radin/my-docs
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

docker doc: update swarm doc

Browse Source
This commit is contained in:
radinpirouz
2025-05-16 00:15:52 +03:30
parent 5e3d651660
commit 92e3749529
1 changed files with 104 additions and 68 deletions
Download Patch File Download Diff File Expand all files Collapse all files

172
Containerization & Orchestration/Docker/5-Docker-Swarm.md
View File

@@ -1,89 +1,85 @@
# Docker Swarm Documentation # 📦 Docker Swarm Documentation
This guide covers key commands and workflows for managing a Docker Swarm cluster. It includes information on initializing the cluster, node management, service management, and best practices to ensure your cluster remains operational. Comprehensive guide to managing a **Docker Swarm** cluster. This document includes core commands, workflows, and best practices for maintaining a healthy and operational environment.
--- ---
## Table of Contents ## 📚 Table of Contents
1. [Cluster Health & Manager Count](#cluster-health--manager-count) 1. [🔧 Cluster Health & Manager Count](#-cluster-health--manager-count)
2. [Example Workflow: Setting Up Nginx](#example-workflow-setting-up-nginx) 2. [🚀 Example Workflow: Setting Up Nginx](#-example-workflow-setting-up-nginx)
3. [Cluster Initialization and Management](#cluster-initialization-and-management) 3. [⚙️ Cluster Initialization and Management](#️-cluster-initialization-and-management)
4. [Node Management](#node-management) 4. [🖥️ Node Management](#️-node-management)
5. [Service Management](#service-management) 5. [🛠️ Service Management](#️-service-management)
6. [Join Tokens](#join-tokens) 6. [🔑 Join Tokens & Node Configuration](#-join-tokens--node-configuration)
--- ---
## Cluster Health & Manager Count ## 🔧 Cluster Health & Manager Count
To maintain a healthy and functional Docker Swarm cluster, **the number of manager nodes must exceed 50% of the total nodes**. If the manager nodes fall below 51% of the cluster, the system will lose quorum and become non-operational. Always monitor your manager count to ensure high availability. A Docker Swarm cluster requires a **majority of manager nodes** to be functional for quorum.
> **Best Practice:** Always maintain **more than 50%** manager nodes online. Losing quorum will render the cluster non-operational.
--- ---
## Example Workflow: Setting Up Nginx ## 🚀 Example Workflow: Setting Up Nginx
When deploying services such as Nginx, the typical workflow in Docker Swarm is: Docker Swarm handles service deployment through several internal components:
1. **API**: Handle incoming requests. 1. **API** – Receives service requests.
2. **Allocator**: Distribute workload to available nodes. 2. **Allocator** – Determines resource allocation.
3. **Dispatcher**: Manage task assignments. 3. **Dispatcher** – Assigns tasks to nodes.
4. **Scheduler**: Ensure tasks run on the optimal nodes. 4. **Scheduler** – Places tasks on optimal nodes.
This workflow ensures that your service is efficiently distributed and scaled across the cluster. This process ensures resilient and efficient service distribution.
--- ---
## Cluster Initialization and Management ## ⚙️ Cluster Initialization and Management
### Create Cluster ### 🔹 Initialize Cluster
Initialize a new Docker Swarm cluster:
```bash ```bash
docker swarm init docker swarm init
``` ```
This command sets up the current node as the manager of a new cluster.
### Create Cluster with a Specific Interface ### 🔹 Initialize with Specific Interface
Specify an IP address or interface name when initializing the cluster:
```bash ```bash
docker swarm init --advertise-addr <ip or interface name> docker swarm init --advertise-addr <ip-or-interface>
``` ```
This ensures that the node advertises the correct network interface for other nodes to join.
### Join Cluster ### 🔹 Join Existing Cluster
Add a node to an existing Docker Swarm cluster:
```bash ```bash
docker swarm join docker swarm join
``` ```
### Leave Cluster ### 🔹 Leave Cluster
Remove a node from the cluster:
```bash ```bash
docker swarm leave docker swarm leave
``` ```
### Unlock Locked Manager ### 🔹 Unlock a Manager Node
Unlock a manager node that has been locked:
```bash ```bash
docker swarm unlock docker swarm unlock
``` ```
--- ---
## Node Management ## 🖥️ Node Management
### List Nodes ### 🔸 List Nodes
View the status and details of all nodes in your cluster:
```bash ```bash
docker node ls docker node ls
``` ```
**Example Output:** **Example Output:**
``` ```
ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION ID HOSTNAME STATUS AVAILABILITY MANAGER STATUS ENGINE VERSION
8yw8jrjeqczaci0qkuy060g09 * docker-1 Ready Active Leader 24.0.5 8yw8jrjeqczaci0qkuy060g09 * docker-1 Ready Active Leader 24.0.5
@@ -92,101 +88,141 @@ kd3ujmt1ey3pw6v9189fouxfa docker-3 Ready Active Reachable
tm1msy58ztcltt36rs1lb76p7 docker-4 Down Active 24.0.5 tm1msy58ztcltt36rs1lb76p7 docker-4 Down Active 24.0.5
``` ```
### Remove Node ### 🔸 Remove a Node
Remove a node from the cluster by specifying its ID:
```bash ```bash
docker node rm <node-id> docker node rm <node-id>
``` ```
**Example:**
### 🔸 Promote to Manager
```bash ```bash
docker node rm tm1msy58ztcltt36rs1lb76p7 docker node promote <hostname-or-id>
``` ```
### Promote Node to Manager ### 🔸 Inspect a Node
Upgrade a worker node to a manager node:
```bash ```bash
docker node promote <hostname or ID> docker node inspect <nodename>
``` ```
**Example:**
### 🔸 Change Node Role
```bash ```bash
docker node promote v4gvf7xenw0izmxgvhr6hb2rj docker node update --role manager <nodename>
docker node update --role worker <nodename>
```
### 🔸 Change Node Availability
```bash
docker node update --availability active <nodename>
docker node update --availability pause <nodename>
docker node update --availability drain <nodename>
```
### 🔸 Add or Remove Labels
**Add:**
```bash
docker node update --label-add env=development <nodename>
docker node update --label-add env=testing <nodename>
```
**Remove:**
```bash
docker node update --label-rm env= <nodename>
docker node update --label-rm env <nodename>
```
**Use label constraints in service deployment:**
```yaml
deploy:
placement:
constraints:
- node.labels.env == development
``` ```
This command promotes the specified node, enabling it to participate in cluster management decisions.
--- ---
## Service Management ## 🛠️ Service Management
### Show Task Status on Cluster ### 🔹 Show Tasks on a Node
Display the status of tasks running on each node:
```bash ```bash
docker node ps docker node ps
``` ```
### List Services ### 🔹 List All Services
List all services currently running in the cluster:
```bash ```bash
docker service ls docker service ls
``` ```
### Create Service ### 🔹 Create a New Service
Create a new service with the specified name and Docker image:
```bash ```bash
docker service create --name <service-name> <image-name> docker service create --name <service-name> <image-name>
``` ```
### Scale Service ### 🔹 Scale a Service
Adjust the number of replicas for an existing service:
```bash ```bash
docker service scale <service-name>=<replica-count> docker service scale <service-name>=<replica-count>
``` ```
**Example:** **Example:**
```bash ```bash
docker service scale nginx=5 docker service scale nginx=5
``` ```
This command scales the `nginx` service to 5 replicas.
### Inspect Service ### 🔹 Inspect a Service
View detailed information about a specific service:
```bash ```bash
docker service inspect <service-name> docker service inspect <service-name>
``` ```
### Create Service with Replicas and Environment Variables ### 🔹 Create Service with Replicas, Env Vars, and Port Mapping
Launch a new service with multiple replicas, environment variables, and port publishing:
```bash ```bash
docker service create --name <service-name> --replicas <replica-count> --env <env-variable> --publish <host-port>:<container-port> <image-name> docker service create \
--name <service-name> \
--replicas <count> \
--env <ENV_VAR=value> \
--publish <host-port>:<container-port> \
<image-name>
``` ```
**Example:** **Example:**
```bash ```bash
docker service create --name nginx --replicas 3 --env MY_ENV_VAR=value --publish 8080:80 nginx docker service create \
--name nginx \
--replicas 3 \
--env MY_ENV_VAR=value \
--publish 8080:80 \
nginx
``` ```
This creates an `nginx` service with 3 replicas, sets the environment variable `MY_ENV_VAR` to `value`, and maps port 8080 on the host to port 80 in the container.
--- ---
## Join Tokens ## 🔑 Join Tokens & Node Configuration
To securely add new nodes to the swarm, use the join tokens provided by Docker Swarm. Securely add nodes to your Swarm using join tokens.
### Get Worker Join Token ### 🔹 Get Worker Token
Display the token required for a node to join as a worker:
```bash ```bash
docker swarm join-token worker docker swarm join-token worker
``` ```
### Get Manager Join Token ### 🔹 Get Manager Token
Display the token required for a node to join as a manager:
```bash ```bash
docker swarm join-token manager docker swarm join-token manager
``` ```