bosima / fireflysoft.ratelimit

It is a rate limiting library based on .Net standard.

FireflySoft.RateLimit　　　　　　　　　　　中文

• FireflySoft.RateLimit　　　　　　　　　　　中文
  • Introduction
  • Features
  • Projects
  • Usage
    • ASP.NET Core
    • ASP.NET
    • Others

Introduction

Fireflysoft.RateLimit is a rate limiting library based on .Net standard. Its core is simple and lightweight, and can flexibly meet the rate limiting needs of many scenarios.

Features

• Multiple rate limiting algorithms: built-in fixed window, sliding window, leaky bucket, token bucket, and can be extended.
• Multiple counting storage: memory and Redis (including cluster).
• Distributed friendly: supports unified counting of distributed programs with Redis storage.
• Flexible rate limiting targets: each data can be extracted from the request to set rate limiting targets.
• Support rate limit penalty: the client can be locked for a period of time after the rate limit is triggered.
• Time window enhancement: support to the millisecond level; support starting from the starting point of time periods such as seconds, minutes, hours, dates, etc.
• Real-time tracking: the number of requests processed and the remaining allowed requests in the current counting cycle, as well as the reset time of the counting cycle.
• Dynamically change the rules: support the dynamic change of the rate limiting rules when the program is running.
• Custom error: you can customize the error code and error message after the current limit is triggered.
• Universality: in principle, it can meet any scenario that requires rate limiting.

Projects

Project	Descriptioin
FireflySoft.RateLmit.Core	algorithm, rules, persistence and other core codes.
FireflySoft.RateLimit.AspNet	ASP.NET rate-limit middleware based on .NET Framework.
FireflySoft.RateLimit.AspNetCore	ASP.NET Core rate-limit middleware.
FireflySoft.RateLimit.Core.UnitTest	Unit test for FireflySoft.RateLimit.Core.
FireflySoft.RateLimit.Core.BenchmarkTest	Benchmark test for FireflySoft.RateLimit.Core.
Samples/Console	FireflySoft.RateLmit.Core sample program.
Samples/AspNet	FireflySoft.RateLimit.AspNet sample program.
Samples/AspNetCore	FireflySoft.RateLimit.AspNetCore sample program.
Samples/RuleAutoUpdate	A sample that can automatic update rate limiting rules.

Usage

ASP.NET Core

1、Install Nuget Package

Package Manager:

Install-Package FireflySoft.RateLimit.AspNetCore

Or .NET CLI:

dotnet add package FireflySoft.RateLimit.AspNetCore

Or Project file：

<ItemGroup>
<PackageReference Include="FireflySoft.RateLimit.AspNetCore" Version="2.*" />
</ItemGroup>

2、Use Middleware

The following code calls the rate-limit middleware from Startup.Configure:

public void ConfigureServices(IServiceCollection services)
{
    ...

    services.AddRateLimit(new InProcessFixedWindowAlgorithm(
        new[] {
            new FixedWindowRule()
            {
                ExtractTarget = context =>
                {
                    return (context as HttpContext).Request.Path.Value;
                },
                CheckRuleMatching = context =>
                {
                    return true;
                },
                Name="default limit rule",
                LimitNumber=30,
                StatWindow=TimeSpan.FromSeconds(1)
            }
        })
    );

    ...
}

public void Configure(IApplicationBuilder app, IWebHostEnvironment env)
{
    ...

    app.UseRateLimit();

    ...
}

ASP.NET

1、Install Nuget Package:

Package Manager:

Install-Package FireflySoft.RateLimit.AspNet

2、Register MessageHandler

Open Global.asax.cs, the following code adds the rate limit message handle:

protected void Application_Start()
{
    ...

    GlobalConfiguration.Configuration.MessageHandlers.Add(
        new RateLimitHandler(
            new Core.InProcessAlgorithm.InProcessFixedWindowAlgorithm(
                new[] {
                    new FixedWindowRule()
                    {
                        ExtractTarget = context =>
                        {
                            return (context as HttpRequestMessage).RequestUri.AbsolutePath;
                        },
                        CheckRuleMatching = context =>
                        {
                            return true;
                        },
                        Name="default limit rule",
                        LimitNumber=30,
                        StatWindow=TimeSpan.FromSeconds(1)
                    }
                })
        ));

    ...
}

Others

1、Install Nuget Package

Package Manager:

Install-Package FireflySoft.RateLimit.Core

Or .NET CLI:

dotnet add package FireflySoft.RateLimit.Core

2、Use IAlgorithm

Use IAlgorithm to filter every request, process the return value of Check method.

