Merge 1d26a37bbf into 3888a29bb1

.github/ISSUE_TEMPLATE.md

@@ -51,9 +51,6 @@ extra_agent_args: ""
 
 kube_vip_tag_version: ""
 
-kube_vip_cloud_provider_tag_version: ""
-kube_vip_lb_ip_range: ""
-
 metal_lb_speaker_tag_version: ""
 metal_lb_controller_tag_version: ""
 

.github/workflows/cache.yml

@@ -11,12 +11,12 @@ jobs:
 
     steps:
       - name: Check out the codebase
-        uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v3 4.1.1
+        uses: actions/checkout@e2f20e631ae6d7dd3b768f56a5d2af784dd54791 # v3 2.5.0
         with:
           ref: ${{ github.event.pull_request.head.sha }}
 
       - name: Set up Python ${{ env.PYTHON_VERSION }}
-        uses: actions/setup-python@0a5c61591373683505ea898e09a3ea4f39ef2b9c # 5.0.0
+        uses: actions/setup-python@75f3110429a8c05be0e1bf360334e4cced2b63fa # 2.3.3
         with:
           python-version: ${{ env.PYTHON_VERSION }}
           cache: 'pip' # caching pip dependencies

.github/workflows/ci.yml

@@ -6,16 +6,7 @@ on:
       - opened
       - synchronize
     paths-ignore:
-      - '**/.gitignore'
-      - '**/FUNDING.yml'
-      - '**/host.ini'
-      - '**/*.md'
-      - '**/.editorconfig'
-      - '**/ansible.example.cfg'
-      - '**/deploy.sh'
-      - '**/LICENSE'
-      - '**/reboot.sh'
-      - '**/reset.sh'
+      - '**/README.md'
 jobs:
   pre:
     uses: ./.github/workflows/cache.yml

.github/workflows/lint.yml

@@ -11,12 +11,12 @@ jobs:
 
     steps:
       - name: Check out the codebase
-        uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v3 4.1.1
+        uses: actions/checkout@e2f20e631ae6d7dd3b768f56a5d2af784dd54791 # v3 2.5.0
         with:
           ref: ${{ github.event.pull_request.head.sha }}
 
       - name: Set up Python ${{ env.PYTHON_VERSION }}
-        uses: actions/setup-python@0a5c61591373683505ea898e09a3ea4f39ef2b9c # 5.0.0
+        uses: actions/setup-python@75f3110429a8c05be0e1bf360334e4cced2b63fa # 2.3.3
         with:
           python-version: ${{ env.PYTHON_VERSION }}
           cache: 'pip' # caching pip dependencies
@@ -45,9 +45,9 @@ jobs:
     runs-on: self-hosted
     steps:
       - name: Checkout code
-        uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v3 4.1.1
+        uses: actions/checkout@e2f20e631ae6d7dd3b768f56a5d2af784dd54791 # v3 2.5.0
       - name: Ensure SHA pinned actions
-        uses: zgosalvez/github-actions-ensure-sha-pinned-actions@ba37328d4ea95eaf8b3bd6c6cef308f709a5f2ec # 3.0.3
+        uses: zgosalvez/github-actions-ensure-sha-pinned-actions@af2eb3226618e2494e3d9084f515ad6dcf16e229 # 2.0.1
         with:
           allowlist: |
             aws-actions/

.github/workflows/test.yml

@@ -13,14 +13,13 @@ jobs:
           - ipv6
           - single_node
           - calico
-          - kube-vip
       fail-fast: false
     env:
       PYTHON_VERSION: "3.11"
 
     steps:
       - name: Check out the codebase
-        uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v3 4.1.1
+        uses: actions/checkout@e2f20e631ae6d7dd3b768f56a5d2af784dd54791 # v3 2.5.0
         with:
           ref: ${{ github.event.pull_request.head.sha }}
 
@@ -58,7 +57,7 @@ jobs:
           EOF
 
       - name: Set up Python ${{ env.PYTHON_VERSION }}
-        uses: actions/setup-python@0a5c61591373683505ea898e09a3ea4f39ef2b9c # 5.0.0
+        uses: actions/setup-python@75f3110429a8c05be0e1bf360334e4cced2b63fa # 2.3.3
         with:
           python-version: ${{ env.PYTHON_VERSION }}
           cache: 'pip' # caching pip dependencies
@@ -117,9 +116,8 @@ jobs:
 
