Post

Containers series IV: Docker Swarm

Posted Updated
By Jose Castillo Lema
6 min read
Containers series IV: Docker Swarm

Serie de artigos sobre containers (Docker, Docker Compose, Docker Swarm e Kubernetes).

GitHub watchers GitHub stars GitHub forks

Orquestrando containers

Docker Swarm permite orquestrar containers em um cluster formado por vários servidores. Desta forma conseguimos garantir as seguintes propriedades nos containers gerenciados pelo orquestrador:

  • tolerância a falhas: se um dos servidores do cluster cair, o container automaticamente será iniciado em outro servidor do cluster
  • alta disponibilidade: várias réplicas de cada container podem ser executadas em vários servidores do cluster
  • escalabilidade: o número de réplicas de cada container pode ser aumentado a qualquer momento em funçao da demanda

Vamos trabalhar com duas máquinas virtuais (T1 e T2). Se elas tiver rodando na nuvem, devem pertencer ao mesmo security group (ou ter um security group em común).

Requisitos

  • 2 VMs Ubuntu 18.04 ou Amazon Linux com Docker instalado com no mínimo 2 GB de memória cada (na AWS usar flavor t2.small ou maior)
  • As duas VMs devem pertencer ao mesmo security group (ou ter um security group em común). Se nao for suficiente, liberar a porta 2377 no security group da VM que for ser manager do cluster. Pode usar o comando telnet para testar a comunicaçao (telnet $ip $porta). Um exemplo de security group para o manager:

Criação do cluster

  1. [T1] Inicialização do manager:

    1
2
3
4
5
6
7
8

     $ docker swarm init
 Swarm initialized: current node (1y4bix4oby6nq2jxx5ft4rhd0) is now a manager.

 To add a worker to this swarm, run the following command:

     docker swarm join --token SWMTKN-1-5it2k13vtptja3tl2xpgjywr856a4r7siuve20r2ev9h98gfrj-498uqdu6x8o74b816orz6s5gn 172.31.47.198:2377

 To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.

  2. [T2] Na segunda maquina virtual, inicialização do worker:

    1
2

     $ docker swarm join --token SWMTKN-1-5it2k13vtptja3tl2xpgjywr856a4r7siuve20r2ev9h98gfrj-498uqdu6x8o74b816orz6s5gn 172.31.47.198:2377
 This node joined a swarm as a worker.

  3. [T1] Listar os servidores que fazem parte do cluster Docker Swarm desde o manager:

    1
2
3
4

     $ docker node ls
 ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS      ENGINE VERSION
 4hx9zm7docinufxy1mm62yynm     ip-172-31-32-255    Ready               Active                                  19.03.6
 1y4bix4oby6nq2jxx5ft4rhd0 *   ip-172-31-47-198    Ready               Active              Leader              19.03.6

  4. [T2] Desde o worker não é possível executar nenhum comando do Docker Swarm, p.ex.:

    1
2

     $ docker node ls
 Error response from daemon: This node is not a swarm manager. Worker nodes can't be used to view or modify cluster state. Please run this command on a manager node or promote the current node to a manager.

Uso

  1. [T1] Navegar até a pasta fiap/bdt/microservices/swarm/v1 deste repositório git:

    1
2
3
4
5

     $ cd fiap/aso/microservices/swarm/v1
 $ pwd
 /home/ubuntu/fiap/aso/microservices/swarm/v1
 $ ls
 docker-compose.yaml
  2. [T1] Mostrar o conteúdo do arquivo docker-compose.yaml. São definidos dois serviços:
    • api: a API escrita em Python, que tem dependência (consulta) o serviço mysql, com 3 réplicas
    • mysql: o servidor MySQL, com mapeamento de portas (porta 3306), persistência de dados (pasta /var/lib/mysql) e algumas variáveis de ambiente, com 1 réplica
    1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41

