Multi-threading Lecture 3: Thread Local

Introduction

• Third video in multi-threading series.
• Focus: ThreadLocal in Java applications.
• Recommended to watch previous videos for foundational understanding.
• Next video will cover virtual threads and more.

Scenario and Benefits

• Scenario: Handling multiple user sessions in a web application.
• Example: 100 users create 100 threads; 1 million users create 1 million threads.
• Need to associate a user ID with each thread.

Challenges with Shared Variables

• Using shared variables leads to:
  • Synchronization issues.
  • Complexity with volatile keywords, synchronized blocks, atomic classes.

ThreadLocal Benefits

• Java's ThreadLocal class allows each thread to store its own local copy of a variable.
• Ensures thread safety as data isn't shared between threads.
• Improves performance by avoiding synchronization overhead.

How ThreadLocal Works

• Creates variables unique to each thread.
• Useful for user sessions, database connections, thread contexts.
• Each thread holds its own variable copy, reducing synchronization needs.
• Be cautious of memory leaks if references are stored too long.

Code Example

• Setup: Create ThreadLocal of Long type for user IDs.
• Actions:
  • set: Assigns user ID to thread.
  • get: Retrieves own thread's user ID.
  • remove: Clears the variable to prevent memory leaks.

Code Walkthrough

• Example of setting user ID for threads.
• Shows thread-specific storage and retrieval.
• Demonstrates null result after remove method.

InheritableThreadLocal

• Used when child threads need to inherit values from parent threads.
• Example:
  • Parent sets value in InheritableThreadLocal.
  • Child can access value using get method.
• Regular ThreadLocal is limited to the setting thread.

Conclusion

• ThreadLocal simplifies handling thread-specific data.
• Always implement remove to manage resources.
• Use InheritableThreadLocal for parent-child thread inheritance.

Next Steps

• Upcoming video on virtual threads in Java 21.
• Differences between virtual threads and platform threads.
• Subscribe for more content on programming.