How can I optimize a Dockerized legacy Java app for startup time and memory?

[Ella.W]

I'm containerizing our team's legacy Java application using Docker to standardize our development and deployment environments, but I'm running into persistent issues with container startup time and memory usage that are much higher than when running the app directly on the host. The Dockerfile is based on an official OpenJDK image, but I suspect I'm not leveraging multi-stage builds or JVM tuning flags effectively. For developers who have optimized Java applications in Docker, what are the key configuration tweaks and best practices for reducing both image size and runtime footprint, and how do you properly set JVM heap and garbage collection parameters inside a container to prevent it from being killed by the host's OOM killer? I need this to be efficient for our CI/CD pipeline.