Skip to content

README.md

Latest commit

 

History

History
71 lines (46 loc) · 1.74 KB

README.md

File metadata and controls

71 lines (46 loc) · 1.74 KB

MySQL

Operational reference for MySQL — local Docker setup, CLI access, and connection strategy benchmarks.

Table of Contents

Local Setup

# Start MySQL container (see docker-compose.yml)
docker-compose up mysqldb

Docker image: hub.docker.com/_/mysql

Connecting

# Connect via TCP from host machine
mysql -h 127.0.0.1 -u root -p   # password: r00t

# Connect via exec into the running container
docker exec -it <container_id> mysql -u root -p

Running Migrations

# Apply schema/seed SQL from within the container
# (after connecting via exec above)
source /path/to/db/mysql/1.sql

Benchmarks — Connection Strategies

These benchmarks compare the cost of shared connection pooling versus opening a new connection per request, isolating the connection overhead from query execution time.

Shared Connection Pool

One connection reused across all requests:

total:   157ms
average:   1ms per request

Per-Request Connection

A new TCP connection is opened for every request. Includes OS-level socket setup, authentication handshake, and session initialisation overhead.

total:   259ms
average:   2ms per request

Takeaway: Per-request connections add ~65% overhead in this test. At scale, connection creation latency accumulates and exhausts the server's max_connections limit. Always use a connection pool (e.g., HikariCP, c3p0, DBCP) in production. See the Connection Pooling section for sizing guidance.