       - name: Upload log files
         if: always() # do this even if a step before has failed
-        uses: actions/upload-artifact@26f96dfa697d77e81fd5907df203aa23a56210a8 # 4.3.0
+        uses: actions/upload-artifact@83fd05a356d7e2593de66fc9913b3002723633cb # 3.1.1
         with:
           name: logs
           path: |
             ${{ runner.temp }}/logs
-          overwrite: true

.gitignore

@@ -1,4 +1,6 @@
 .env/
 *.log
 ansible.cfg
-kubeconfig
+
+__pycache__/
+*.py[cod]

README.md

@@ -73,6 +73,21 @@ This requires at least k3s version `1.19.1` however the version is configurable
 
 If needed, you can also edit `inventory/my-cluster/group_vars/all.yml` to match your environment.
 
+#### verify routing prefixes and ip addresses in all.yml
+Ensure that the ip routing prefix matches the the one from the flannel_iface (`eth0` by default) for each host.
+Example `192.168.30` is the default routing prefix, and it is used in the following variables:
+- apiserver_endpoint
+- metal_lb_bgp_peer_address (commented by default)
+- metal_lb_ip_range
+Also Ensure that the apiserver_endpoint is not in the metal_lb_ip_range
+
+For your convience The playbook site.yml verifies the above elements of the config are valid.
+* Optionally to skip these verifications set, in all.yml `verify_config: false`  
+* Optionally to just verify the config Run
+    `ansible-playbook -i inventory/my-cluster/hosts.ini verify_config.yml`
+* Optionally manually recover ip routing prefix based on inet and netmask from
+    `ansible -i inventory/my-cluster/hosts.ini --become -m shell -a 'ifconfig eth0 | grep "inet "' all`
+
 ### ☸️ Create Cluster
 
 Start provisioning of the cluster using the following command:

inventory/sample/group_vars/all.yml

@@ -56,13 +56,6 @@ extra_agent_args: >-
 # image tag for kube-vip
 kube_vip_tag_version: "v0.6.4"
 
-# tag for kube-vip-cloud-provider manifest
-# kube_vip_cloud_provider_tag_version: "main"
-
-# kube-vip ip range for load balancer
-# (uncomment to use kube-vip for services instead of MetalLB)
-# kube_vip_lb_ip_range: "192.168.30.80-192.168.30.90"
-
 # metallb type frr or native
 metal_lb_type: "native"
 
@@ -147,3 +140,5 @@ custom_registries_yaml: |
 #   HTTP_PROXY: "http://proxy.domain.local:3128"
 #   HTTPS_PROXY: "http://proxy.domain.local:3128"
 #   NO_PROXY: "*.domain.local,127.0.0.0/8,10.0.0.0/8,172.16.0.0/12,192.168.0.0/16"
+
+verify_config: true

molecule/README.md

@@ -15,8 +15,6 @@ We have these scenarios:
   Very similar to the default scenario, but uses only a single node for all cluster functionality.
 - **calico**:
   The same as single node, but uses calico cni instead of flannel.
-- **kube-vip**
-  The same as single node, but uses kube-vip as service loadbalancer instead of MetalLB
 
 ## How to execute
 

molecule/kube-vip/molecule.yml

@@ -1,49 +0,0 @@
----
-dependency:
-  name: galaxy
-driver:
-  name: vagrant
-platforms:
-  - name: control1
-    box: generic/ubuntu2204
-    memory: 4096
-    cpus: 4
-    config_options:
-      # We currently can not use public-key based authentication on Ubuntu 22.04,
-      # see: https://github.com/chef/bento/issues/1405
-      ssh.username: "vagrant"
-      ssh.password: "vagrant"
-    groups:
-      - k3s_cluster
-      - master
-    interfaces:
-      - network_name: private_network
-        ip: 192.168.30.62
-provisioner:
-  name: ansible
-  env:
-    ANSIBLE_VERBOSITY: 1
-  playbooks:
-    converge: ../resources/converge.yml
-    side_effect: ../resources/reset.yml
-    verify: ../resources/verify.yml
-  inventory:
-    links:
-      group_vars: ../../inventory/sample/group_vars
-scenario:
-  test_sequence:
-    - dependency
-    - cleanup
-    - destroy
-    - syntax
-    - create
-    - prepare
-    - converge
-    # idempotence is not possible with the playbook in its current form.
-    - verify
-    # We are repurposing side_effect here to test the reset playbook.
-    # This is why we do not run it before verify (which tests the cluster),
-    # but after the verify step.
-    - side_effect
-    - cleanup
-    - destroy

