Unit 5: Performance, Monitoring & Troubleshooting

Part J: Performance, Monitoring & Optimization

1. Image Optimization

Use slim/alpine where appropriate, multi-stage builds, .dockerignore, cache-friendly ordering, remove package caches and avoid unnecessary runtime tools.

2. Resource Monitoring & Limits

CPU and memory limits can prevent one container from consuming uncontrolled resources. Monitor usage using:

docker stats
docker top CONTAINER
docker logs -f CONTAINER

3. Disk Usage & Cleanup

docker system df
docker system prune

4. Image Reproducibility

Use deliberate image tags/versioning and deterministic dependency installation. Avoid blindly relying on latest for production.

Part K: Complete Troubleshooting

1. Master Debugging Method

Follow this exact flow when things go wrong:

1. docker ps -a (Check if it exited)
2. docker logs CONTAINER (Read the error)
3. docker inspect CONTAINER (Check configuration)
4. docker exec -it CONTAINER sh (Go inside and debug)
5. Check ports / network / volume
6. Reproduce
7. Fix Dockerfile/Compose
8. Rebuild and validate

2. Common Scenarios

Container Immediately Exits: Containers stay alive only while their main process is running. Run docker logs to see why the main process died.

Database Connection Failed:

• Check both services.
• Use the Compose service name (e.g., db), not localhost.
• Verify credentials/port.
• Remember that depends_on does not necessarily mean the database application is ready to accept connections.

Code Changes Not Visible:

Build Context Too Large: Improve .dockerignore; exclude .git, virtual environments, caches, local secrets and unnecessary datasets.