_conf
-  retries: 4
-  delay: "{{ retry_stagger | random + 3 }}"
-  run_once: true
-  changed_when: false
-
-- name: Calico | Configure calico network pool
-  shell: >
-    echo '{
-    "kind": "ipPool",
-    "spec": {"disabled": false, "ipip": {"enabled": {{ ipip }}, "mode": "{{ ipip_mode }}"},
-             "nat-outgoing": {{ nat_outgoing|default(false) and not peer_with_router|default(false) }}},
-    "apiVersion": "v1",
-    "metadata": {"cidr": "{{ kube_pods_subnet }}"}
-    }'
-    | {{ bin_dir }}/calicoctl apply -f -
-  environment:
-    NO_DEFAULT_POOLS: true
-  run_once: true
-  when: ("Key not found" in calico_conf.stdout or "nodes" not in calico_conf.stdout)
-
-- name: Calico | Get calico configuration from etcd
-  command: |-
-    curl \
-      --cacert {{ etcd_cert_dir }}/ca.pem \
-      --cert {{ etcd_cert_dir}}/node-{{ inventory_hostname }}.pem \
-      --key {{ etcd_cert_dir }}/node-{{ inventory_hostname }}-key.pem \
-      {{ etcd_access_addresses.split(',') | first }}/v2/keys/calico/v1/ipam/v4/pool
-  register: calico_pools_raw
-  retries: 4
-  delay: "{{ retry_stagger | random + 3 }}"
-  run_once: true
-
-- set_fact:
-    calico_pools: "{{ calico_pools_raw.stdout | from_json }}"
-  run_once: true
-
-- name: Calico | Set global as_num
-  command: "{{ bin_dir}}/calicoctl config set asNumber {{ global_as_num }}"
-  run_once: true
-
-- name: Calico | Disable node mesh
-  shell: "{{ bin_dir }}/calicoctl config set nodeToNodeMesh off"
-  retries: 4
-  delay: "{{ retry_stagger | random + 3 }}"
-  when: ((peer_with_router|default(false) or peer_with_calico_rr|default(false))
-          and inventory_hostname in groups['k8s-cluster'])
-  run_once: true
-
-- name: Calico | Configure peering with router(s)
-  shell: >
-   echo '{
-   "kind": "bgpPeer",
-   "spec": {"asNumber": "{{ item.as }}"},
-   "apiVersion": "v1",
-   "metadata": {"node": "{{ inventory_hostname }}", "scope": "node", "peerIP": "{{ item.router_id }}"}
-   }'
-   | {{ bin_dir }}/calicoctl create --skip-exists -f -
-  retries: 4
-  delay: "{{ retry_stagger | random + 3 }}"
-  with_items: "{{ peers|default([]) }}"
-  when: peer_with_router|default(false) and inventory_hostname in groups['k8s-cluster']
-
-- name: Calico | Configure peering with route reflectors
-  shell: >
-   echo '{
-   "kind": "bgpPeer",
-   "spec": {"asNumber": "{{ local_as | default(global_as_num)}}"},
-   "apiVersion": "v1",
-   "metadata": {"node": "{{ inventory_hostname }}",
-     "scope": "node",
-     "peerIP": "{{ hostvars[item]["calico_rr_ip"]|default(hostvars[item]["ip"])|default(hostvars[item]["ansible_default_ipv4"]["address"]) }}"}
-   }'
-   | {{ bin_dir }}/calicoctl create --skip-exists -f -
-  retries: 4
-  delay: "{{ retry_stagger | random + 3 }}"
-  with_items: "{{ groups['calico-rr'] | default([]) }}"
-  when: (peer_with_calico_rr|default(false) and inventory_hostname in groups['k8s-cluster']
-         and hostvars[item]['cluster_id'] == cluster_id)
-
-- name: Calico | Create calico manifests
-  template:
-    src: "{{item.file}}.j2"
-    dest: "{{kube_config_dir}}/{{item.file}}"
-  with_items:
-    - {name: calico-config, file: calico-config.yml, type: cm}
-    - {name: calico-node, file: calico-node.yml, type: ds}
-    - {name: calico, file: calico-node-sa.yml, type: sa}
-    - {name: calico, file: calico-cr.yml, type: clusterrole}
-    - {name: calico, file: calico-crb.yml, type: clusterrolebinding}
-  register: calico_node_manifests
   when:
-    - inventory_hostname in groups['kube-master']
+    - calico_upgrade_needed
+  run_once: yes
+  delegate_to: "{{ groups['kube-master'][0] }}"
+
+- include_tasks: install.yml