molecule/kube-vip/overrides.yml

@@ -1,17 +0,0 @@
----
-- name: Apply overrides
-  hosts: all
-  tasks:
-    - name: Override host variables
-      ansible.builtin.set_fact:
-        # See:
-        # https://github.com/flannel-io/flannel/blob/67d603aaf45ef80f5dd39f43714fc5e6f8a637eb/Documentation/troubleshooting.md#Vagrant
-        flannel_iface: eth1
-
-        # The test VMs might be a bit slow, so we give them more time to join the cluster:
-        retry_count: 45
-
-        # Make sure that our IP ranges do not collide with those of the other scenarios
-        apiserver_endpoint: "192.168.30.225"
-        # Use kube-vip instead of MetalLB
-        kube_vip_lb_ip_range: "192.168.30.110-192.168.30.119"

roles/k3s_server/tasks/kube-vip.yml

@@ -1,27 +0,0 @@
----
-- name: Create manifests directory on first master
-  file:
-    path: /var/lib/rancher/k3s/server/manifests
-    state: directory
-    owner: root
-    group: root
-    mode: 0644
-  when: ansible_hostname == hostvars[groups[group_name_master | default('master')][0]]['ansible_hostname']
-
-- name: Download vip cloud provider manifest to first master
-  ansible.builtin.get_url:
-    url: "https://raw.githubusercontent.com/kube-vip/kube-vip-cloud-provider/{{ kube_vip_cloud_provider_tag_version | default('main') }}/manifest/kube-vip-cloud-controller.yaml"  # noqa yaml[line-length]
-    dest: "/var/lib/rancher/k3s/server/manifests/kube-vip-cloud-controller.yaml"
-    owner: root
-    group: root
-    mode: 0644
-  when: ansible_hostname == hostvars[groups[group_name_master | default('master')][0]]['ansible_hostname']
-
-- name: Copy kubevip configMap manifest to first master
-  template:
-    src: "kubevip.yaml.j2"
-    dest: "/var/lib/rancher/k3s/server/manifests/kubevip.yaml"
-    owner: root
-    group: root
-    mode: 0644
-  when: ansible_hostname == hostvars[groups[group_name_master | default('master')][0]]['ansible_hostname']

roles/k3s_server/tasks/main.yml

@@ -29,12 +29,6 @@
 - name: Deploy metallb manifest
   include_tasks: metallb.yml
   tags: metallb
-  when: kube_vip_lb_ip_range is not defined
-
-- name: Deploy kube-vip manifest
-  include_tasks: kube-vip.yml
-  tags: kubevip
-  when: kube_vip_lb_ip_range is defined
 
 - name: Init cluster inside the transient k3s-init service
   command:

roles/k3s_server/templates/kubevip.yaml.j2

@@ -1,13 +0,0 @@
----
-apiVersion: v1
-kind: ConfigMap
-metadata:
-  name: kubevip
-  namespace: kube-system
-data:
-{% if kube_vip_lb_ip_range is string %}
-{# kube_vip_lb_ip_range was used in the legacy way: single string instead of a list #}
-{#   => transform to list with single element #}
-{% set kube_vip_lb_ip_range = [kube_vip_lb_ip_range] %}
-{% endif %}
-  range-global: {{ kube_vip_lb_ip_range | join(',') }}

roles/k3s_server/templates/vip.yaml.j2

@@ -43,7 +43,7 @@ spec:
         - name: vip_ddns
           value: "false"
         - name: svc_enable
-          value: "{{ 'true' if kube_vip_lb_ip_range is defined else 'false' }}"
+          value: "false"
         - name: vip_leaderelection
           value: "true"
         - name: vip_leaseduration

roles/k3s_server_post/tasks/main.yml

@@ -7,7 +7,6 @@
 - name: Deploy metallb pool
   include_tasks: metallb.yml
   tags: metallb
-  when: kube_vip_lb_ip_range is not defined
 
