17 files changed, 101 insertions, 180 deletions
diff --git a/ansible/README.md b/ansible/README.md
index 0d0c607..db8d960 100644
--- a/ansible/README.md
+++ b/ansible/README.md
@@ -1,52 +1,15 @@
-
-## Provisionning
-
- 1. Need a public IP address
- 2. Deploy Debian sid/buster
- 3. Add a DNS entry like xxxx.machine.deuxfleurs.fr A 0.0.0.0 in Cloudflare + Havelock
- 4. Setup the fqdn in /etc/hosts (127.0.1.1 xxxx.machine.deuxfleurs.fr)
- 5. Switch the SSH port to the port 110
- 6. Add the server to the ./production file 
- 7. Reboot machine
- 8. Deploy Ansible
- 9. Check that everything works as intended
- 10. Update NS 1.cluster.deuxfleurs.fr
-
-## Useful commands
-
-Show every variables collected by Ansible for a given host:
-
-```
-ansible -i production villequin.machine.deuxfleurs.fr -m setup
-```
-
-Run playbook for only one host:
-
-```
-ansible-playbook -i production --limit villequin.machine.deuxfleurs.fr site.yml
-```
-
-Dump hostvars:
-
-```
-ansible -m debug villequin.machine.deuxfleurs.fr -i ./production -a "var=hostvars"
-```
-
-Deploy only one tag:
-
-```
-ansible-playbook -i production site.yml --tags "container"
-```
-
-Redeploy everything:
-
-```
-ansible-playbook -i production site.yml
-```
-
-Upgrade packages and force overwirte to fix bad packing done by GlusterFS:
-
-```
-apt-get -o Dpkg::Options::="--force-overwrite"  dist-upgrade -y
-```
+# ANSIBLE
+
+## How to proceed
+
+For each machine, **one by one** do:
+  - Check that cluster is healthy
+    - `sudo gluster peer status`
+    - `sudo gluster volume status all` (check Online Col, only `Y` must appear)
+    - Check that Nomad is healthy
+    - Check that Consul is healthy
+    - Check that Postgres is healthy
+  - Run `ansible-playbook -i production --limit <machine> site.yml`
+  - Reboot
+  - Check that cluster is healthy
 
diff --git a/ansible/README.more.md b/ansible/README.more.md
new file mode 100644
index 0000000..0d0c607
--- /dev/null
+++ b/ansible/README.more.md
@@ -0,0 +1,52 @@
+
+## Provisionning
+
+ 1. Need a public IP address
+ 2. Deploy Debian sid/buster
+ 3. Add a DNS entry like xxxx.machine.deuxfleurs.fr A 0.0.0.0 in Cloudflare + Havelock
+ 4. Setup the fqdn in /etc/hosts (127.0.1.1 xxxx.machine.deuxfleurs.fr)
+ 5. Switch the SSH port to the port 110
+ 6. Add the server to the ./production file 
+ 7. Reboot machine
+ 8. Deploy Ansible
+ 9. Check that everything works as intended
+ 10. Update NS 1.cluster.deuxfleurs.fr
+
+## Useful commands
+
+Show every variables collected by Ansible for a given host:
+
+```
+ansible -i production villequin.machine.deuxfleurs.fr -m setup
+```
+
+Run playbook for only one host:
+
+```
+ansible-playbook -i production --limit villequin.machine.deuxfleurs.fr site.yml
+```
+
+Dump hostvars:
+
+```
+ansible -m debug villequin.machine.deuxfleurs.fr -i ./production -a "var=hostvars"
+```
+
+Deploy only one tag:
+
+```
+ansible-playbook -i production site.yml --tags "container"
+```
+
+Redeploy everything:
+
+```
+ansible-playbook -i production site.yml
+```
+
+Upgrade packages and force overwirte to fix bad packing done by GlusterFS:
+
+```
+apt-get -o Dpkg::Options::="--force-overwrite"  dist-upgrade -y
+```
+
diff --git a/ansible/cluster_nodes.yml b/ansible/cluster_nodes.yml
index df73e79..ea58630 100644
--- a/ansible/cluster_nodes.yml
+++ b/ansible/cluster_nodes.yml
@@ -11,17 +11,8 @@
       tags: kv
     - role: nomad
       tags: orchestrator
-
-# UNSAFE!! This section is disabled by default, to run it the flags -t net should be added
-# to the ansible playbook command line.
-# Reason: when rules.{v4,v6} are changed, the whole iptables configuration is reloaded.
-# This creates issues with Docker, which injects its own configuration in iptables when it starts.
-# In practice, most (all?) containers will break if rules.{v4,v6} are changed,
-# and docker will have to be restared.
-- hosts: cluster_nodes
-  roles:
-   - role: network
-     tags: [ net, never ]
+    - role: network
+      tags: net
 
