HassanAbouelela/DevopsDemoLive

Steps

1. Install python and minikube
2. Start minikube: minikube start --driver docker
3. Prepare a virtual environment: python -m venv venv
4. pip install flask "psycopg[binary]" python-dotenv
5. Start the database: docker compose up -d postgres-db
6. Obtain the docker container ID: docker ps
7. Run a shell in the container: docker exec -it <ID> /bin/bash
8. Connect to the postgres shell psql -U postgres
9. Switch to the site database: \c site
10. Create a table in the database. Refer to postgres-init.sql for sample SQL commands.
11. Check the table's existence: \dt
12. Integrate the database with the API and test it
13. Build the API docker image by hand: docker build -t devopsdemo:latest .
14. Run the devopsdemo image, either by hand with docker run or via compose as a new service
15. Move the build command into the docker-compose service
16. Run the new compose service: docker compose up -d site

Postgres Structure

graph TD;
d["Database (site)"]
s["Schema (public)"]
t["Table (important_data)"]

d --> Users;
d --> Views;
d --> etc;
d --> s;
s --> t;
t --> data;

Loading

Errata

• Added a volume to persist postgres data in the docker compose:

<snip>
postgres-db:
    volumes:
        # /var/lib/postgresql/data is the location of the actual database files
        - pg-data:/var/lib/postgresql/data
<snip>
volumes:
    pg-data:

• Postgres will load the init file on startup only if there is no data in the DB
• Healthcheck should specify the condition to wait for:

depends_on:
    postgres-db:
        condition: service_healthy