 - name: Remove tmp directory used for manifests
   file:

roles/k3s_server_post/tasks/metallb.yml

@@ -8,27 +8,6 @@
   with_items: "{{ groups[group_name_master | default('master')] }}"
   run_once: true
 
-- name: Delete outdated metallb replicas
-  shell: |-
-    set -o pipefail
-
-    REPLICAS=$(k3s kubectl --namespace='metallb-system' get replicasets \
-      -l 'component=controller,app=metallb' \
-      -o jsonpath='{.items[0].spec.template.spec.containers[0].image}, {.items[0].metadata.name}' 2>/dev/null || true)
-    REPLICAS_SETS=$(echo ${REPLICAS} | grep -v '{{ metal_lb_controller_tag_version }}' | sed -e "s/^.*\s//g")
-    if [ -n "${REPLICAS_SETS}" ] ; then
-      for REPLICAS in "${REPLICAS_SETS}"
-      do
-        k3s kubectl --namespace='metallb-system' \
-          delete rs "${REPLICAS}"
-      done
-    fi
-  args:
-    executable: /bin/bash
-  changed_when: false
-  run_once: true
-  with_items: "{{ groups[group_name_master | default('master')] }}"
-
 - name: Copy metallb CRs manifest to first master
   template:
     src: "metallb.crs.j2"

roles/verify_config/tasks/main.yml

@@ -0,0 +1,13 @@
+---
+# To be ran on localhost after verify_config_gather role.
+# Verifies gatherd facts.
+- name: Collect all routing4_prefixes into a list
+  set_fact:
+    all_routing4_prefixes: "{{ groups['all'] | map('extract', hostvars, 'routing4_prefix') | list }}"
+    routing4_prefix: "{{ groups['all'] | map('extract', hostvars, 'routing4_prefix') | list | first }}"
+
+- name: Ensure all hosts have the same routing4_prefix
+  assert:
+    that: all_routing4_prefixes | unique | length == 1
+    fail_msg: "Not all hosts have the same routing4_prefix."
+    success_msg: "Using verified routing prefix {{ routing4_prefix }} across all hosts"

roles/verify_config_gather/filter_plugins/range_to_ips.py

@@ -0,0 +1,53 @@
+import netaddr
+
+''' returns True if ip is in range
+    examples:
+    - see test functions below
+'''
+def test(ip, range, expected):
+    assert netaddr_ip_in_dash_range(ip, range) == expected
+
+def test_netaddr_ip_in_dash_range():
+    # ipv4
+    test('192.168.1.1', '192.168.1.1-192.168.1.2', True)
+    test('192.168.1.1', '192.168.1.1', True)
+    test('192.168.1.1', '192.168.1.2-192.168.1.3', False)
+    test('192.168.1.1', '192.168.1.2', False)
+
+    # ipv6 style
+    test('::ffff:192.168.1.1', '::ffff:192.168.1.1-::ffff:192.168.1.8', True)
+    test('::ffff:192.168.1.1', '::ffff:192.168.1.1', True)
+    test('::ffff:192.168.1.1', '::ffff:192.168.1.2-::ffff:192.168.1.8', False)
+    test('::ffff:192.168.1.1', '::ffff:192.168.1.2', False)
+
+    # Note I expedted true but apperently the netaddr library does not support this?? or I don't understand ipv6 :)
+    test('::2:1', '::2:1-::2:2', False)
+    '''
+    todo: ?
+    - netaddr_ip_in_dash_range('192.168.1.1',  '192.168.1.0/24')             => True (TODO: test, implement)
+    - netaddr_ip_in_dash_range('192.168.99.1', '192.168.1.0/24')            => False (TODO: test, implement)
+    '''
+
+def netaddr_ip_in_dash_range(ip, range):
+    # return False early if range is invalid
+    if '-' not in range:
+        ip_start = range
+        ip_end = range
+    else:
+        ip_start = range.split('-')[0]
+        ip_end = range.split('-')[1]
+    return ip in [str(ip) for ip in netaddr.iter_iprange(ip_start, ip_end)]
+
+
+class FilterModule(object):
+        ''' Ansible filters. Interface to custom netaddr methods.
+            https://pypi.org/project/netaddr/
+        '''
+
+        def filters(self):
+            return {
+                'netaddr_ip_in_dash_range' : netaddr_ip_in_dash_range
+                }
+
+if __name__ == '__main__':
+    test_netaddr_ip_in_dash_range()

