1, first look at two cluster examples: nginx cluster, Redis cluster

2. Then summarize

1. Symmetric clusters

Different service instances with equal functional status (each instance offers the same functional opportunity)

Positioning: data calculation

2. Asymmetric clusters

Different service instances with unequal functional status (each instance has a different opportunity to provide functionality)

Location: Data store

Docker cluster diagram to analyze, Docker inside the running container, if Docker down?

Docker container provides external access. If the number of visits is larger, the concurrency of a Docker is limited after all

Conclusion:

1. Single point of failure

2. Performance issues

So in order to solve these two problems, docker clusters appear

1. Node

2. Task concept – Introduced later

3. The concept of Service – introduced later

4. Stack concept – Introduced later

The docker cluster is maintained by assigning roles to nodes

Node is the alias of docker, just like, we go to work, everyone has a work number, the work number is our alias, so node is the alias of Docker, but node has two types, manager and work, so docker is divided into two types, just like, There are bosses and employees in a company. “Manager” is the boss and “work” is the employee. The boss is in charge and the employee is responsible for the work. Sometimes when the labor is not enough, the boss also comes to do things.

1. Node: docker host

2. Management Node (Manager)

Manage the Docker cluster

1. Cluster configuration

2. Container service management.

3. Load balancing

4. Cluster Management?

3. Work node (Worker)

1. Provide container services

conditions

1. More than two Docker hosts

2, docker swarm

steps

1. Install two docker hosts using Linux or centos7.0 or above

Docker 1.12 or later will be supported by default on the host

How to use Docker swarm?

Docker docker docker docker docker docker docker docker docker

​ docker swarm

2. Enter the docker swarm command

You can see a lot of help commands

3. Create a cluster management node

Docker swarm init –advertise-addr 123.57.164.54

4. Create a work node

​ docker swarm join –token SWMTKN-1-2bomg3tb4cs8f5g7oqrw34k1mql9i0tr2msn8leqyphgac5hct-db4iwy32sdr6m0bmqyhi47jyn 123.57.164.54:2377

5. View cluster messages, working nodes, and management nodes

​ docker info

If you think about the stack, in addition to the human, there are some managers in the company, what else is the general manager, so stack is the general manager

It’s like a department manager in a company

conditions

1. Rmcore mirroring

2. Nginx mirroring

3, the docker – compose. Yml

4, the stack

1. View rMCore images and Nginx images

Use rMCore images for cluster service deployment

How to use Docker Stack?

Docker command output, you will see a service management command

​ docker stack

2. Enter the Docker stack command

There are a lot of help commands

3. Create stack (collection of services)

Create a stack directory (create a stack directory under current directory /root)

​ mkdir stack

2. Configure the YML file

Add configuration to existing docker.compose. Yml

Version: ‘3.4’

Gloab replicas: 2 replicas

3. Create stack

​ docker stack deploy -c docker-compose.yml rmstack

-c specifies the docker-comemage. yml file

Rmcorestack Specifies the stack id (name)

4. View the Stack list

​ docker stack ls

5. View the stack service list

​ docker stack ps rmstack

Or docker service ls

5. Delete the Stack

​ docker stack rm rmstack

Check whether all information is deleted on both nodes

version: '3'
services:
  rmcore:
    image: rmcore
    ports:
 - 8088:80
 - 8089:443
    deploy:
        mode: 
         replicated
        replicas: 2
nginx:
    image: nginx
    ports:
 - 8090:80
    deploy:
        mode: 
         replicated
        replicas: 2	
Copy the code

6. Access the Internet

​ 111.229.181.115:6066

​ 123.57.164.54:6066

Swarm couldn’t do it

Use k8S to do so