25-01-2024- Shai Zambrovski
In a typical Spring application, scheduling tasks is a powerful feature to automate recurring processes.
However, there are scenarios where developers might seek to limit the number of executions for a scheduled task, a feature not directly supported by Spring’s APIs.
The motivation behind such a limitation could stem from specific use cases where the task’s execution should be restricted to a fixed number of occurrences.
For instance, in scenarios where resource constraints or business requirements dictate that a particular process should only run a predetermined number of times.
While Spring provides robust support for task scheduling, it encourages developers to implement custom logic within the scheduled method to control the number of executions programmatically.
This approach ensures flexibility and allows developers to tailor the scheduling behavior according to the unique requirements of their applications.
A wrapper class is a class that contains and manages another class or object, often providing additional functionality or abstraction.
In the context of scheduling tasks with a fixed number of executions, a wrapper class can encapsulate a ScheduledFuture<?> and a TaskScheduler.
The FixedExecutionTaskWrapper class will implements the ISchedulerWrapper interface, which includes the cancelSchedulerTask() and startSchedulerTask(long delay) methods.
These methods provide control over canceling and starting the scheduler task with a specified delay.
This wrapper class can implement the Runnable interface to achieve the desired behavior. Let’s create a simple example:
public interface ISchedulerWrapper {
void startSchedulerTask(long delay);
void cancelSchedulerTask();
}
@RequiredArgsConstructor
public class FixedExecutionTaskWrapper implements ISchedulerWrapper, Runnable {
private static final Logger logger = Logger.getLogger(FixedExecutionTaskWrapper.class.getName());
private ScheduledFuture<?> scheduledFuture;
private int currentCounter;
private final TaskScheduler taskScheduler;
private final int maxRetries;
@Override
public void startSchedulerTask(long delay) {
logger.info("Register scheduler task");
this.scheduledFuture = taskScheduler.scheduleWithFixedDelay(
this,
delay
);
}
@Override
public void cancelSchedulerTask() {
logger.info("Cancel scheduler task");
if (this.scheduledFuture != null) {
this.scheduledFuture.cancel(true);
this.scheduledFuture = null;
}
}
@Override
public void run() {
if (currentCounter < maxRetries) {
logger.info("iteration " + currentCounter + " out from " + maxRetries);
// do your logic ...
} else {
this.cancelSchedulerTask();
currentCounter = 0;
return;
}
currentCounter++;
}
}
The FixedExecutionTaskWrapper class is designed to manage a scheduled task with a fixed number of iterations or retries.
It implements the ISchedulerWrapper interface and the Runnable interface.
The class uses a TaskScheduler for scheduling tasks and allows the task to be started with a specified delay through the startSchedulerTask method.
The run method is executed on each iteration, logging information about the current iteration count and canceling the scheduled task if the maximum number of retries (maxRetries) is reached.
The cancelSchedulerTask method is provided to manually cancel the scheduled task.
This wrapper class provides a way to control the number of iterations for a scheduled task by encapsulating the original ScheduledFuture<?> and TaskScheduler instances.
It demonstrates a dynamic approach to scheduling tasks with a fixed number of executions.
@Configuration
public class SchedulerConfiguration {
@Value("${scheduler.task.max.retries:6}")
private int schedulerTaskMaxRetries;
@Bean
public TaskScheduler schedulerTaskScheduler() {
ThreadPoolTaskScheduler taskScheduler = new ThreadPoolTaskScheduler();
taskScheduler.setPoolSize(1); // Set the number of threads in the pool
taskScheduler.setThreadNamePrefix("scheduler-task-");
taskScheduler.setAwaitTerminationSeconds(60); // Set the time to wait for scheduled tasks to finish
taskScheduler.setWaitForTasksToCompleteOnShutdown(true);
return taskScheduler;
}
@Bean
ISchedulerWrapper FixedExecutionTaskWrapper(TaskScheduler schedulerTaskScheduler) {
return new FixedExecutionTaskWrapper(schedulerTaskScheduler, schedulerTaskMaxRetries);
}
}
The SchedulerConfiguration class is a Spring configuration class responsible for defining and configuring beans related to task scheduling.
It reads a property scheduler.task.max.retries with a default value of 6. It creates a TaskScheduler bean using ThreadPoolTaskScheduler and sets properties such as pool size and thread prefix for task execution.
Additionally, it defines a bean of type ISchedulerWrapper named FixedExecutionTaskWrapper, injecting the configured TaskScheduler and the maximum number of retries into the FixedExecutionTaskWrapper constructor.
This configuration class centralizes the setup for scheduling tasks and provides a flexible way to manage scheduled tasks with a specified maximum retry count.
@SpringBootApplication
public class Application implements CommandLineRunner {
private static final Logger logger = Logger.getLogger(Application.class.getName());
@Autowired
private ISchedulerWrapper schedulerWrapper;
@Value("${scheduler.task.delay:1000}")
private long schedulerTaskMaxDelay;
public static void main(String[] args) {
SpringApplication.run(Application.class, args);
}
@Override
public void run(String... args) {
logger.info("Application started...");
schedulerWrapper.startSchedulerTask(schedulerTaskMaxDelay);
}
}
In the last part of our demonstration, the Application class serves as the entry point for the Spring Boot application.
It implements CommandLineRunner (Spring Boot interface that provides a convenient way to execute code after the application context is fully initialized, allowing the execution of custom logic or tasks on application startup), and upon startup, it logs an info message confirming the application’s start.
The class injects an ISchedulerWrapper bean and starts a scheduled task with a specified delay (defaulting to 1000 milliseconds) using the injected scheduler wrapper.
The above demo can be run after you clone my repository