roles/verify_config_gather/tasks/main.yml

@@ -0,0 +1,117 @@
+---
+# To be ran on all hosts before role verify_config.
+# Gathers facts and ensures they are set
+
+- name: Set routing4_prefix from regex
+  set_fact:
+    routing4_prefix: "{{ hostvars[inventory_hostname]['ansible_' ~ flannel_iface]['ipv4']['broadcast']
+      | regex_replace('\\.?255', '') }}"
+    routing6_cidr: "{{ hostvars[inventory_hostname]['ansible_' ~ flannel_iface]['ipv6'][0]['address'] }}/{{
+      hostvars[inventory_hostname]['ansible_' ~ flannel_iface]['ipv6'][0]['prefix'] }}"
+
+- name: Check if fact routing4_prefix exists and is not empty
+  assert:
+    that:
+      - routing4_prefix is defined
+      - routing4_prefix is not none
+      - routing4_prefix != ''
+    fail_msg: >-
+      The fact 'routing_4prefix' is not defined, is null, or is empty
+      (based on flannel_iface: {{ flannel_iface }} ipv4 broadcast).
+
+# metal_lb_bgp_peer_address
+- name: Assert that metal_lb_bgp_peer_address starts with routing4_prefix, or is ipv6
+  assert:
+    that:
+      - >
+        metal_lb_bgp_peer_address.startswith(routing4_prefix)
+        or metal_lb_bgp_peer_address.matches('.*:.*')
+    fail_msg: >
+      The fact 'metal_lb_bgp_peer_address' <{{ metal_lb_bgp_peer_address }}>
+      doesn't start with the routing prefix <{{ routing4_prefix }}>
+  when: metal_lb_bgp_peer_address is defined
+
+# metal_lb_ip_range
+- name: >
+    Assert that metal_lb_ip_range (when string) contains <ipv4>-<ipv4>
+    and both ips start with routing4_prefix, skip any containing ':' (ipv6)
+  assert:
+    that:
+      - metal_lb_ip_range | regex_search('^{{ routing4_prefix }}\.[0-9]{1,3}-{{ routing4_prefix }}\.[0-9]{1,3}$|.*:.*')
+    fail_msg: >
+      metal_lb_ip_range <{{ metal_lb_ip_range }}> has one or more ipv4s
+      that don't start with the routing prefix <{{ routing4_prefix }}>
+  when: metal_lb_ip_range is string
+
+- name: Assert that metal_lb_ip_ranges (when list) has only strings that match the regexes in the task above
+  assert:
+    that:
+      - >
+        ( metal_lb_ip_range
+        | select('match', '^{{ routing4_prefix }}\.[0-9]{1,3}-{{ routing4_prefix }}\.[0-9]{1,3}$|.*:.*')
+        | list | length
+        )
+        ==
+        (metal_lb_ip_range
+        | length
+        )
+    fail_msg: >
+      metal_lb_ip_range <{{ metal_lb_ip_range }}> has one or more values with ipv4s
+      that don't start with the routing prefix <{{ routing4_prefix }}>
+  when: metal_lb_ip_range is not string and metal_lb_ip_range is not mapping and metal_lb_ip_range is iterable
+
+# apiserver_endpoint
+- name: Assert that apiserver_endpoint is not in metal_lb_ip_range (when string) using network_in_usable
+  # For /<mask> ranges
+  assert:
+    that:
+      - not ( metal_lb_ip_range | ansible.utils.network_in_usable( apiserver_endpoint ))
+    fail_msg: "apiserver_endpoint {{ apiserver_endpoint }} cannot be in the metal_lb_ip_range {{ metal_lb_ip_range }}"
+    success_msg: >
+      apiserver_endpoint {{ apiserver_endpoint }} is *probably* not in the metal_lb_ip_range {{ metal_lb_ip_range }}
+  when: metal_lb_ip_range is string
+
+- name: >
+    Assert that apiserver_endpoint is not in metal_lb_ip_range (when string)
+    using custom filter netaddr_ip_in_dash_range
+  # For <ip>-<ip> ranges. Not sure this works for ipv6
+  assert:
+    that:
+      - not (apiserver_endpoint | netaddr_ip_in_dash_range(metal_lb_ip_range))
+    fail_msg: "apiserver_endpoint {{ apiserver_endpoint }} cannot be in the metal_lb_ip_range {{ metal_lb_ip_range }}"
+    success_msg: >
+      apiserver_endpoint {{ apiserver_endpoint }} is *probably* not in the metal_lb_ip_range {{ metal_lb_ip_range }}
+  when: metal_lb_ip_range is string
+  # *probably* in the success_msg sections of the prior two tasks because not all cases may work
+
+- name: Assert that apiserver_endpoint is not in metal_lb_ip_ranges (when list) using logic of the task above
+  assert:
+    that:
+      - not (apiserver_endpoint | netaddr_ip_in_dash_range(item))
+    # this probably fails on an ipv6 range ''::1-::2', and on <ipv4or6>/<mask> ranges
+    fail_msg: "apiserver_endpoint {{ apiserver_endpoint }} cannot be in the metal_lb_ip_range item {{ item }}"
+    success_msg: >
+      apiserver_endpoint {{ apiserver_endpoint }} is *probably* not in the metal_lb_ip_range item {{ item }}
+  loop: "{{ metal_lb_ip_range | list }}"
+  when: metal_lb_ip_range is not string   and metal_lb_ip_range is not mapping   and metal_lb_ip_range is iterable
+  # these (when string, when list tasks) smell funny.
+  # It seems there should be a way in one task to handle an object that is a string or a list
+  #    and loop on {{ [metal_lb_ip_range] }}  or {{ metal_lb_ip_range }} respectively
+  # when it is a string it skips each char :(
+  # I tried the select pattern like in the task
+  #    'assert that metal_lb_ip_ranges (when list) has only strings that match the regexes in the task above'
+  #  but it didn't work
+
+- name: Assert that apiserver_endpoint, is ipv4 and starts with routing4_prefix, or is ipv6 and is in routing6_cidr
+  assert:
+    that:
+      - apiserver_endpoint is defined
+      - apiserver_endpoint is not none
+      - >-
+        ( apiserver_endpoint | ansible.utils.ipv4 and apiserver_endpoint.startswith(routing4_prefix) )
+        or
+        ( apiserver_endpoint | ansible.utils.ipv6 and apiserver_endpoint | ansible.utils.ipaddr(routing6_cidr))
+    fail_msg: >
+      The fact 'apiserver_endpoint' <{{ apiserver_endpoint }}>
+      doesn't start with the routing prefix <{{ routing4_prefix }}>
+      or is not in the routing6_cidr <{{ routing6_cidr }}>

