MODOP – Installation d’un Cluster etcd

Aucun commentaire

Inventaire

  • Node-etcd01
    • IP : 192.168.1.40/24
    • OS : Rocky Linux 8.5
    • RAM : 2Go
    • CPU : 1
    • Disk : 32Go
  • Node-etcd02
    • IP : 192.168.1.41/24
    • OS : Rocky Linux 8.5
    • RAM : 2Go
    • CPU : 1
    • Disk : 32Go
  • Node-etcd03
    • IP : 192.168.1.42/24
    • OS : Rocky Linux 8.5
    • RAM : 2Go
    • CPU : 1
    • Disk : 32Go

Pool des machines du cluster ETCD

1. Prérequis des machines du Pool ETCD

Mise à jour de la machine (3 nodes)

[root@node-etcd0x ~]# dnf update -y

Ajout des Hosts (pas de DNS)

[root@node-etcd01 ~]# echo "192.168.1.40 node-etcd01 " >> /etc/hosts
[root@node-etcd01 ~]# echo "192.168.1.41 node-etcd02 " >> /etc/hosts
[root@node-etcd01 ~]# echo "192.168.1.42 node-etcd03 " >> /etc/hosts
[root@node-etcd01 ~]# scp /etc/hosts root@node-etcd02:/etc/hosts
[root@node-etcd01 ~]# scp /etc/hosts root@node-etcd03:/etc/hosts
[root@node-etcd01 ~]# ping -c 2 node-etcd01
[root@node-etcd01 ~]# ping -c 2 node-etcd02
[root@node-etcd01 ~]# ping -c 2 node-etcd03

Désactiver SELinux (3 nodes)

[root@node-etcd0x ~]# getenforce
Enforcing

[root@node-etcd0x ~]# setenforce 0
[root@node-etcd0x ~]# getenforce
Permissive

[root@node-etcd0x ~]# sed -i 's/SELINUX=enforcing/SELINUX=disabled/g' /etc/selinux/config

Synchroniser DateTime sur le fuseau de PARIS (3 nodes)

[root@node-etcd0x ~]# timedatectl

[root@node-etcd0x ~]# timedatectl set-timezone Europe/Paris
[root@node-etcd0x ~]# timedatectl

Installation des middlewares utiles (3 nodes)

[root@node-etcd0x ~]# dnf install epel-release net-tools nmap wget tar -y

2. Récupération « latest » etcd (3 nodes)

[root@node-etcd0x ~]# echo "ETCD_RELEASE=$(curl -s https://api.github.com/repos/etcd-io/etcd/releases/latest|grep tag_name | cut -d '"' -f 4)" >> ~/.bashrc
[root@node-etcd0x ~]# source ~/.bashrc
[root@node-etcd0x ~]# echo $ETCD_RELEASE
v3.5.7

[root@node0x-psql tmp]# cd /tmp && wget https://github.com/etcd-io/etcd/releases/download/${ETCD_RELEASE}/etcd-${ETCD_RELEASE}-linux-amd64.tar.gz
[root@node-etcd0x tmp]# ls -al |grep etc
-rw-r--r--. 1 root root 18458320 20 janv. 11:23 etcd-v3.5.7-linux-amd64.tar.gz

3. Installation des binaires etcd (3 nodes)

[root@node-etcd0x tmp]# tar -xzvf etcd-${ETCD_RELEASE}-linux-amd64.tar.gz
[root@node-etcd0x tmp]# cd etcd-${ETCD_RELEASE}-linux-amd64/
[root@node-etcd0x etcd-v3.5.7-linux-amd64]# ls -al

[root@node-etcd0x etcd-v3.5.7-linux-amd64]# mv etcd* /usr/local/bin
[root@node-etcd0x etcd-v3.5.7-linux-amd64]# ls /usr/local/bin
etcd etcdctl etcdutl
[root@node-etcd0x etcd-v3.5.7-linux-amd64]# etcdutl version
[root@node-etcd0x etcd-v3.5.7-linux-amd64]# etcdctl version
[root@node-etcd0x etcd-v3.5.7-linux-amd64]# etcd --version

4. Installation/création du service etcd (3 nodes)

Création de la strucure etcd (3 nodes)

[root@node-etcd0x etcd-v3.5.7-linux-amd64]# cd ~
[root@node-etcd0x ~]# mkdir -p /var/lib/etcd/
[root@node-etcd0x ~]# mkdir /etc/etcd

Création droit user/group etcd (3 nodes)

[root@node-etcd0x ~]# groupadd --system etcd
[root@node-etcd0x ~]# useradd -s /sbin/nologin --system -g etcd etcd

Ajout des droits user/group etcd sur la structure (3 nodes)

[root@node-etcd0x ~]# chown -R etcd:etcd /var/lib/etcd/
[root@node-etcd0x ~]# chmod -R 0700 /var/lib/etcd