// Rule
var fixedWindowRules = new FixedWindowRule[]
    {
        new FixedWindowRule()
        {
            Id = "3",
            StatWindow=TimeSpan.FromSeconds(1),
            LimitNumber=30,
            ExtractTarget = (request) =>
            {
                return (request as SimulationRequest).RequestResource;
            },
            CheckRuleMatching = (request) =>
            {
                return true;
            },
        }
    };

// Algorithm
IAlgorithm algorithm = new InProcessFixedWindowAlgorithm(fixedWindowRules);

// Check
var result = algorithm.Check(new SimulationRequest()
    {
        RequestId = Guid.NewGuid().ToString(),
        RequestResource = "home",
        Parameters = new Dictionary<string, string>() {
                    { "from","sample" },
            }
    });

SimulationRequest is a custom request that you can modify to any type.

fireflysoft.ratelimit's Issues

Dynamically change the flow limiting rule configuration at program runtime

我看Sample项目都是直接在服务里面注入的，如何支持在程序运行时动态更改限流算法的参数和配置呢

DoLeakyBucketWait Issue when checkResult.RuleCheckResults.Count() == 0 , if I add some filter or Middleware

private static async Task DoLeakyBucketWait(AlgorithmCheckResult checkResult)
{
if (checkResult.RuleCheckResults.Count() == 0)
return;
// Simulation leaky bucket algorithm queuing mechanism
var wait = checkResult.RuleCheckResults.Max(d => d.Wait);
if (wait > 0)
{
await Task.Delay((int)wait).ConfigureAwait(false);
}
}

是否可以实现按照IP进行限制

是否可以实现按照ip进行限制？

FireflySoft.RateLimit.AspNetCore.Sample project startup fail

be cause AddRateLimit inner
builder.AddSingleton(interceptor);
interceptor Default value cannot be null

同一个程序是否支持多个限流算法

期望使用固定窗口算法来控制ip某个时间端不要超过特定数量
期望使用令牌桶算法来控制访问的限流
有什么方法可以同时注册使用这两种算法吗？

请问可以用这个来限制httprequest请求的下载速度吗

rt

LeakyBucket bug when the first period has over outflow_quantity_per_unit quantity let in bucket

asume that the bucket
capacity : 20
outflow_unit: 500 ms
outflow_quantity_per_unit: 5
amount: 1

in the first period ( the first outflow unit: the first 500ms), has 25 requests, end of the period, 5 requests pass, and 20 requests in the buecket.
in the second period, in the theory, also has 5 requests pass (wait is 0, ret = [0, x, 0]), but actually, in the second period all request all return wait > 0

that is: when capacity >= outflow_quantity_per_unit * n (n >= 2), and the first period let in the bucket quantity greater than outflow_quantity_per_unit , then cause the problem.
because the below sentence:
if(current_value>outflow_quantity_per_unit)

在ASP.NET MVC 项目中如何注册限流处理器

很好的库，已经在ASP.NET WEB API 中实现限流。可否告知一下如何在ASP.NET MVC项目模版中注册限流组件。

Redis storage have a defect

这是令牌桶算法方式在redis的存储，这个key的剩余令牌桶数量不会自动更新，只有当每次程序访问的时候才能去更新key剩余的容量还有多少，this one of tow defect,

每次新的请求都会重新创建一个Key，而且这个key不会自动过期，也就是说随着程序运行的越来越久，访问量越来越高，那么redis的key也会越来越多，只能去手动维护？

CheckRuleMatching return false Will it take effect

还有个问题，就是实例化限流规则类的 自定义CheckRuleMatching 委托规则 如果返回false，是不是应该就是不对改规则进行限流，可是实际上貌似没什么作用，甚至还抛出一个异常

Defects of redis storage

这是令牌桶算法方式在redis的存储，这个key的剩余令牌桶数量不会自动更新，只有当每次程序访问的时候才能去更新key剩余的容量还有多少，this one of tow defect,

每次新的请求都会重新创建一个Key，而且这个key不会自动过期，也就是说随着程序运行的越来越久，访问量越来越高，那么redis的key也会越来越多，只能去手动维护？

FixedWindowRule可以支持动态往容器内增加可用数量么？

您好，

我有这样的一个场景，控制10分钟内只能有1000个用户进入系统，在这期间如果进入的用户有退出的话，我想递增可进入人数

假设：
StatWindow=TimeSpan.FromSeconds(10*60),
LimitNumber=1000,

一分钟之后，剩余可以登录的用户数为800，此时有10个用户退出

我想把当前剩余可以登录的用户数置为810

请问这种在程序运行过程中修改剩余数量的场景支持吗？或者说未来打算支持么？