Explore Project Loom's revolutionary approach to lightweight concurrency in the JVM through this Devoxx conference talk. Dive into the world of concurrent programming, unraveling concepts like fibers, threads, and coroutines. Examine code samples, learn about Loom's inception and recent prototypes, and discover how it aims to transform the JVM beyond simple primitives. Investigate CPU utilization, thread creation, visual DVM, system calls, and context switching. Compare first and second implementations, exploring topics such as service threads, socket channel API, asynchronous circuits, and handling multiple clients efficiently. Delve into continuations, asynchronous programming, and structured concurrency. Gain insights into Quasar as an alternative solution while awaiting Loom's official release.
Overview
Syllabus
Introduction
Project Loom
Use case
First implementation
CPU utilization
Thread creation
Visual DVM
Thread dumb
System call
Socket channel
Blocking calls
Context switch
IANA scheduler
TCP connections open
Second implementation
Threads
Service Threads
Tweed Method
SocketChannel API
asynchronous circuits channel
promote efficiency
handle 200 clients
asynchronous calls
asynchronous implementation
evenloop
Nettie
Reactive streams
Atomic references
User threads
continuations
asynchronous programming
continuation as runnable
continuation as state machine
iterator as state machine
continuation as lightweight thread scheduler
nonblocking
scheduler threads
play with continuations
cosa
automatic tail call optimization
structured concurrency
References
Taught by
Devoxx
Related Courses
-
Untangled - Sneak Peek at Project Loom
-
Project Loom - Fibers and Continuations for Java
-
Project Loom: Lightweight Concurrency with Fibers - Part 2
-
Project Loom - Modern Scalable Concurrency for the Java Platform
-
Project Loom - Helping Write Concurrent Applications on the Java Platform
-
Project Loom - Introducing Fibers for Efficient Concurrent Java Applications