site.yml

@@ -1,4 +1,7 @@
 ---
+- name: Import verify_config playbook
+  ansible.builtin.import_playbook: verify_config.yml
+
 - name: Prepare Proxmox cluster
   hosts: proxmox
   gather_facts: true
@@ -46,14 +49,3 @@
   roles:
     - role: k3s_server_post
       become: true
-
-- name: Storing kubeconfig in the playbook directory
-  hosts: master
-  environment: "{{ proxy_env | default({}) }}"
-  tasks:
-    - name: Copying kubeconfig from {{ hostvars[groups[group_name_master | default('master')][0]]['ansible_hostname'] }}
-      ansible.builtin.fetch:
-        src: "{{ ansible_user_dir }}/.kube/config"
-        dest: ./kubeconfig
-        flat: true
-      when: ansible_hostname == hostvars[groups[group_name_master | default('master')][0]]['ansible_hostname']

verify_config.yml

@@ -0,0 +1,14 @@
+---
+- name: Gather config for verify
+  hosts: all
+  gather_facts: true
+  roles:
+    - role: verify_config_gather
+      when: verify_config is not defined or verify_config
+
+- name: Verify config
+  hosts: localhost
+  gather_facts: false
+  roles:
+    - role: verify_config
+      when: verify_config is not defined or verify_config