Thread Pool Tests

Introduction

The main goal of this project is to study how threads works in Java and how can be optimized. All the study are done with endpoints, and jmeter.

In the middle of the request process, I coded a Thread.sleep in order to randomize the execution.

Technologies

Java 17
Spring Boot 3.0.4

Synchronous

In the first file threads_1.txt we can see that in default configurations java runs 200 at same time. In the jmeter file, we started with 300 threads. But when the applications achieved 200 threads at the same time, stopped to process until a thread will be free.

The longest process: Start time - End time = 10.017985900698633 (sec)

Asynchronous

threads_2.txt

In the file threads_2.txt we can see that all the 300 threads executed instantly, but the amount of threads that executed in parallel are very low.

The longest process: Start time - End time = 16.18498095989369 (sec)

This happened without any configuration of any Thread Pool.

Spring JAVA DOC Api: ThreadPoolTaskExecutor

The default configuration is a core pool size of 1, with unlimited max pool size and unlimited queue capacity. This is roughly equivalent to Executors.newSingleThreadExecutor(), sharing a single thread for all tasks. Setting "queueCapacity" to 0 mimics Executors.newCachedThreadPool(), with immediate scaling of threads in the pool to a potentially very high number. Consider also setting a "maxPoolSize" at that point, as well as possibly a higher "corePoolSize" (see also the "allowCoreThreadTimeOut" mode of scaling). NOTE: This class implements Spring's TaskExecutor interface as well as the Executor interface, with the former being the primary interface, the other just serving as secondary convenience. For this reason, the exception handling follows the TaskExecutor contract rather than the Executor contract, in particular regarding the TaskRejectedException. For an alternative, you may set up a ThreadPoolExecutor instance directly using constructor injection, or use a factory method definition that points to the Executors class. To expose such a raw Executor as a Spring TaskExecutor, simply wrap it with a ConcurrentTaskExecutor adapter.

threads_3.txt

With the following configurations of the Thread Pool in order to execute in @Async:

    @Bean(name = "threadPoolTaskExecutor1")
    public Executor threadPoolTaskExecutor1() {
        var executor = new ThreadPoolTaskExecutor();
        executor.setCorePoolSize(1);
        executor.setMaxPoolSize(1);
        executor.setQueueCapacity(1000);
        executor.setThreadNamePrefix("threadPoolTaskExecutor1-");
        return executor;
    }

We can see in the file that we initialize all the tasks, and the executed one at a time. Ending: Start time - End time = 69370 (millis) for just 10 parallel requests.

threads_4.txt

With:

    @Bean(name = "threadPoolTaskExecutor1")
    public Executor threadPoolTaskExecutor1() {
        var executor = new ThreadPoolTaskExecutor();
        executor.setCorePoolSize(100);
        executor.setQueueCapacity(1000);
        executor.setThreadNamePrefix("threadPoolTaskExecutor1-");
        return executor;
    }

means that setMaxPoolSize get INT.MAX_VALUE, and setCorePoolSize that is the number of threads of the pool, finished the whole process in: Start time - End time = 17.55855784541837 (sec) for 300 parallel requests.

mdymen85 / threadpool