Ouverture des rules Firewall (3 nodes)

[root@node-etcd0x ~]# firewall-cmd --zone=public --permanent --add-port={2379,2380}/tcp
[root@node-etcd0x ~]# firewall-cmd --remove-service={cockpit,dhcpv6-client} --permanent
[root@node-etcd0x ~]# firewall-cmd --reload

[root@node-etcd0x ~]# firewall-cmd --list-port
2379/tcp 2380/tcp

5. Configuration du cluster etcd

Service etcd node-etcd01

[root@node-etcd01 ~]# systemctl stop etcd.service
[root@node-etcd01 ~]# vi /etc/systemd/system/etcd.service

[Unit]
Description=etcd key-value store
Documentation=https://github.com/etcd-io/etcd
After=network.target
[Service]
User=etcd
Type=notify
ExecStart=/usr/local/bin/etcd \
 --name node-etcd01 \
 --data-dir /var/lib/etcd/node-etcd01 \
 --initial-advertise-peer-urls http://192.168.1.40:2380 \
 --listen-peer-urls http://192.168.1.40:2380 \
 --listen-client-urls http://192.168.1.40:2379,http://127.0.0.1:2379 \
 --advertise-client-urls http://192.168.1.40:2379 \
 --initial-cluster-token clusterpsql \
 --initial-cluster node-etcd01=http://192.168.1.40:2380 \
 --initial-cluster-state new \
 --heartbeat-interval 1000 \
 --election-timeout 5000
Restart=always
RestartSec=10s
LimitNOFILE=40000

[Install]
WantedBy=multi-user.target
On recharge le service
[root@node-etcd01 ~]# systemctl daemon-reload

Service etcd node-etcd02

[root@node-etcd02 ~]# systemctl stop etcd.service
[root@node-etcd02 ~]# vi /etc/systemd/system/etcd.service

[Unit]
Description=etcd key-value store
Documentation=https://github.com/etcd-io/etcd
After=network.target
[Service]
User=etcd
Type=notify
ExecStart=/usr/local/bin/etcd \
--name node-etcd02 \
--data-dir /var/lib/etcd/node-etcd02 \
--initial-advertise-peer-urls http://192.168.1.41:2380 \
--listen-peer-urls http://192.168.1.41:2380 \
--listen-client-urls http://192.168.1.41:2379,http://127.0.0.1:2379 \
--advertise-client-urls http://192.168.1.41:2379 \
--initial-cluster-token clusterpsql \
--initial-cluster node-etcd01=http://192.168.1.40:2380,node-etcd02=http://192.168.1.41:2380\
--initial-cluster-state existing \
--heartbeat-interval 1000 \
--election-timeout 5000
Restart=always
RestartSec=10s
LimitNOFILE=40000

[Install]
WantedBy=multi-user.target
On recharge le service
[root@node-etcd02 ~]# systemctl daemon-reload

Service etcd node-etcd03

[root@node-etcd03 ~]# systemctl stop etcd.service
[root@node-etcd03 ~]# vi /etc/systemd/system/etcd.service

[Unit]
Description=etcd key-value store
Documentation=https://github.com/etcd-io/etcd
After=network.target

[Service]
User=etcd
Type=notify
ExecStart=/usr/local/bin/etcd \
--name node-etcd03 \
--data-dir /var/lib/etcd/node-etcd03 \
--initial-advertise-peer-urls http://192.168.1.42:2380 \
--listen-peer-urls http://192.168.1.42:2380 \
--listen-client-urls http://192.168.1.42:2379,http://127.0.0.1:2379 \
--advertise-client-urls http://192.168.1.42:2379 \
--initial-cluster-token clusterpsql \
--initial-cluster node-etcd01=http://192.168.1.40:2380,node-etcd02=http://192.168.1.41:2380,node-etcd03=http://192.168.1.42:2380 \
--initial-cluster-state existing \
--heartbeat-interval 1000 \
--election-timeout 5000
Restart=always
RestartSec=10s
LimitNOFILE=40000

[Install]

WantedBy=multi-user.target
On recharge le service
[root@node-etcd03 ~]# systemctl daemon-reload

6. Lancement du cluster etcd

Lancement etcd node-etcd01

[root@node-etcd01 ~]# rm -rf /var/lib/etcd/node-etcd01
[root@node-etcd01 ~]# systemctl start etcd.service
[root@node-etcd01 ~]# systemctl status etcd.service

[root@node-etcd01 ~]# ls -al /var/lib/etcd/node-etcd01
[root@node-etcd01 ~]# ls -al /var/lib/etcd/node-etcd01/member/

Ajout des membres au cluster sur node-etcd01
[root@node-etcd02 ~]# systemctl start --now etcd.service
[root@node-etcd01 ~]# etcdctl member add node-etcd02 --peer-urls=http://192.168.1.41:2380