 # UNSAFE!! This section configures glusterfs. Once done, don't run it ever again as it may break stuff.
 #   - role: storage
diff --git a/ansible/production b/ansible/production
index 6266502..c8f08f2 100644
--- a/ansible/production
+++ b/ansible/production
@@ -1,4 +1,4 @@
 [cluster_nodes]
-veterini ansible_host=fbx-rennes2.machine.deuxfleurs.fr ansible_port=110 ansible_user=root public_ip=192.168.1.2 private_ip=192.168.1.2 interface=eno1
-silicareux ansible_host=fbx-rennes2.machine.deuxfleurs.fr ansible_port=111 ansible_user=root public_ip=192.168.1.3 private_ip=192.168.1.3 interface=eno1
-wonse ansible_host=fbx-rennes2.machine.deuxfleurs.fr ansible_port=112 ansible_user=root public_ip=192.168.1.4 private_ip=192.168.1.4 interface=eno1
+veterini ansible_host=fbx-rennes2.machine.deuxfleurs.fr ansible_port=110 ansible_user=root public_ip=192.168.1.2 private_ip=192.168.1.2 interface=eno1 dns_server=80.67.169.40
+silicareux ansible_host=fbx-rennes2.machine.deuxfleurs.fr ansible_port=111 ansible_user=root public_ip=192.168.1.3 private_ip=192.168.1.3 interface=eno1 dns_server=80.67.169.40
+wonse ansible_host=fbx-rennes2.machine.deuxfleurs.fr ansible_port=112 ansible_user=root public_ip=192.168.1.4 private_ip=192.168.1.4 interface=eno1 dns_server=80.67.169.40
diff --git a/ansible/roles/common/tasks/main.yml b/ansible/roles/common/tasks/main.yml
index b4d00bb..0b15790 100644
--- a/ansible/roles/common/tasks/main.yml
+++ b/ansible/roles/common/tasks/main.yml
@@ -12,6 +12,12 @@
     autoclean: yes
     autoremove: yes
 
+- name: "Remove base tools"
+  apt:
+    name:
+      - systemd-resolved
+    state: absent
+
 - name: "Install base tools"
   apt:
     name:
diff --git a/ansible/roles/consul/handlers/main.yml b/ansible/roles/consul/handlers/main.yml
deleted file mode 100644
index e8cd4a4..0000000
--- a/ansible/roles/consul/handlers/main.yml
+++ /dev/null
@@ -1,4 +0,0 @@
----
-
-- name: restart consul
-  service: name=consul state=restarted
diff --git a/ansible/roles/consul/tasks/main.yml b/ansible/roles/consul/tasks/main.yml
index bb1d9ef..2b77080 100644
--- a/ansible/roles/consul/tasks/main.yml
+++ b/ansible/roles/consul/tasks/main.yml
@@ -1,16 +1,6 @@
 - name: "Set consul version"
   set_fact:
-    consul_version: 1.7.4
-
-- name: "Download and install Consul for armv7l"
-  unarchive:
-    src: "https://releases.hashicorp.com/consul/{{ consul_version }}/consul_{{ consul_version }}_linux_arm.zip"
-    dest: /usr/local/bin
-    remote_src: yes
-  when:
-    - "ansible_architecture == 'armv7l'"
-  notify:
-    - restart consul
+    consul_version: 1.8.0
 
 - name: "Download and install Consul for x86_64"
   unarchive:
@@ -19,31 +9,18 @@
     remote_src: yes
   when:
     - "ansible_architecture == 'x86_64'"
-  notify:
-    - restart consul
-
-- name: "Download and install Consul for arm64"
-  unarchive:
-    src: "https://releases.hashicorp.com/consul/{{ consul_version }}/consul_{{ consul_version }}_linux_arm64.zip"
-    dest: /usr/local/bin
-    remote_src: yes
-  when:
-    - "ansible_architecture == 'aarch64'"
-  notify:
-    - restart consul
 
 - name: "Create consul configuration directory"
   file: path=/etc/consul/ state=directory
 
 - name: "Deploy consul configuration"
   template: src=consul.json.j2 dest=/etc/consul/consul.json
-  notify:
-    - restart consul
 
 - name: "Deploy consul systemd service"
   copy: src=consul.service dest=/etc/systemd/system/consul.service
-  notify:
-    - restart consul
 
