上一篇博文中说到Prometheus有四种指标类型:Counter(计数器)、Gauge(仪表盘)、Histogram(直方图)、Summary(摘要),并且我们做了一个Counter的Demo,接下来看看Histogram。
3、Histogram:直方图
直方图,维基百科的定义:是一种对数据分布情况的图形表示,是一种二维统计图表,它的两个坐标分别是统计样本和该样本对应的某个属性的度量,以长条图(bar)的形式具体表现。因为直方图的长度及宽度很适合用来表现数量上的变化,所以较容易解读差异小的数值。还是拿上一篇的Sample来说明,假如每个订单都有一个金额,在order时在返回值{result=true,data=1000}的data属性中返回,这里,我们就可以用直方图来收集这个金额,由于订单的金额不一样,我们就可以用直方图来展示一定范围金额内订单的数据,监控出一定金额范围内的订单比例了。就是说在一定数量的订单里,少于1000元的有多少个订单,少于2000元有多少个订单,少于3000元的有多少个订单……
首先,我们得修改BusinessController中Order Action的业务逻辑,把订单金额作为返回值:
[HttpGet("/order")] public IActionResult Order(string orderno) { try { _logger.LogInformation("下单"); //返回订单金额 var random = new Random(); return new JsonResult(new { Result = true, data = random.Next(1, 8000) }); } catch (Exception exc) { _logger.LogCritical(exc, exc.Message); return new JsonResult(new { Result = false, Message = exc.Message }); } }
这里的金额为了方便demo,是随机生成一个1到8000的随机数。
需要在MetricsHub.cs中添加Histogram类型的指标收集集合:
using Prometheus; using System.Collections.Generic; namespace PrometheusSample.Middlewares { public class MetricsHub { private static Dictionary<string, Counter> _counterDictionary = new Dictionary<string, Counter>(); private static Dictionary<string, Dictionary<string, Gauge>> _gaugeDictionary = new Dictionary<string, Dictionary<string, Gauge>>(); private static Dictionary<string, Histogram> _histogramDictionary = new Dictionary<string, Histogram>(); public Counter GetCounter(string key) { if (_counterDictionary.ContainsKey(key)) { return _counterDictionary[key]; } else { return null; } } public Dictionary<string, Gauge> GetGauge(string key) { if (_gaugeDictionary.ContainsKey(key)) { return _gaugeDictionary[key]; } else { return null; } } public Histogram GetHistogram(string key) { if (_histogramDictionary.ContainsKey(key)) { return _histogramDictionary[key]; } else { return null; } } public void AddCounter(string key, Counter counter) { _counterDictionary.Add(key, counter); } public void AddGauge(string key, Dictionary<string, Gauge> gauges) { _gaugeDictionary.Add(key, gauges); } public void AddHistogram(string key, Histogram histogram) { _histogramDictionary.Add(key, histogram); } } }
接下来就要在BusinessMetricsMiddleware的中间件中添加处理Histogram指标的代码了:
using Microsoft.AspNetCore.Http; using PrometheusSample.Models; using System.IO; using System.Threading.Tasks; namespace PrometheusSample.Middlewares { /// <summary> /// 请求记录中间件 /// </summary> public class BusinessMetricsMiddleware { private readonly RequestDelegate _next; public BusinessMetricsMiddleware(RequestDelegate next) { _next = next; } public async Task InvokeAsync(HttpContext context, MetricsHub metricsHub) { var originalBody = context.Response.Body; try { using (var memStream = new MemoryStream()) { //从管理返回的Response中取出返回数据,根据返回值进行监控指标计数 context.Response.Body = memStream; await _next(context); memStream.Position = 0; string responseBody = new StreamReader(memStream).ReadToEnd(); memStream.Position = 0; await memStream.CopyToAsync(originalBody); if (metricsHub.GetCounter(context.Request.Path) != null || metricsHub.GetGauge(context.Request.Path) != null) { //这里约定所有action返回值是一个APIResult类型 var result = System.Text.Json.JsonSerializer.Deserialize<APIResult>(responseBody, new System.Text.Json.JsonSerializerOptions { PropertyNameCaseInsensitive = true }); if (result != null && result.Result) { //获取到Counter var counter = metricsHub.GetCounter(context.Request.Path); if (counter != null) { //计数 counter.Inc(); } var gauges = metricsHub.GetGauge(context.Request.Path); if (gauges != null) { //存在增加指标+就Inc if (gauges.ContainsKey("+")) { gauges["+"].Inc(); } //存在减少指标-就Dec if (gauges.ContainsKey("-")) { gauges["-"].Dec(); } } var histogram = metricsHub.GetHistogram(context.Request.Path); if (histogram != null) { var parseResult = int.TryParse(result.Data.ToString(), out int i); if (parseResult) { histogram.Observe(i); } } } } } } finally { context.Response.Body = originalBody; } } } }
