Rate limiter with sliding window algorithm

Why need another rate limiter?

• this rate limiter is based on the algorithm sliding window
• not found any solution in popular libraries (bucket4j, resilience4j, guava)

Briefly about the problem

Some resources (example1 , example2) has some limits for accept user requests.

So application need to know (fast, without sent packets to any additional system) - can they send request or wait or skip

How to use

add dependency for Maven

<dependency>
    <groupId>io.github.axel-n</groupId>
    <artifactId>limiter-sliding-window</artifactId>
    <version>0.3</version>
</dependency>

add dependency for Gradle

compile "io.github.axel-n:limiter-sliding-window:0.3"

Usage

examples availble in another repository

with annotation

@LimiterConfig(
        instanceName = "common", // name of limiter config. you can use several limiters
        maxTimeWait = @TimeConfig(value = 10, interval = TimeUnit.SECONDS), // optional
        limitType = ExecutionLimitType.EXECUTE_OR_WAIT // also available EXECUTE_OR_THROW_EXCEPTION
)
public void limiterWrapperService() {
  // call some your service
}

manual check limiter and execute

// use executeOrWait or executeOrThrowException
limiter.executeOrWait(() -> { 
  // run runnable or callable
});

How it works sliding window algorithm

Plans

• integrate deploy to maven repository
• add example without spring boot
• add methods for run something with limiter wrapper
• add tests in concurrency execution
• add example with spring boot
• add usage with annotation
• increase code coverage
• load values from values in specific environments (like slow github runner)
• add ability setup several limiters (for example 30 requests per 1 minute and 5 requests per second)
• add integration with monitoring
• add normal logging