roles/network_plugin/calico/tasks/pre.yml

@@ -0,0 +1,16 @@
+---
+- name: Calico | Disable calico-node service if it exists
+  service:
+    name: calico-node
+    state: stopped
+    enabled: no
+  failed_when: false
+
+- name: Calico | Get kubelet hostname
+  shell: >-
+    {{ bin_dir }}/kubectl get node -o custom-columns='NAME:.metadata.name,INTERNAL-IP:.status.addresses[?(@.type=="InternalIP")].address'
+    | egrep "{{ ansible_all_ipv4_addresses | join('$|') }}$" | cut -d" " -f1
+  register: calico_kubelet_name
+  delegate_to: "{{ groups['kube-master'][0] }}"
+  when:
+    - "cloud_provider is defined"

roles/network_plugin/calico/tasks/upgrade.yml

@@ -0,0 +1,26 @@
+---
+- name: "Download calico-upgrade tool (force version)"
+  get_url:
+    url: "https://github.com/projectcalico/calico-upgrade/releases/download/{{ calico_upgrade_version }}/calico-upgrade"
+    dest: "{{ bin_dir }}/calico-upgrade"
+    mode: 0755
+    owner: root
+    group: root
+    force: yes
+
+- name: "Create etcdv2 and etcdv3 calicoApiConfig"
+  template:
+    src: "{{ item }}-store.yml.j2"
+    dest: "/etc/calico/{{ item }}.yml"
+  with_items:
+    - "etcdv2"
+    - "etcdv3"
+
+- name: "Tests data migration (dry-run)"
+  shell: "{{ bin_dir }}/calico-upgrade dry-run --output-dir=/tmp --apiconfigv1 /etc/calico/etcdv2.yml --apiconfigv3 /etc/calico/etcdv3.yml"
+  register: calico_upgrade_test_data
+  failed_when: '"Successfully" not in calico_upgrade_test_data.stdout'
+
+- name: "If test migration is success continue with calico data real migration"
+  shell: "{{ bin_dir }}/calico-upgrade start --no-prompts --apiconfigv1 /etc/calico/etcdv2.yml --apiconfigv3 /etc/calico/etcdv3.yml --output-dir=/tmp/calico_upgrade"
+  register: calico_upgrade_migration_data

roles/network_plugin/calico/templates/calico-node.yml.j2

@@ -77,12 +77,6 @@ spec:
             # Set Felix logging to "info"
             - name: FELIX_LOGSEVERITYSCREEN
               value: "info"
-            # Disable autocreation of pools
-            - name: CALICO_NO_DEFAULT_POOLS
-              value: "true"
-            # Enable libnetwork
-            - name: CALICO_LIBNETWORK_ENABLED
-              value: "true"
             # Set MTU for tunnel device used if ipip is enabled
 {% if calico_mtu is defined %}
             - name: FELIX_IPINIPMTU
@@ -158,6 +152,8 @@ spec:
               readOnly: true
             - mountPath: /var/run/calico
               name: var-run-calico
+            - mountPath: /var/lib/calico
+              name: var-lib-calico
               readOnly: false
             - mountPath: /calico-secrets
               name: etcd-certs
@@ -169,6 +165,9 @@ spec:
         - name: var-run-calico
           hostPath:
             path: /var/run/calico
+        - name: var-lib-calico
+          hostPath:
+            path: /var/lib/calico
         # Used to install CNI.
         - name: cni-bin-dir
           hostPath:
@@ -184,4 +183,3 @@ spec:
     rollingUpdate:
       maxUnavailable: {{ serial | default('20%') }}
     type: RollingUpdate
-

