Project Loom Lab

Experiments with Project Loom's features based on these JEPs:

• JEP 444: Virtual Threads
• JEP 453: Structured Concurrency

This includes experimenting with these features directly as well as activating them in web frameworks.

You need Java 21.

Experiments

• Disk Stats
• Echo Client & Server
• GitHub Crawler

Change into the project folder experiments and build it with mvn package to get target/loom-experiments.jar. To run it:

java --enable-preview -p target/loom-experiments.jar -m loom.experiments $EXPERIMENT $ARGUMENTS

Where:

• $EXPERIMENT selects one of the experiments by name
• $ARGUMENTS configures the experiment

For details on these, see specific experiments below.

Disk Stats

Walks over all folders and files in a given directory to gather their respective sizes. Can be configured to either run as a single thread or with one virtual thread for each file/folder.

• name: DiskStats
• arguments: see DiskStats.java.
• package: dev.nipafx.lab.loom.disk

Echo Client & Server

A client and server that exchange messages via sockets on localhost:8080. Client protocol:

• sends a single line, terminated by a newline
• waits for a single line (i.e. a string terminated by a newline) to be received

Server protocol:

• reads a single line (i.e. a string terminated by a newline) from that socket
• waits a predetermined amount of time
• replies with the same string, including the newline

To try this out, run the client and the server in different shells.

• server
  • name: EchoServer
  • arguments: see Echo.java.
• client
  • name: EchoClient
  • arguments: see Send.java.,
• package: dev.nipafx.lab.loom.echo

Note: For a much more thorough experiment with an echo server, check out Elliot Barlas' project-loom-experiment.

GitHub Crawler

Starting from a given seed URL, crawls GitHub pages and prints their connections and statistics. Only runs with virtual threads but also uses/demonstrates some data-oriented programming concepts.

• name: GitHubCrawl
• arguments: see GitHubCrawl.java.
• package: dev.nipafx.lab.loom.crawl

Frameworks

• Spring Boot
• Quarkus

Spring Boot

Change into the project folder frameworks/spring_boot and build with mvn package, then run it with:

java -jar target/loom-spring-boot.jar

Once launched, visit http://localhost:8080/api/current-thread. To switch to virtual threads, append the command line parameter --spring.threads.virtual.enabled=true, which sets the Spring property spring.threads.virtual.enabled, which in turn leads to Spring using a virtual thread for each request.

As far as I know, the release notes for version 3.2 are the most up-to-date documentation on Spring and virtual threads.

Quarkus

Change into the project folder frameworks/quarkus and build with mvn package, then run it with:

java --add-opens java.base/java.lang=ALL-UNNAMED -jar target/quarkus-app/quarkus-run.jar

Once launched, visit http://localhost:8080/api/current-thread. To switch to virtual threads, edit QuarkusEndpoints and uncomment @RunOnVirtualThread, which allows configuring virtual threads for individual endpoints. This annotation requires the --add-opens. The artifact quarkus-netty-loom-adaptor (see pom.xml) seems to improve performance of Quarkus on virtual threads.

As far as I know, the blog post Writing simpler reactive REST services with Quarkus Virtual Thread support is the most up-to-date documentation on Quarkus and virtual threads.