Analyst Visualization Experiment

One-command Docker lab for comparing and experimenting with BI & data visualization tools — Metabase, Apache Superset, Grafana, and Redash — all connected to a shared PostgreSQL with sample data. Runs on Docker Desktop with WSL2 backend.

---

Table of Contents

• Overview
• Quick Start
• Tools Included
• Shared Database
• WSL / Docker Desktop Notes
• Starting Individual Tools
• Access URLs
• First-Time Setup Per Tool
• Sample Queries to Try
• Desktop-Only Tools
• Comparison Matrix
• Troubleshooting

---

Overview

This project lets you spin up 4 popular BI/visualization tools side-by-side to compare features, learn their interfaces, and decide which fits your workflow. All tools share a single PostgreSQL database pre-loaded with 250K rows of sample data (customers, orders, products).

┌─────────────────────────────────────────────────────────┐
│  Docker (WSL2 backend)                                   │
│                                                          │
│  ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐   │
│  │ Metabase │ │ Superset │ │ Grafana  │ │ Redash   │   │
│  │ :5000    │ │ :8088    │ │ :3000    │ │ :5001    │   │
│  └────┬─────┘ └────┬─────┘ └────┬─────┘ └────┬─────┘   │
│       │            │            │            │          │
│       └────────────┴────────────┴────────────┘          │
│                        │                                 │
│                 ┌──────┴──────┐                          │
│                 │ PostgreSQL  │  ← 250K sample rows      │
│                 │ :5439       │                          │
│                 └─────────────┘                          │
└─────────────────────────────────────────────────────────┘
         │
         ▼
  Windows Browser
  http://localhost:5000  → Metabase
  http://localhost:8088  → Superset
  http://localhost:3000  → Grafana
  http://localhost:5001  → Redash

---

Quick Start

Prerequisites

• Docker Desktop installed on Windows (with WSL2 backend enabled)
• 4 GB+ RAM free (all tools together use ~2-3 GB)
• Git Bash or WSL terminal to run the scripts

1. Clone and start

cd ~/Desktop/Analyst_Visualization_experiment

# Make scripts executable
chmod +x start.sh stop.sh

# Start everything
./start.sh

First boot downloads images (2-5 minutes). Subsequent boots are instant.

2. Open in browser

Tool	URL	Login
Metabase	http://localhost:5000	Setup wizard on first visit
Superset	http://localhost:8088	admin / admin
Grafana	http://localhost:3000	admin / admin
Redash	http://localhost:5001	Create account on first visit
Shared DB	localhost:5439	analyst / analyst_pass / analytics

3. Stop

./stop.sh           # Stop all, keep data
./stop.sh clean     # Stop all, remove everything

---

Tools Included

Metabase

• Type: Open-source BI (Java)
• Best for: Quick dashboards, non-technical users, simple SQL questions
• Strengths: Zero-config setup, beautiful UI, "Ask a question" natural language
• Weaknesses: Limited advanced chart types, less flexible than Superset

Apache Superset

• Type: Open-source data exploration (Python)
• Best for: Power users, complex dashboards, SQL Lab
• Strengths: 40+ chart types, SQL IDE, semantic layer, row-level security
• Weaknesses: Heavier setup, steeper learning curve

Grafana

• Type: Open-source observability (Go)
• Best for: Time-series monitoring, infrastructure dashboards, alerts
• Strengths: Industry standard for monitoring, excellent time-series, plugins
• Weaknesses: Not ideal for ad-hoc BI queries, SQL support is secondary

Redash

• Type: Open-source SQL dashboarding (Python)
• Best for: SQL-heavy teams, scheduled queries, alerts
• Strengths: Pure SQL focus, query snippets, easy sharing
• Weaknesses: Fewer chart types, smaller community than Metabase/Superset

---

Shared Database

All tools connect to the same PostgreSQL instance with pre-loaded data:

Parameter	Value
Host	bi_postgres (inside Docker) or localhost (from Windows)
Port	5439
Database	analytics
Username	analyst
Password	analyst_pass

Sample Data

Table	Rows	Description
customers	100,000	Customer profiles with location and status
orders	100,000	Orders with amounts, categories, and statuses
products	50,000	Product catalog with pricing and stock status

Connecting from each tool

When adding a data source in any tool, use:

Host:     bi_postgres
Port:     5432           (internal Docker port, NOT 5439)
Database: analytics
User:     analyst
Password: analyst_pass

⚠ Important: Use port 5432 inside Docker (container-to-container) and 5439 from Windows.

---

WSL / Docker Desktop Notes

How it works

Docker Desktop on Windows runs containers inside a WSL2 virtual machine. From Windows:

Windows Browser → localhost:PORT → Docker proxy → Container

This is automatic — Docker Desktop forwards localhost ports to WSL2. No special IP configuration needed.

Accessing from WSL terminal

# From WSL bash, same URLs work:
curl http://localhost:5000

Accessing from another machine on the network

# 1. Get your WSL IP
ip addr show eth0 | grep inet

# 2. From another machine, use:
http://<WSL_IP>:5000

For persistent network access, configure Windows Firewall to allow the ports.

Resource Limits

Docker Desktop → Settings → Resources → WSL Integration:

• Memory: Set to at least 8 GB if running all tools
• CPUs: 4+ recommended
• Enable integration with your WSL distro

Docker Compose vs Docker Compose (V2)

This project uses the modern docker compose (V2) syntax. If you see 'compose' is not a docker command, use docker-compose (with hyphen) instead.