     $ cat docker-compose.yaml 
 version: '3'

 services:

   api:
     image: josecastillolema/api
     ports:
       - "3000:5000"
     depends_on:
       - mysql
     deploy:
       replicas: 3
       # update_config:
       #   parallelism: 1
       #   delay: 10is
       mode: replicated
       restart_policy:
         condition: on-failure

   mysql:
     image: josecastillolema/mysql
     ports:
       - "3306:3306"
     volumes:
       - db-data:/var/lib/mysql
     environment:
       MYSQL_USER: root
       MYSQL_DATABASE: fiapdb
       MYSQL_ROOT_PASSWORD: senhaFiap
     deploy:
       replicas: 1
       # resources:
       #   limits:
       #     cpus: "0.1"
       #     memory: 50M
       restart_policy:
         condition: on-failure

 volumes:
   db-data:

  3. [T1] Criar o stack definido no arquivo docker-compose.yml:

    1
2
3
4

     $ docker stack deploy -c docker-compose.yaml stackFiap
 Creating network stackFiap_default
 Creating service stackFiap_api
 Creating service stackFiap_mysql

  4. [T1] Conferir que o stack foi criado corretamente:

    1
2
3

     $ docker stack ls
 NAME                SERVICES            ORCHESTRATOR
 stackFiap           2                   Swarm

  5. [T1] Conferir que os serviços foram criados corretamente:

    1
2
3
4

     $ docker service ls
 ID                  NAME                MODE                REPLICAS            IMAGE                           PORTS
 vlh0cjv5nd65        stackFiap_api       replicated          3/3                 josecastillolema/api:latest     *:3000->5000/tcp
 tmns8lwyrb9f        stackFiap_mysql     replicated          1/1                 josecastillolema/mysql:latest   *:3306->3306/tcp

  6. [T1] Testar a API:

    1
2

    $ curl localhost:3000
Benvido a API FIAP!

  7. [T1] Conferir quais containers foram criados na primeira máquina virtual (neste caso o banco de dados e uma instância da API):

    1
2
3
4

    $ docker ps
CONTAINER ID        IMAGE                           COMMAND                  CREATED             STATUS                   PORTS                 NAMES
44cfaf700b20        josecastillolema/mysql:latest   "docker-entrypoint.s…"   2 minutes ago       Up 2 minutes             3306/tcp, 33060/tcp   stackFiap_mysql.1.wr5ia70abxco05ypimayagrm0
2239ae187e86        josecastillolema/api:latest     "./api.py"               2 minutes ago       Up 2 minutes (healthy)   5000/tcp              stackFiap_api.1.qedwp50z4l5dhskg66txj91d7

  8. [T2] Conferir quais containers foram criados na segunda máquina virtual (neste caso duas instâncias da API):

    1
2
3
4

    $ docker ps
CONTAINER ID        IMAGE                         COMMAND             CREATED             STATUS                   PORTS               NAMES
48fdb1fb8cb7        josecastillolema/api:latest   "./api.py"          2 minutes ago       Up 2 minutes (healthy)   5000/tcp            stackFiap_api.3.rpl9ggawgeo7edietartr7ca0
3201f87fc015        josecastillolema/api:latest   "./api.py"          2 minutes ago       Up 2 minutes (healthy)   5000/tcp            stackFiap_api.2.rt2zby2s1s8igdzysd3gbmv9w

  9. [T1] Aumentar o numero de replicas da API (scale out):

    1
2
3
4
5
6
7
8
9

    $ docker service scale stackFiap_api=5
stackFiap_api scaled to 5
overall progress: 5 out of 5 tasks 
1/5: running   [==================================================>] 
2/5: running   [==================================================>] 
3/5: running   [==================================================>] 
4/5: running   [==================================================>] 
5/5: running   [==================================================>]  
verify: Service converged

  10. [T1] Confirmar o novo numero de replicas:

    1
2
3
4

    $ docker service ls
ID                  NAME                MODE                REPLICAS            IMAGE                           PORTS
vlh0cjv5nd65        stackFiap_api       replicated          5/5                 josecastillolema/api:latest     *:3000->5000/tcp
tmns8lwyrb9f        stackFiap_mysql     replicated          1/1                 josecastillolema/mysql:latest   *:3306->3306/tcp

