learning docker

swarm

install docker with user data

1#!/bin/bash
2curl -fsSL https://get.docker.com -o get-docker.sh
3sh get-docker.sh
4sudo usermod -aG docker <username eg root>

swarm lab - part 01 setup the swarm

setup three droplets with docker installed as above.
on node-01 initiate swarm on a public ip address

1docker swarm init --advertise--addr <IP_ADDRESS>

add the node-02 as worker.

ssh on node-02 and ask it to join the swarm

1docker swarm join --token <token> <IP>

PS this command is printed for you after you do swarm init

update the role of node-02 to be a manager

on node-01 execute

1docker node update --role manager node-02

now if you do docker node ls you will see node-02 being reachable.

take the join token to add also the third node

1docker swarm join-token manager

paste on node-03 the join command

1docker swarm join --token <token> <IP_ADDRESS>

The manager status indicates that these nodes are managers.

swarm lab - part 02 install two services. a website and a database

create a network overlay

1docker network create --driver overlay mydrupal

create the database service and add it to the network we created

1docker service create --name psql --network mydrupal -e POSTGRES_PASSWORD=mypass postgres

runs on node-01

create the drupal service and add it to the same network

1docker service create --name drupal --network mydrupal -p 80:80 drupal

visit a droplet and see the service running

drupal setup must be the same as the service name so that they can communicate.
drupal complete

swarm lab - part 03 routing mesh

creating one more service. elasticsearch

1docker service create --name search --replicas 3 -p 9200:9200 elasticsearch:2

on node-01 use curl to see the loadbalancer working

1curl localhost:9200

swarm stacks

we will be using

1docker stack deploy

example

1version: "3"
2services:
3
4  redis:
5    image: redis:alpine
6    ports:
7      - "6379"
8    networks:
9      - frontend
10    deploy:
11      replicas: 1
12      update_config:
13        parallelism: 2
14        delay: 10s
15      restart_policy:
16        condition: on-failure
17  db:
18    image: postgres:9.4
19    volumes:
20      - db-data:/var/lib/postgresql/data
21    networks:
22      - backend
23    environment:
24      - POSTGRES_HOST_AUTH_METHOD=trust
25    deploy:
26      placement:
27        constraints: [node.role == manager]
28  vote:
29    image: bretfisher/examplevotingapp_vote
30    ports:
31      - 5000:80
32    networks:
33      - frontend
34    depends_on:
35      - redis
36    deploy:
37      replicas: 5
38      update_config:
39        parallelism: 2
40      restart_policy:
41        condition: on-failure
42  result:
43    image: bretfisher/examplevotingapp_result
44    ports:
45      - 5001:80
46    networks:
47      - backend
48    depends_on:
49      - db
50    deploy:
51      replicas: 1
52      update_config:
53        parallelism: 2
54        delay: 10s
55      restart_policy:
56        condition: on-failure
57
58  worker:
59    image: bretfisher/examplevotingapp_worker:java
60    networks:
61      - frontend
62      - backend
63    depends_on:
64      - db
65      - redis
66    deploy:
67      mode: replicated
68      replicas: 1
69      labels: [APP=VOTING]
70      restart_policy:
71        condition: on-failure
72        delay: 10s
73        max_attempts: 3
74        window: 120s
75      placement:
76        constraints: [node.role == manager]
77
78  visualizer:
79    image: dockersamples/visualizer
80    ports:
81      - "8080:8080"
82    stop_grace_period: 1m30s
83    volumes:
84      - "/var/run/docker.sock:/var/run/docker.sock"
85    deploy:
86      placement:
87        constraints: [node.role == manager]
88
89networks:
90  frontend:
91  backend:
92
93volumes:
94  db-data:

learning docker - day 5

swarm

install docker with user data

swarm lab - part 01 setup the swarm

swarm lab - part 02 install two services. a website and a database

swarm lab - part 03 routing mesh

swarm stacks

More articles from blog for tech

learning docker - day 4

learning docker - day 3