Prometheus 是一套开源的系统监控报警框架。它由工作在 SoundCloud 的 员工创建,并在 2015 年正式发布的开源项目。2016 年,Prometheus 正式加入 Cloud Native Computing Foundation,非常的受欢迎。
声明:该小节参考了文章[Prometheus 入门与实践]
Prometheus 生态圈中包含了多个组件,其中许多组件是可选的:
其大概的工作流程是:
在spring boot工程中引入actuator的起步依赖,以及micrometer-registry-prometheus的依赖。
<dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-web</artifactId> </dependency> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-actuator</artifactId> </dependency> <dependency> <groupId>io.micrometer</groupId> <artifactId>micrometer-registry-prometheus</artifactId> </dependency>
暴露prometheus的接口;暴露metrics.tags,和spring.application.name一致。
server: port: 8081 spring: application: name: my-prometheus management: endpoints: web: exposure: include: 'prometheus' metrics: tags: application: ${spring.application.name}
写一个API接口,用作测试。代码如下:
@RestController public class TestController { Logger logger = LoggerFactory.getLogger(TestController.class); @GetMapping("/test") public String test() { logger.info("test"); return "ok"; } @GetMapping("") public String home() { logger.info("home"); return "ok"; } }
在浏览器上访问http://localhost:8081/actuator/prometheus,展示的信息如下,这些信息都是actuator的一些监控信息。
# HELP jvm_gc_max_data_size_bytes Max size of old generation memory pool # TYPE jvm_gc_max_data_size_bytes gauge jvm_gc_max_data_size_bytes{application="my-prometheus",} 2.863661056E9 # HELP http_server_requests_seconds # TYPE http_server_requests_seconds summary http_server_requests_seconds_count{application="my-prometheus",exception="None",method="GET",outcome="CLIENT_ERROR",status="404",uri="/**",} 1.0 http_server_requests_seconds_sum{application="my-prometheus",exception="None",method="GET",outcome="CLIENT_ERROR",status="404",uri="/**",} 0.018082327 # HELP http_server_requests_seconds_max # TYPE http_server_requests_seconds_max gauge http_server_requests_seconds_max{application="my-prometheus",exception="None",method="GET",outcome="CLIENT_ERROR",status="404",uri="/**",} 0.018082327 # HELP jvm_threads_states_threads The current number of threads having NEW state # TYPE jvm_threads_states_threads gauge jvm_threads_states_threads{application="my-prometheus",state="waiting",} 12.0 jvm_threads_states_threads{application="my-prometheus",state="runnable",} 8.0 jvm_threads_states_threads{application="my-prometheus",state="timed-waiting",} 2.0 jvm_threads_states_threads{application="my-prometheus",state="terminated",} 0.0 jvm_threads_states_threads{application="my-prometheus",state="blocked",} 0.0 jvm_threads_states_threads{application="my-prometheus",state="new",} 0.0 # HELP process_files_open_files The open file descriptor count # TYPE process_files_open_files gauge ...省略更多
安装Prometheus很简单,在linux系统上安装,执行以下的安装命令。其他的操作系统,比如windows、mac等在官网上(https://prometheus.io/download/)下载并安装。
wget https://github.com/prometheus/prometheus/releases/download/v2.19.2/prometheus-2.19.2.darwin-amd64.tar.gz tar xvfz prometheus-*.tar.gz cd prometheus-*
修改Prometheus的配置文件prometheus.yml,代码如下:
global: scrape_interval: 15s # By default, scrape targets every 15 seconds. # Attach these labels to any time series or alerts when communicating with # external systems (federation, remote storage, Alertmanager). external_labels: monitor: 'codelab-monitor' # A scrape configuration containing exactly one endpoint to scrape: # Here it's Prometheus itself. scrape_configs: # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config. - job_name: 'prometheus' # Override the global default and scrape targets from this job every 5 seconds. scrape_interval: 5s static_configs: - targets: ['localhost:9090'] - job_name: 'springboot_prometheus' scrape_interval: 5s metrics_path: '/actuator/prometheus' static_configs: - targets: ['127.0.0.1:8081']
./prometheus --config.file=prometheus.yml
多次请求springboot项目的接口http://localhost:8081/test , 并访问prometheus的控制台http://localhost:9090/,展示的界面如下:
prometheus提供了一些可视化图,比如使用柱状图来展示每秒请求数:
grafana 是一款采用 go 语言编写的开源应用,主要用于大规模指标数据的可视化展现,是网络架构和应用分析中最流行的 时序数据展示 工具,目前已经支持绝大部分常用的时序数据库。使用grafana去展示prometheus上的数据。先安装,安装命令如下:
wget https://dl.grafana.com/oss/release/grafana-7.0.4.darwin-amd64.tar.gz tar -zxvf grafana-7.0.4.darwin-amd64.tar.gz ./grafana-server
访问http://localhost:3000/,初始密码:admin/admin。
配置数据源,如图:
配置prometheus的地址:http://localhost:9090 ,如图所示:
在dashboard界面新建panel,展示的metrics为http_server_request_seconds_count,即展示了以时间为横轴,请求数为纵轴的请求曲线,如图所示:
Prometheus 入门与实践