Production Environment Deployment

This document describes a deployment option suitable for production environments.

For production environment deployment, you can use the https://github.com/muyantech/installer (opens new window) all-in-one installation package, which supports the following features:

1. This all-in-one installation package contains all the related dependent components needed to run the system.
2. This all-in-one installation package is based on Docker Swarm and GlusterFS distributed file system, supporting multi-node and high availability.

TIP

If you purchased the system from https://lcdp.ai, you can get help from our technical support team.

Target Audience

The target audience for this document is: System implementers

Prerequisites

1. This installation package has been tested on English versions of Ubuntu 20.04 LTS and 22.04 LTS operating systems.
2. The system needs to create an operating system user named ubuntu who can use sudo without a password.

Component Description

This installer will install all components into the my stack in the swarm cluster.

The following lists the main related software and their versions used:

Software	Current Version	Description
Docker	24.0.6	Docker Engine
glusterfs	10.1	GlusterFS distributed storage
ntp		Time server

The following lists the services included in the current all-in-one installation package:

Component	Image and Version	Description	Service Name in Swarm
proxy	nginx:1.25.2	Web server, running in front of client and server services, used for service routing	my_proxy
client	nginx:1.25.2	Frontend service	my_client
redis	redis:7.0.10	Redis server, used for backend service caching and queue service	my_redis
database	timescale/timescaledb-ha:pg14-all	Database server, used for backend service data storage with time-series database functionality	my_database
pgadmin	dpage/pgadmin4:7.7	Database maintenance tool, used to connect to the database via web interface and execute SQL, etc.	my_pgadmin
server	adoptopenjdk:11-jdk-hotspot	Backend service	my_server

System Installation

Installation steps are as follows:

1. Log in as root user, create ubuntu user and set its sudo permissions

# Create ubuntu user
useradd -m -s /bin/bash ubuntu

# Edit sudo configuration file
visudo

# Add the following line at the end of the file to set ubuntu user's sudo permissions
ubuntu ALL=(ALL) NOPASSWD:ALL

# Save and exit

2. Clone the repository from https://github.com/muyantech/installer (opens new window).

Large files like installation packages are saved using git lfs and need to be downloaded using git lfs.

# Install git lfs
apt update
apt install git-lfs

# Clone the repository
git clone [email protected]:muyantech/installer.git

# Download all files managed by git lfs
cd installer
git lfs pull

3. Edit the .env file and modify the relevant installation parameters. The relevant installation parameters are described in detail as follows:

# Whether to enable GlusterFS (NFS, only needed for multi-node mode)
# If the system needs to run on multiple nodes, please set this option to true
GLUSTER_ENABLED=false

# Whether to enable HTTPS, if enabled, you need to provide SSL certificate to /assets/configs/proxy/ssl folder
# If you want to enable HTTPS access to the system, you need to put the SSL certificate in the /assets/configs/proxy/ssl directory before starting the installation
HTTPS_ENABLED=false

# Extra hosts, will be injected to backend service, the docker swarm service is called my_server
# Format: "hostname:ip"
# Please keep this placeholder even if you don't need it
# Additional hosts DNS mapping needed in the docker container of the backend service, the
 service name in the swarm cluster is my_server
# Format: "hostname:ip address"
# If not needed, please keep the placeholder line below unchanged
EXTRA_HOSTS="samt:192.168.0.1"

# Domain name to access the system
# Domain name to access the system
NGINX_HOST=samt.myddm.com

4. In the installation package directory, run the following command to install the system:

# Run the installer
bash +x cluster-setup.sh

The following is the default interface of the installer. Please read the relevant prompts carefully and record the installation log output file path. You can use tail -f <installation log output path> to track the detailed installation process log.

After confirming that the relevant configuration is correct, enter y to start the installation process.

The system will display the following interface and stop the currently running services:

According to the prompt,

1. If installing the system on a single server, just press Enter.
2. If installing the system on a cluster of multiple servers, please enter the IP addresses of other servers in the cluster. Press Enter after entering each one, and press Enter twice when finished.

After the input is complete, the system will start the installation process and automatically start all services after the installation is complete.

Service Refresh

Run the following command to refresh one or all services:

# Refresh a specific service
bash +x scripts/refresh-service.sh <service-name>

# Refresh all services, refresh Swarm cluster configuration,
# After modifying the docker-compose.yml file, execute the following command to refresh the cluster configuration
bash +x scripts/refresh-service.sh 

Available service names are as follows:

• my_proxy
• my_client
• my_redis
• my_database
• my_pgadmin
• my_server

Program Update

Update Frontend Program

1. Copy the compiled files of the frontend to the directory <installer>/assets/applications/client.
2. In the <installer> directory, execute bash +x scripts/refresh-service.sh my_client

After completing the above operations, you can deploy the new version of the frontend and refresh the corresponding service.