 - name: "Enable consul systemd service at boot"
   service: name=consul state=started enabled=yes daemon_reload=yes
+
+- name: "Deploy resolv.conf to use Consul"
+  template: src=resolv.conf.j2 dest=/etc/resolv.conf
diff --git a/ansible/roles/consul/templates/consul.json.j2 b/ansible/roles/consul/templates/consul.json.j2
index d1bd2d8..b6c86aa 100644
--- a/ansible/roles/consul/templates/consul.json.j2
+++ b/ansible/roles/consul/templates/consul.json.j2
@@ -17,6 +17,9 @@
   "ports": {
      "dns": 53
   },
+  "recursors": [
+    "{{ dns_server }}"
+  ],
   "encrypt": "{{ consul_gossip_encrypt }}",
   "domain": "2.cluster.deuxfleurs.fr",
   "performance": {
diff --git a/ansible/roles/consul/templates/resolv.conf.j2 b/ansible/roles/consul/templates/resolv.conf.j2
new file mode 100644
index 0000000..2404034
--- /dev/null
+++ b/ansible/roles/consul/templates/resolv.conf.j2
@@ -0,0 +1,2 @@
+nameserver {{ private_ip }}
+nameserver {{ dns_server }}
diff --git a/ansible/roles/network/files/nsswitch.conf b/ansible/roles/network/files/nsswitch.conf
index f4c3149..a84e024 100644
--- a/ansible/roles/network/files/nsswitch.conf
+++ b/ansible/roles/network/files/nsswitch.conf
@@ -9,8 +9,7 @@ group:          files systemd
 shadow:         files
 gshadow:        files
 
-#hosts:          files dns
-hosts:          files mymachines resolve [!UNAVAIL=return] dns myhostname
+hosts:          files dns
 networks:       files
 
 protocols:      db files
diff --git a/ansible/roles/network/files/systemd-resolve-no-listen.conf b/ansible/roles/network/files/systemd-resolve-no-listen.conf
deleted file mode 100644
index 6e95967..0000000
--- a/ansible/roles/network/files/systemd-resolve-no-listen.conf
+++ /dev/null
@@ -1,2 +0,0 @@
-[Resolve]
-DNSStubListener=no
diff --git a/ansible/roles/network/handlers/main.yml b/ansible/roles/network/handlers/main.yml
deleted file mode 100644
index 3454894..0000000
--- a/ansible/roles/network/handlers/main.yml
+++ /dev/null
@@ -1,12 +0,0 @@
----
-- name: reload iptables
-  shell: iptables-restore < /etc/iptables/rules.v4 && systemctl restart docker && ifdown nomad1 || true && ifup nomad1 || true
-
-- name: reload ip6tables
-  shell: ip6tables-restore < /etc/iptables/rules.v6
-
-- name: reload nomad interface
-  shell: ifdown nomad1 || true ; ifup nomad1
-
-- name: reload systemd-resolved
-  service: name=systemd-resolved state=restarted 
diff --git a/ansible/roles/network/tasks/main.yml b/ansible/roles/network/tasks/main.yml
index 7f95b0f..2087765 100644
--- a/ansible/roles/network/tasks/main.yml
+++ b/ansible/roles/network/tasks/main.yml
@@ -1,42 +1,23 @@
-- name: "Add dummy interface to handle Nomad NAT restriction nomad#2770"
-  template: src=nomad-interface.j2 dest=/etc/network/interfaces.d/nomad.cfg
-  when: public_ip != private_ip
-  notify:
-    - reload nomad interface
-
 - name: "Deploy iptablesv4 configuration"
   template: src=rules.v4.j2 dest=/etc/iptables/rules.v4
-  notify:
-    - reload iptables
 
 - name: "Deploy iptablesv6 configuration"
   copy: src=rules.v6 dest=/etc/iptables/rules.v6
-  notify:
-    - reload ip6tables
 
 - name: "Activate IP forwarding"
   sysctl:
     name: net.ipv4.ip_forward
-    value: 1
+    value: "1"
     sysctl_set: yes
 
-- name: "Create systemd-resolved override directory"
-  file: path=/etc/systemd/resolved.conf.d/ state=directory
-
-- name: "Prevent systemd-resolved from listening on port 53 (DNS)"
-  copy: src=systemd-resolve-no-listen.conf dest=/etc/systemd/resolved.conf.d/systemd-resolve-no-listen.conf
-  notify: reload systemd-resolved
+# These two lines are used to undo previous config, remove them once it is done
+- name: "Update nsswitch.conf to not use systemd-resolved"
+  copy: src=nsswitch.conf dest=/etc/nsswitch.conf
 