再就是在Starsup中配置对应url的Histogram参数了:
using Microsoft.AspNetCore.Builder; using Microsoft.AspNetCore.Hosting; using Microsoft.Extensions.Configuration; using Microsoft.Extensions.DependencyInjection; using Microsoft.Extensions.Hosting; using Microsoft.OpenApi.Models; using Prometheus; using PrometheusSample.Middlewares; using PrometheusSample.Services; using System.Collections.Generic; namespace PrometheusSample { public class Startup { public Startup(IConfiguration configuration) { Configuration = configuration; } public IConfiguration Configuration { get; } public void ConfigureServices(IServiceCollection services) { MetricsHandle(services); services.AddScoped<IOrderService, OrderService>(); services.AddControllers(); services.AddSwaggerGen(c => { c.SwaggerDoc("v1", new OpenApiInfo { Title = "PrometheusSample", Version = "v1" }); }); } public void Configure(IApplicationBuilder app, IWebHostEnvironment env) { if (env.IsDevelopment()) { app.UseDeveloperExceptionPage(); app.UseSwagger(); app.UseSwaggerUI(c => c.SwaggerEndpoint("/swagger/v1/swagger.json", "PrometheusSample v1")); } app.UseRouting(); //http请求的中间件 app.UseHttpMetrics(); app.UseAuthorization(); //自定义业务跟踪 app.UseBusinessMetrics(); app.UseEndpoints(endpoints => { //映射监控地址为 /metrics endpoints.MapMetrics(); endpoints.MapControllers(); }); } /// <summary> /// 处理监控事项 /// </summary> /// <param name="services"></param> void MetricsHandle(IServiceCollection services) { var metricsHub = new MetricsHub(); //counter metricsHub.AddCounter("/register", Metrics.CreateCounter("business_register_user", "注册用户数。")); metricsHub.AddCounter("/order", Metrics.CreateCounter("business_order_total", "下单总数。")); metricsHub.AddCounter("/pay", Metrics.CreateCounter("business_pay_total", "支付总数。")); metricsHub.AddCounter("/ship", Metrics.CreateCounter("business_ship_total", "发货总数。")); //gauge var orderGauge = Metrics.CreateGauge("business_order_count", "当前下单数量。"); var payGauge = Metrics.CreateGauge("business_pay_count", "当前支付数量。"); var shipGauge = Metrics.CreateGauge("business_ship_count", "当前发货数据。"); metricsHub.AddGauge("/order", new Dictionary<string, Gauge> { { "+", orderGauge} }); metricsHub.AddGauge("/pay", new Dictionary<string, Gauge> { {"-",orderGauge}, {"+",payGauge} }); metricsHub.AddGauge("/ship", new Dictionary<string, Gauge> { {"+",shipGauge}, {"-",payGauge} }); //histogram var orderHistogram = Metrics.CreateHistogram("business_order_histogram", "订单直方图。", new HistogramConfiguration { Buckets = Histogram.LinearBuckets(start: 1000, width: 1000, count: 5) }); metricsHub.AddHistogram("/order", orderHistogram); services.AddSingleton(metricsHub); } } }
Histogram.LinearBuckets(start: 1000, width: 1000, count: 5)是金额从1000开始,每1000为一个台阶,一共6个台阶:0~1000,1001~2000,2001~3000,3001~4000,4001~5000,还有一个是大于5000的。
最后一步,就是打开Grafana来配置展示图表了。
订单金额分布图
订单比例分布图
图中histogram_quantile(0.80, sum(rate(business_order_histogram_bucket[5m])) by (le))的意思是“80%的订单金额小于等于这个值”5分钟内的值。
最终展示结果:
聪明的你一定发现,这篇博文与上一篇如出一辙,是的,只是监控指标展示类型不同而以。
想要更快更方便的了解相关知识,可以关注微信公众号