Update Backend Program

1. Copy the compiled jar file of the backend to the directory <installer>/assets/applications/server, and rename it to app.jar.
2. In the <installer> directory, execute bash +x scripts/refresh-service.sh my_server

After completing the above operations, you can deploy the new version of the backend and refresh the corresponding service.

Additional Information

Program File Paths

Component	Directory	Description
client	/opt/muyan/runtime/applications/client	Path for frontend service's css, html, and js files
server	/opt/muyan/runtime/applications/server	Path for backend service's jar file, the jar file should be named app.jar

Component Configuration File Paths

Component	Configuration File	Description
proxy	/opt/muyan/runtime/configs/proxy/proxy.conf.template	Nginx configuration for Proxy service
client	/opt/muyan/runtime/configs/client/nginx.conf	Nginx configuration for Client service
database	/opt/muyan/runtime/configs/postgres/postgresql.conf	Main configuration for database service
database	/opt/muyan/runtime/configs/postgres/pg_hba.conf	Access permission config for database
redis	/opt/muyan/runtime/configs/redis/redis.conf	Configuration for Redis service
pgadmin	/opt/muyan/runtime/pgadmin/config/servers.json	Configuration for pgadmin service
pgadmin	/opt/muyan/runtime/pgadmin/pgpass	Server configuration for pgadmin service

Component Data File Paths

Component	Directory	Description
server	/opt/muyan/runtime/data/attachments	Storage path for attachment files uploaded by users when using LOCAL_FILE storage engine
server	/opt/muyan/runtime/data/seed	Seed data directory, the system will read seed data from this directory during startup
database	/opt/muyan/runtime/data/postgres	Data directory for database service
redis	/opt/muyan/runtime/data/redis	Data persistence directory for Redis service
pgadmin	/opt/muyan/runtime/pgadmin/data	Data persistence directory for pgadmin service

TIP

If there is seed data in the directory /opt/muyan/runtime/data/seed, the seed data should be complete. The priority of seed data in this directory is higher than the priority of seed data packaged in the running jar file.

Component Log Directory List

Component	Directory	Description
proxy	/opt/muyan/runtime/logs/proxy	Access and error logs for proxy service
client	/opt/muyan/runtime/logs/client	Access and error logs for client service
server	/opt/muyan/runtime/logs/server	Logs for server service
database	/opt/muyan/runtime/logs/postgres	Logs for database service
redis	/opt/muyan/runtime/logs/redis	Logs for redis service
pgadmin	/opt/muyan/runtime/pgadmin/log	Logs for pgadmin service

Command Reference

Docker Swarm Related Commands

• View logs of a specific service

# View logs of the service named my_client
docker service logs my_client

• Force restart a service

# Force restart the service named my_client
docker service update my_client --force

• View the current running instance list and status of a service

# View the instance list and status of the service named my_server
docker service ps my_server

• List all nodes in the swarm cluster and their status

# List all nodes
docker node ls
# View the status of the current node
docker node inspect self
 --pretty
# View the detailed status of a specific node
docker node inspect <NODE-ID> --pretty

• Inspect a specific service

# View the detailed status of a service
docker service inspect --pretty <service_name>

GlusterFS Related Commands

• Mount remote volume to local directory

# Please note that the glusterfs volume must be mounted to be accessed locally
# And the client should write to this mounted directory,
# Not the gluster brick folder
mount -t glusterfs 172.104.
32.234:/my-runtime /opt/muyan/runtime

• Enable self-healing every ten minutes to repair files that need repair (to be repaired)

gluster volume heal my-runtime enable

• View files in split brain status

gluster volume heal my-runtime info split-brain

• Mount disk according to /etc/fstab

mount -av

• Delete existing volume

gluster volume stop my-runtime

• Remove a brick, set replica to 2 --> If replica is reduced, need to use force

gluster volume remove-brick my-runtime replica 2 139.177.191.133:/opt/muyan/gluster force
gluster volume remove-brick my-runtime replica 2 172.104.179.24:/opt/muyan/gluster force

• Remove a node from peer pool

gluster peer detach 172.104.179.24

• Delete volume

gluster volume delete my-runtime

Redis Related Commands

• Monitor Redis key deletions

# Execute inside redis-cli
# Config event log for redis key changes
# https://stackoverflow.com/questions/40436417/redis-keys-are-getting-deleted-automatically
CONFIG SET notify-keyspace-events AKE