-- name: "Use systemd-resolved as a source for /etc/resolv.conf"
-  file:
-    src: "/run/systemd/resolve/resolv.conf"
-    dest: "/etc/resolv.conf"
-    state: link
-    force: yes
-  notify: reload systemd-resolved
+- name: "Disable systemd-resolved"
+  systemd:
+    name: systemd-resolved
+    state: stopped
+    enabled: false
 
-- name: "Update nsswitch.conf to use systemd-resolved"
-  copy: src=nsswitch.conf dest=/etc/nsswitch.conf
 
-- name: "Flush handlers"
-  meta: flush_handlers
diff --git a/ansible/roles/network/templates/nomad-interface.j2 b/ansible/roles/network/templates/nomad-interface.j2
deleted file mode 100644
index 74e9cd4..0000000
--- a/ansible/roles/network/templates/nomad-interface.j2
+++ /dev/null
@@ -1,8 +0,0 @@
-auto nomad1
-iface nomad1 inet manual
-    pre-up /sbin/ip link add nomad1 type dummy
-    up /sbin/ip addr add {{ public_ip }} dev nomad1
-    up /sbin/iptables -t nat -A PREROUTING -d {{ private_ip }}/32 -j NETMAP --to {{ public_ip }}/32
-    down /sbin/iptables -t nat -D PREROUTING -d {{ private_ip }}/32 -j NETMAP --to {{ public_ip }}/32
-    post-down /sbin/ip link del nomad1
-
diff --git a/ansible/roles/nomad/handlers/main.yml b/ansible/roles/nomad/handlers/main.yml
deleted file mode 100644
index 0274673..0000000
--- a/ansible/roles/nomad/handlers/main.yml
+++ /dev/null
@@ -1,5 +0,0 @@
----
-
-- name: restart nomad
-  service: name=nomad state=restarted
-
diff --git a/ansible/roles/nomad/tasks/main.yml b/ansible/roles/nomad/tasks/main.yml
index 0b7b65c..7c73362 100644
--- a/ansible/roles/nomad/tasks/main.yml
+++ b/ansible/roles/nomad/tasks/main.yml
@@ -1,16 +1,6 @@
 - name: "Set nomad version"
   set_fact:
-    nomad_version: 0.11.3
-
-- name: "Download and install Nomad for armv7l"
-  unarchive:
-    src: "https://releases.hashicorp.com/nomad/{{ nomad_version }}/nomad_{{ nomad_version }}_linux_arm.zip"
-    dest: /usr/local/bin
-    remote_src: yes
-  when:
-    - "ansible_architecture == 'armv7l'"
-  notify:
-    - restart nomad
+    nomad_version: 0.12.0-beta2
 
 - name: "Download and install Nomad for x86_64"
   unarchive:
@@ -19,31 +9,15 @@
     remote_src: yes
   when:
     - "ansible_architecture == 'x86_64'"
-  notify:
-    - restart nomad
-
-- name: "Download and install Nomad for arm64"
-  unarchive:
-    src: "https://releases.hashicorp.com/nomad/{{ nomad_version }}/nomad_{{ nomad_version }}_linux_arm64.zip"
-    dest: /usr/local/bin
-    remote_src: yes
-  when:
-    - "ansible_architecture == 'aarch64'"
-  notify:
-    - restart nomad
 
 - name: "Create Nomad configuration directory"
   file: path=/etc/nomad/ state=directory
 
 - name: "Deploy Nomad configuration"
   template: src=nomad.hcl.j2 dest=/etc/nomad/nomad.hcl
-  notify:
-    - restart nomad
 
 - name: "Deploy Nomad systemd service"
   copy: src=nomad.service dest=/etc/systemd/system/nomad.service
-  notify:
-    - restart nomad
 
 - name: "Enable Nomad systemd service at boot"
   service: name=nomad state=started enabled=yes daemon_reload=yes
diff --git a/ansible/roles/nomad/templates/nomad.hcl.j2 b/ansible/roles/nomad/templates/nomad.hcl.j2
index 8107410..b0be6a8 100644
--- a/ansible/roles/nomad/templates/nomad.hcl.j2
+++ b/ansible/roles/nomad/templates/nomad.hcl.j2
@@ -26,5 +26,9 @@ client {
   #cpu_total_compute = 4000
   servers = ["127.0.0.1:4648"]
   network_interface = "{{ interface }}"
+  options {
+    docker.privileged.enabled = "true"
+    docker.volumes.enabled = "true"
+  } 
 }