roles/network_plugin/calico/templates/etcdv2-store.yml.j2

@@ -0,0 +1,9 @@
+apiVersion: v1
+kind: calicoApiConfig
+metadata:
+spec:
+  datastoreType: "etcdv2"
+  etcdEndpoints: "{{ etcd_access_addresses }}"
+  etcdKeyFile: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}-key.pem"
+  etcdCertFile: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}.pem"
+  etcdCACertFile: "{{ etcd_cert_dir }}/ca.pem"

roles/network_plugin/calico/templates/etcdv3-store.yml.j2

@@ -0,0 +1,9 @@
+apiVersion: projectcalico.org/v3
+kind: CalicoAPIConfig
+metadata:
+spec:
+  datastoreType: "etcdv3"
+  etcdEndpoints: "{{ etcd_access_addresses }}"
+  etcdKeyFile: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}-key.pem"
+  etcdCertFile: "{{ etcd_cert_dir }}/node-{{ inventory_hostname }}.pem"
+  etcdCACertFile: "{{ etcd_cert_dir }}/ca.pem"

roles/network_plugin/canal/templates/canal-node.yaml.j2

@@ -28,6 +28,9 @@ spec:
         - name: lib-modules
           hostPath:
             path: /lib/modules
+        - name: var-lib-calico
+          hostPath:
+            path: /var/lib/calico
         - name: var-run-calico
           hostPath:
             path: /var/run/calico
@@ -202,6 +205,9 @@ spec:
             - mountPath: /var/run/calico
               name: var-run-calico
               readOnly: false
+            - mountPath: /var/lib/calico
+              name: var-lib-calico
+              readOnly: false
             - name: "canal-certs"
               mountPath: "{{ canal_cert_dir }}"
               readOnly: true

roles/reset/tasks/main.yml

@@ -136,6 +136,7 @@
     - "{{ bin_dir }}/kubeadm"
     - "{{ bin_dir }}/helm"
     - "{{ bin_dir }}/calicoctl"
+    - "{{ bin_dir }}/calico-upgrade"
     - "{{ bin_dir }}/weave"
     - /var/lib/rkt
     - /etc/vault

upgrade-cluster.yml

@@ -78,10 +78,19 @@
     - { role: kubernetes/master, tags: master }
     - { role: kubernetes/client, tags: client }
     - { role: kubernetes-apps/cluster_roles, tags: cluster-roles }
-    - { role: network_plugin, tags: network }
     - { role: upgrade/post-upgrade, tags: post-upgrade }
   environment: "{{proxy_env}}"
 
+#Upgrade calico on all masters and nodes
+- hosts: kube-master:kube-node
+  any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
+  serial: "{{ serial | default('20%') }}"
+  roles:
+    - { role: kubespray-defaults}
+    - { role: network_plugin, tags: network }
+    - { role: kubernetes-apps/network_plugin, tags: network }
+    - { role: kubernetes-apps/policy_controller, tags: policy-controller }
+
 #Finally handle worker upgrades, based on given batch size
 - hosts: kube-node:!kube-master
   any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
@@ -90,7 +99,6 @@
     - { role: kubespray-defaults}
     - { role: upgrade/pre-upgrade, tags: pre-upgrade }
     - { role: kubernetes/node, tags: node }
-    - { role: network_plugin, tags: network }
     - { role: upgrade/post-upgrade, tags: post-upgrade }
     - { role: kubernetes/kubeadm, tags: kubeadm, when: "kubeadm_enabled" }
     - { role: kubespray-defaults}
@@ -102,14 +110,6 @@
     - { role: kubespray-defaults}
     - { role: kubernetes-apps/rotate_tokens, tags: rotate_tokens, when: "secret_changed|default(false)" }
 
-- hosts: kube-master
-  any_errors_fatal: true
-  roles:
-    - { role: kubespray-defaults}
-    - { role: kubernetes-apps/network_plugin, tags: network }
-    - { role: kubernetes-apps/policy_controller, tags: policy-controller }
-    - { role: kubernetes/client, tags: client }
-
 - hosts: calico-rr
   any_errors_fatal: "{{ any_errors_fatal | default(true) }}"
   roles: