Skip to content

A simple but efficient C++ thread/worker pool library for asynchronous task management.

License

Notifications You must be signed in to change notification settings

geru-scotland/ThreadPoolLib

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

47 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ThreadPoolLib

ThreadPoolLib is a minimalist C++ thread pool library, designed with simplicity and efficiency in mind. It provides an easy and straightforward way to manage and execute tasks concurrently in a multi-threaded environment, and is an essential tool for developers working on high-performance applications.

Features

  • Ease of Use: ThreadPoolLib allows developers to easily add tasks with callbacks to be processed concurrently, through a simple and intuitive interface.

  • Efficiency: ThreadPoolLib leverages modern C++ features and best practices to ensure that your tasks are processed as efficiently as possible. It uses rvalue references and move semantics to avoid unnecessary copies, and manages all thread-related resources for you to eliminate the risk of common threading pitfalls such as zombie threads.

  • Performance: The library utilizes C++ std::thread to create and manage threads. It optimizes the number of threads created based on the number of cores available, ensuring optimal utilization of hardware resources.

Updates

Improved performance and flexibility with std::unique_lock

  • Use of std::unique_lock: To provide a finer-grained control over locking and synchronization, the library has been updated to use std::unique_lock. This provides greater flexibility compared to the previous std::lock_guard, allowing for unlocking and re-locking a mutex within the same scope. This flexibility is especially useful in facilitating lengthy operations without holding the lock, thus reducing contention and enhancing performance. The implementation now leverages std::condition_variable and its wait() method, which puts threads to sleep when there are no tasks to process, rather than constantly polling for new tasks. This helps to reduce CPU usage. Additionally, the introduction of a predicate function with wait(), manages the wake-up conditions for the sleeping threads, ensuring that they only awake when there are tasks to be consumed. This results in a more efficient and responsive system that optimizes task handling and thread utilization.

Callbacks and Result Handling

Instead of simply adding and running tasks concurrently, ThreadPoolLib supports advanced use cases through task callbacks and result handling. Here's what you can do:

  • Task Callbacks: Add a callback function that will be invoked once a task has completed its execution. This can be useful for handling results, logging, or chaining tasks.

  • Result Handling: When a task finishes execution, it might return a result. With ThreadPoolLib, you can easily handle these results within the callback function, allowing you to use or store the outcome of the task.

Future Enhancements

This library is under active development and future enhancements include:

  • Thread Metrics and Visualization: In the interest of providing valuable insights about the efficiency of the tasks and the overall performance of the applications, I am planning to introduce a feature that collects and exposes thread metrics. These metrics will enable users to visualize and analyze the behavior of their thread pools over time using tools such as Grafana and InfluxDB.
  • Task dependency: (suggested by @tugrul512bit here It will consist in creating dependencies between tasks, so one task can not be consumed until all of its dependencies are resolved, generating this way graphs of tasks, such as:
task 1 <--- (task 2 + task 3)
 |
 |
 V 
task 4 + task 5 ----> task 6

Building

This project uses cmake as a build system.

mkdir build && cd build
cmake ..
make

Basic usage

Include include/ThreadPool.h and create a ThreadPool object by specifying the number of threads you want to use. Then you can add tasks using the overloaded CreateTask() method from the pool object. The Examples mentioned here refers to a namespace containing some trivial sample functions and methods.

// Create a unique pointer to the pool
std::unique_ptr<ThreadPool> pool = std::make_unique<ThreadPool>(std::thread::hardware_concurrency());

// Shared pointer for the tasks:
std::shared_ptr<Task> task1, task2, task3, task4;
  • Example 1:
task1 = pool->CreateTask(normalFunction, normalCallback);
  • Example 2:
task2 = pool->CreateTask(
    [](){
        printf("\n Lambda main function \n");
    },
    []() {
        printf("\n Lambda callback \n");
    }
);
  • Example 3:
std::tuple<int, int, int> args = std::make_tuple(2, 555, 999);
task3 = pool->CreateTask(normalFunctionParams, normalCallbackParams, args);
  • Example 4:
Examples::Foo fooObj1;
auto fooObj2 = std::make_shared<Examples::Foo>();

Task task4;
task4(
        [&fooObj1, &fooObj2]() {
            printf("Lambda Task %i", fooObj1.MyTask(9));
            fooObj2->MyTask(1399);
        },
        [&fooObj1](){
            fooObj1.MyCallback(fooObj1.MyTask(4 * 2));
        }
);

pool.AddTask(std::move(task4));
  • Note: The testing mode is enabled in the CMakeLists.txt file. To use the library without the testing mode, simply comment out the line:
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DDEBUG")

Docker

To run this code using Docker, you first need to build the Docker images and then run the container. Here are the necessary steps:

  1. First, build the base image with the following command:
docker build -t base-image -f docker/base/Dockerfile .
  1. Then, build the application image with the following command:
docker build -t threadlibpool-image -f docker/app/Dockerfile .
  1. Finally, run the application in a Docker container with the following command:
docker build -t threadlibpool-image -f docker/app/Dockerfile .

License

This project is licensed under the GPL 3 License - see the LICENSE.md file for details.