  11. [T1] Diminuir o numero de replicas da API (scale in):

    1
2
3
4
5
6
7
8

    $ docker service scale stackFiap_api=4
stackFiap_api scaled to 4
overall progress: 4 out of 4 tasks 
1/4: running   [==================================================>] 
2/4: running   [==================================================>] 
3/4: running   [==================================================>] 
4/4: running   [==================================================>] 
verify: Service converged

  12. [T1] Vamos desligar o worker (servidor T2). Antes disso, conferir os containers que estão rodando no manager (neste caso, o banco de dados e 2 replicas da API):

    1
2
3
4
5

    $ docker ps
CONTAINER ID        IMAGE                           COMMAND                  CREATED             STATUS                    PORTS                 NAMES
325de1e84b51        josecastillolema/api:latest     "./api.py"               4 minutes ago       Up 4 minutes (healthy)    5000/tcp              stackFiap_api.4.vciwnzaxakavsj2t3p2tlaekc
44cfaf700b20        josecastillolema/mysql:latest   "docker-entrypoint.s…"   11 minutes ago      Up 11 minutes             3306/tcp, 33060/tcp   stackFiap_mysql.1.wr5ia70abxco05ypimayagrm0
2239ae187e86        josecastillolema/api:latest     "./api.py"               11 minutes ago      Up 11 minutes (healthy)   5000/tcp              stackFiap_api.1.qedwp50z4l5dhskg66txj91d7

  13. [T2] Desligar o worker:

    1
2
3

    $ sudo shutdown -h now
Connection to ec2-3-85-40-189.compute-1.amazonaws.com closed by remote host.
Connection to ec2-3-85-40-189.compute-1.amazonaws.com closed.

  14. [T1] Apos uns instantes, confirmar que o worker aparece como down:

    1
2
3
4

    $ docker node ls
ID                            HOSTNAME            STATUS              AVAILABILITY        MANAGER STATUS      ENGINE VERSION
4hx9zm7docinufxy1mm62yynm     ip-172-31-32-255    Down                Active                                  19.03.6
1y4bix4oby6nq2jxx5ft4rhd0 *   ip-172-31-47-198    Ready               Active              Leader              19.03.6

  15. [T1] Confirmar que os containers que estavam rodando no worker (servidor T2), foram recriados no manager (neste caso, 2 réplicas da API):

    1
2
3
4
5
6
7

    $ docker ps
CONTAINER ID        IMAGE                           COMMAND                  CREATED              STATUS                        PORTS                 NAMES
2bbcefe6fc28        josecastillolema/api:latest     "./api.py"               About a minute ago   Up About a minute (healthy)   5000/tcp              stackFiap_api.3.15hrsir7ac5a2eun79wbp6ftf
41067e32567d        josecastillolema/api:latest     "./api.py"               About a minute ago   Up About a minute (healthy)   5000/tcp              stackFiap_api.7.ldetjf1ykn7kyaokadcaod0di
325de1e84b51        josecastillolema/api:latest     "./api.py"               7 minutes ago        Up 7 minutes (healthy)        5000/tcp              stackFiap_api.4.vciwnzaxakavsj2t3p2tlaekc
44cfaf700b20        josecastillolema/mysql:latest   "docker-entrypoint.s…"   14 minutes ago       Up 14 minutes                 3306/tcp, 33060/tcp   stackFiap_mysql.1.wr5ia70abxco05ypimayagrm0
2239ae187e86        josecastillolema/api:latest     "./api.py"               15 minutes ago       Up 15 minutes (healthy)       5000/tcp              stackFiap_api.1.qedwp50z4l5dhskg66txj91d7

Clean-up

  1. [T1] Remover o stack:

    1
2
3
4

    $ docker stack rm stackFiap
Removing service stackFiap_api
Removing service stackFiap_mysql
Removing network stackFiap_default

GitHub followers

classes fiap iac openshift pt redhat series
This post is licensed under CC BY 4.0 by the author.