[root@node-etcd03 ~]# systemctl start --now etcd.service
[root@node-etcd01 ~]# etcdctl member add node-etcd03 --peer-urls=http://192.168.1.42:2380

Démarrage/Création du Cluster ETCD 

[root@node-etcd01 ~]# systemctl start --now etcd
[root@node-etcd01 ~]# systemctl enable etcd.service
[root@node-etcd01 ~]# systemctl status etcd

Lancement etcd node-etcd02

[root@node-etcd02 ~]# systemctl start --now etcd
[root@node-etcd02 ~]# systemctl enable etcd.service
[root@node-etcd02 ~]# systemctl status etcd

Lancement etcd node-etcd03

[root@node-etcd03 ~]# systemctl start --now etcd
[root@node-etcd03 ~]# systemctl enable etcd.service
[root@node-etcd03 ~]# systemctl status etcd

7. Check du Cluster ETCD

Check « Status » des membres (node01)

[root@node-etcd01 ~]# etcdctl member list

[root@node-etcd01 ~]# etcdctl -w table member list

Check « Health » des membres (node01)

[root@node-etcd01 ~]# etcdctl endpoint health --endpoints=192.168.1.40:2380
[root@node-etcd01 ~]# etcdctl endpoint health --endpoints=192.168.1.41:2380
[root@node-etcd01 ~]# etcdctl endpoint health --endpoints=192.168.1.42:2380
[root@node-etcd01 ~]# etcdctl endpoint health

root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd01:2380 endpoint status
root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd02:2380 endpoint status
root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd02:2380 endpoint status

Check « Transactions » des membres (node01)

root@node-etcd01 ~]# etcdctl --endpoints=192.168.1.40:2380 put user1 chris
OK

root@node-etcd01 ~]# etcdctl --endpoints=192.168.1.41:2380 get user1
user1
chris

root@node-etcd01 ~]# etcdctl --endpoints=192.168.1.42:2380 get user1
user1
chris

root@node-etcd01 ~]# etcdctl --endpoints=192.168.1.40:2380 get user1
user1
chris

Check « Create/Save DBA» des membres (node01)

[root@node-etcd01 ~]# etcdctl --endpoints=192.168.1.40:2379 snapshot save node-etcd01.db

[root@node-etcd01 ~]# etcdctl --endpoints=192.168.1.41:2379 snapshot save node-etcd02.db

[root@node-etcd01 ~]# etcdctl --endpoints=192.168.1.42:2379 snapshot save node-etcd03.db

[root@node-etcd01 ~]# ls -al |grep node

[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd01:2380 endpoint status
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd02:2380 endpoint status
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd03:2380 endpoint status

[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=192.168.1.40:2380 snapshot status node-etcd01.db
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=192.168.1.40:2380 snapshot status node-etcd02.db
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=192.168.1.40:2380 snapshot status node-etcd03.db

[root@node-etcd01 ~]# etcdutl snapshot status node-etcd01.db
7a62a45b, 2, 9, 20 kB

[root@node-etcd01 ~]# etcdutl snapshot status node-etcd02.db
7a62a45b, 2, 9, 20 kB

[root@node-etcd01 ~]# etcdutl snapshot status node-etcd03.db
7a62a45b, 2, 9, 20 kB

Change « leader» des membres (node01)

[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node01-psql:2380 endpoint status
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node02-psql:2380 endpoint status
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node03-psql:2380 endpoint status

Change le leader du node01 vers le node02 (node01)

[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd02:2380 endpoint status
[root@node-etcd01 ~]# etcdctl --endpoints=node-etcd01:2380 move-leader a782ed7d378b33b3

Check nouveau leader node-etcd02 (node01)

[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd01:2380 endpoint status
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd02:2380 endpoint status
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd03:2380 endpoint status

[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 endpoint status

Check Ecriture/Lecture de données (node01)

Ecriture d’un texte (data)

[root@node-etcd01 ~]# etcdctl --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 put msg "Christian - Cluster ETCD"
OK
[root@node-etcd01 ~]# etcdctl --write-out=table --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 endpoint status

Lecture du texte (data)

[root@node-etcd01 ~]# etcdctl --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 get msg

[root@node-etcd02 ~]# etcdctl --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 get msg

[root@node-etcd03 ~]# etcdctl --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 get msg

Effacement du texte (data)

[root@node-etcd03 ~]# etcdctl --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 del msg
1
[root@node-etcd03 ~]# etcdctl --endpoints=node-etcd01:2380,node-etcd02:2380,node-etcd03:2380 get msg

Views: 4

Étiquettes

Laisser un commentaire

Votre adresse e-mail ne sera pas publiée. Les champs obligatoires sont indiqués avec *