喆拉德筆記

Author: Gerard Chen

  • mairadb command, basic 101

    # create database
CREATE DATABASE `databaseName`;

# create user
CREATE USER 'my_user'@'localhost' IDENTIFIED BY 'your_strong_password';

# grant privileges
GRANT ALL PRIVILEGES ON my_new_database.* TO 'my_user'@'localhost';

FLUSH PRIVILEGES;
    # in docker
docker exec -it container_name mariadb -u root -p

# Create the Database
CREATE DATABASE `dbName`;
# Create the User (Using '%' so you can connect from outside the container)
CREATE USER 'userName'@'%' IDENTIFIED BY 'userPassword';

# Grant Permissions
GRANT ALL PRIVILEGES ON `dbName`.* TO 'dbUser'@'%';
# Apply changes
FLUSH PRIVILEGES;

# Exit the prompt
EXIT;

  • erpNEXT, “Multi-Bench, Multi-Tenant” architecture setting

    Bench Name (Domain)Internal Port Project
    alpha.comalpha.com8000alpha
    beta.combeta.com8001beta
    gamma.comgamma.com8002gamma

    Initialize the Bench

    # Replace 'alpha.com' with your actual domain
bench init alpha.com --frappe-branch version-15
cd alpha.com

  • Gitea deploy by using docker yml

  • GitLab deploy by using docker yml

    結論

    GitLab真的超吃資源，沒有優化的情況下，要佔用10.9G的RAM。只有自己用，現在RAM的價格，又跟黃金差不多。可能GitLab功能很強，現階段不適用。ChatGPT建議改用Gitea，就來試試看。

  • enlarge docker space

    問題：Your Docker build environment ran out of disk space.

    # 檢查 docker 狀態

    df -h
docker system df

    # 清理docker 空間

    # Remove all unused data
docker system prune -a --volumes

# Remove build cache
docker builder prune -a

# Check again
df -h

    問題：變更docker 位置

    # Stop Docker
sudo systemctl stop docker

# Edit Docker config
sudo nano /etc/docker/daemon.json
    # add the following code in daemon.json

{
  "data-root": "/new/docker/path"
}
    # Create new directory
sudo mkdir -p /new/docker/path

# Move existing data (optional)
sudo rsync -aP /var/lib/docker/ /new/docker/path/

# Start Docker
sudo systemctl start docker

  • erpNEXT deploy by yml

  • Open-Source Workflow/Automation Tools

    Ranked by Popularity

    1. n8n — Very Popular
      • Widely used for general workflow automation with a visual builder.
      • Large community and many integrations.
      • Often ranked #1 in workflow tool lists.
    2. Node-RED — Highly Popular
      • Strong open-source user base, especially for IoT and integration workflows.
      • Long history and strong community.
    3. Apache Airflow — Very Popular (in data/engineering)
      • Dominant in data engineering and complex scheduled pipelines; huge adoption in enterprise data teams.
    4. Activepieces — Fast-Growing
      • Newer but rapidly rising in many “top open-source automation” lists.
      • Growing ecosystem and community interest.
    5. Huginn — Popular Niche Tool
      • Known in self-hosted automation communities for event-based agents.
    6. Windmill — Emerging
      • Increasingly included in open-source automation tool lists; still smaller compared to older players.
    7. Kestra — Growing
      • Gaining attention for scalable orchestration, less mainstream than Airflow but rising.

    Ranked with Brief Comments

    RankToolWhy it’s popular
    1n8nGeneral workflow automation; strong community; visual UI.
    2Node-REDMature, widely used in IoT and integrations; broad ecosystem.
    3Apache AirflowVery popular in data engineering and ETL use cases.
    4ActivepiecesNewer but rapidly adopted; mentioned in many 2025 tool lists.
    5HuginnKnown in privacy-focused and scriptable automation circles.
    6WindmillEmerging workflow tool, gaining traction.
    7KestraGrowing in developer orchestration space.

  • GPT 比較：chatGPT、Gemini、Claude

    Claude

    # Pron

    在APP佈署上，真的比較強。應該是系統內部，內建系統prompts羅輯就比較精細。

    # Cron

    “You’ve used 75% of your weekly limit”
    “You’ve used 75% of your monthly limit”

    一堆的token使用量限制，只要超過資費的token使用上限 (日限制 / 週限制 / 月限制)，要嘛就等到限制被解除，要嘛就升級方案，就是要多付錢)

    Gemini

    # Pron

    # Cron

    chatGPT

    # Pron

    # Cron

  • Koha deploy by docker-compose.yml

  • deploy onlyoffice by using yml file

    Reference

    1. GitHub / ONLYOFFICE / Docker-CommunityServer
    2. GitHub / ONLYOFFICE / Docker-CommunityServer / docker-compose.workspace.yml