Add Docker build and per-profile image tooling

[ikyrannas]

Summary

This PR adds a production-ready Docker build for LST-Bench and documentation to streamline local usage and Docker Hub publishing. It also introduces a Makefile that builds/pushes per-profile images so the correct JDBC drivers are included.

Changes

• Add multi-stage Dockerfile with non-root runtime and optional Maven profile support (MAVEN_PROFILES).
• Add .dockerignore to keep build context lean.
• Add Makefile with build-<profile>/push-<profile> targets and build-all/push-all.
• Update README.md with Docker build/run instructions and Spark Thrift example.

Motivation / Context

Running LST-Bench from Docker makes it easier to:

• Standardize environments across machines.
• Ship the right JDBC driver set per engine.
  These changes reduce setup time.

How to Test

Dockerhub image

I have already built and pushed the image to my Dockerhub repo for convenience.

1. Run with your configs:

docker run --rm \
  -v "$PWD/config":/work/config:ro \
  ikyrannas/lst-bench:spark-jdbc \
  -c /work/config/connections.yaml \
  -e /work/config/experiment.yaml \
  -l /work/config/library.yaml \
  -t /work/config/telemetry.yaml \
  -w /work/config/workload.yaml

Local build

1. Build a profile image:

docker build -t lst-bench:spark-jdbc --build-arg MAVEN_PROFILES=spark-jdbc .

2. Run with your configs:

docker run --rm \
  -v "$PWD/config":/work/config:ro \
  lst-bench:spark-jdbc \
  -c /work/config/connections.yaml \
  -e /work/config/experiment.yaml \
  -l /work/config/library.yaml \
  -t /work/config/telemetry.yaml \
  -w /work/config/workload.yaml

Notes

• The Makefile provides per-profile tags (e.g., spark-jdbc, trino-jdbc, etc.) to ensure the correct JDBC driver set is included.

Checklist

• README updated
• No breaking changes
• Tests not required / not applicable for README & build tooling

[ikyrannas]

@microsoft-github-policy-service agree