---

Starting Individual Tools

You don't have to run everything. Use the helper script with a tool name:

./start.sh metabase    # Only Metabase + shared DB
./start.sh superset    # Only Superset + shared DB
./start.sh grafana     # Only Grafana + shared DB
./start.sh redash      # Only Redash (has its own DB)

Or use Docker Compose directly:

# Start shared DB first (required for Metabase, Superset, Grafana)
docker compose -f docker-compose-shared-db.yml up -d

# Then start your tool of choice
docker compose -f docker-compose-metabase.yml up -d
docker compose -f docker-compose-superset.yml up -d
docker compose -f docker-compose-grafana.yml up -d
docker compose -f docker-compose-redash.yml up -d

---

Access URLs

Service	URL	Notes
Metabase	http://localhost:5000	First visit → follow setup wizard
Superset	http://localhost:8088	Login: admin / admin
Grafana	http://localhost:3000	Login: admin / admin
Redash	http://localhost:5001	Create account on first visit
PostgreSQL	localhost:5439	analyst / analyst_pass

---

First-Time Setup Per Tool

Metabase

1. Open http://localhost:5000
2. Click "Let's get started"
3. Choose language → fill in your name/email/password
4. Add database: PostgreSQL → host: bi_postgres, port: 5432, db: analytics, user: analyst, pass: analyst_pass
5. Skip sample data (you already have it)
6. Start building questions and dashboards

Superset

1. Open http://localhost:8088 → login admin / admin
2. Settings → Database Connections → + Database
3. Choose PostgreSQL
4. URI: postgresql://analyst:analyst_pass@bi_postgres:5432/analytics
5. Click "Connect" → "Finish"
6. Go to SQL Lab → start querying
7. Create charts and dashboards

Grafana

1. Open http://localhost:3000 → login admin / admin
2. Skip changing password (or change it)
3. Connections → Data Sources → Add data source → PostgreSQL
4. Host: bi_postgres:5432, Database: analytics, User: analyst, Password: analyst_pass
5. Save & Test
6. Create dashboard → Add panel → write SQL queries

Redash

1. Open http://localhost:5001
2. Create admin account (first visit)
3. Settings → Data Sources → New Data Source → PostgreSQL
4. Host: bi_postgres, Port: 5432, Database: analytics, User: analyst, Password: analyst_pass
5. Create → Save
6. Write queries and build visualizations

---

Sample Queries to Try

Once connected, try these queries across all 4 tools to compare the experience:

Revenue by country (bar chart)

SELECT country, ROUND(SUM(amount)::numeric, 2) AS total_revenue
FROM orders o
JOIN customers c ON c.id = o.customer_id
GROUP BY country
ORDER BY total_revenue DESC;

Monthly order trends (line chart)

SELECT
    DATE_TRUNC('month', order_date) AS month,
    COUNT(*) AS order_count,
    ROUND(AVG(amount)::numeric, 2) AS avg_order_value
FROM orders
GROUP BY DATE_TRUNC('month', order_date)
ORDER BY month;

Category breakdown (pie chart)

SELECT product_category, COUNT(*) AS orders
FROM orders
GROUP BY product_category
ORDER BY orders DESC;

Top 10 cities by customer count (table)

SELECT city, country, COUNT(*) AS customers
FROM customers
GROUP BY city, country
ORDER BY customers DESC
LIMIT 10;

High-value pending orders (table)

SELECT o.id, c.name, o.amount, o.order_date
FROM orders o
JOIN customers c ON c.id = o.customer_id
WHERE o.status = 'pending' AND o.amount > 500
ORDER BY o.amount DESC
LIMIT 20;

---

Desktop-Only Tools (Cannot Dockerize)

These tools are Windows/Mac desktop applications. Install them separately to compare:

Tool	Type	Install
Tableau Public	Desktop BI (free)	https://public.tableau.com
Power BI Desktop	Desktop BI (free)	https://powerbi.microsoft.com/desktop
Google Looker Studio	Cloud BI (free, web)	https://lookerstudio.google.com
Excel / Google Sheets	Spreadsheet	Built-in / sheets.google.com

Connecting desktop tools to the Docker DB

Desktop apps run on Windows, not inside Docker. Use:

Host:     localhost
Port:     5439           ← external mapped port
Database: analytics
User:     analyst
Password: analyst_pass

---

Comparison Matrix

Feature	Metabase	Superset	Grafana	Redash
Setup complexity	⭐ Easy	⭐⭐⭐ Medium	⭐ Easy	⭐⭐ Medium
Chart types	15+	40+	20+ (plugin)	15+
SQL editor	✓ Simple	✓✓ SQL Lab (full)	✓ Basic	✓✓ Good
Natural language	✓ "Ask a question"	✗	✗	✗
Dashboards	✓✓ Good	✓✓✓ Best	✓✓ Good	✓ Good
Alerts	✓	✓	✓✓✓ Best	✓
Semantic layer	✓	✓✓	✗	✗
Multi-tenant	✗	✓✓✓ Row-level	✓ Org-based	✓ Basic
Time-series focus	✗	✗	✓✓✓ Best	✗
License	AGPL (OSS)	Apache 2.0	AGPL (OSS)	BSD
Best for	Quick BI	Power users	Monitoring	SQL teams

---

Troubleshooting

Port already in use

# Check what's on the port
netstat -ano | findstr :5000

# Kill the process (Windows)
taskkill /PID <PID> /F

Container fails to start

# Check logs
docker logs metabase
docker logs superset
docker logs grafana
docker logs redash_server

Can't connect to shared DB from tool

Make sure you're using host bi_postgres port 5432 (not localhost or 5439) inside Docker.

Full reset

./stop.sh clean
./start.sh

WSL not running

# In PowerShell (as admin):
wsl --install
wsl --shutdown
wsl

Docker Desktop not connected to WSL

Docker Desktop → Settings → Resources → WSL Integration → Enable your distro → Apply & Restart.

---

License

MIT — free to use and modify for learning and comparison purposes.