# Alibaba Druid — High-Performance JDBC Connection Pool with Monitoring

> A production-grade JDBC connection pool for Java that provides built-in SQL monitoring, wall-based security filtering, and detailed performance statistics.

## Install

Save in your project root:

# Alibaba Druid — High-Performance JDBC Connection Pool with Monitoring

## Quick Use
```xml
<dependency>
  <groupId>com.alibaba</groupId>
  <artifactId>druid-spring-boot-starter</artifactId>
  <version>1.2.24</version>
</dependency>
```
```yaml
spring:
  datasource:
    type: com.alibaba.druid.pool.DruidDataSource
    url: jdbc:mysql://localhost:3306/mydb
    druid:
      stat-view-servlet:
        enabled: true
```

## Introduction
Alibaba Druid is a JDBC connection pool designed for monitoring and performance analysis. Beyond pooling connections, it provides a built-in web console that displays SQL execution statistics, slow query logs, and connection usage in real time.

## What Alibaba Druid Does
- Manages JDBC connection pooling with configurable min/max pool sizes and timeouts
- Collects per-SQL execution metrics including count, time, rows, and error rate
- Provides a built-in web dashboard (StatViewServlet) for real-time monitoring
- Includes a WallFilter that blocks SQL injection patterns at the driver level
- Supports SQL parsing and formatting for MySQL, PostgreSQL, Oracle, and SQL Server

## Architecture Overview
Druid wraps the standard JDBC DataSource interface and intercepts every SQL call through a filter chain. The StatFilter collects timing and count data per SQL template. The WallFilter parses each statement through a dialect-specific SQL parser and rejects dangerous patterns. All statistics are stored in memory and exposed via a Servlet-based dashboard or JMX.

## Self-Hosting & Configuration
- Add druid-spring-boot-starter for automatic Spring Boot integration
- Enable the stat view servlet in application.yml for the monitoring dashboard
- Configure connection pool sizes via initialSize, minIdle, and maxActive
- Set slowSqlMillis to flag queries exceeding your latency threshold
- Optionally export metrics to Prometheus via the druid-spring-boot-actuator module

## Key Features
- Built-in SQL monitoring dashboard with no external dependencies
- SQL injection firewall (WallFilter) that works at the connection pool level
- Detailed per-SQL statistics: execution count, total time, max time, affected rows
- Full SQL parser supporting 6+ database dialects
- Connection leak detection with stack-trace logging

## Comparison with Similar Tools
- **HikariCP** — fastest raw throughput; Druid trades marginal speed for built-in monitoring and security
- **Apache DBCP2** — Apache's pool; Druid adds SQL-level metrics and the wall filter
- **c3p0** — legacy pool; Druid is more actively maintained with modern diagnostics
- **PgBouncer** — PostgreSQL-specific external pooler; Druid is a JVM-embedded, database-agnostic solution

## FAQ
**Q: Is Druid slower than HikariCP?**
A: In raw connection-acquire benchmarks HikariCP is slightly faster, but Druid's monitoring overhead is negligible in most production workloads.

**Q: Can I use Druid without Spring Boot?**
A: Yes. Druid works as a standalone DataSource in any Java application.

**Q: Does the WallFilter protect against all SQL injection?**
A: It blocks common injection patterns at the SQL parser level, but should complement application-level parameterized queries, not replace them.

**Q: How do I access the monitoring console?**
A: Enable stat-view-servlet and visit /druid/index.html on your application server.

## Sources
- https://github.com/alibaba/druid
- https://github.com/alibaba/druid/wiki

---
Source: https://tokrepo.com/en/workflows/asset-f53bfd3c
Author: AI Open Source