如何在默认端点/指标下覆盖普罗米修斯

Question

I am trying to expose metrics using prometheus httpserver (io.prometheus.client.exporter.HTTPServer) for which i am using below given dependency.

 <dependency>
  <groupId>io.prometheus</groupId>
  <artifactId>simpleclient_httpserver</artifactId>
  <version>0.2.0</version>
</dependency>

Prometheus httpserver by default exposes all the metrics in "/metrics" endpoint and i want to override it and restrict it to expose metrics for any particular endpoint only (like "/prometheusMetrics"). Using below code i running prometheus httpserver and exposing metrics.

@ServiceBean(singleton = true)
public class PrometheusCustomHTTPServer {

private static Logger s_logger = LoggerFactory.getLogger(PrometheusCustomHTTPServer.class);

private HTTPServer httpServer;
 private static CollectorRegistry registry = CollectorRegistry.defaultRegistry;
 private static boolean initailized = false;


@PostConstruct
public void start() {
    if(httpServer == null) {
        String hostAddress;
        try {
            int port = 9090;
            httpServer = new HTTPServer(port,true); 
            s_logger.error("started prometheus at http://"+port);
        } catch (IOException e) {
            System.out.println("Exception occured while starting of prometheus server" + e);
        }
        initialize(registry);
    }
}

private static  synchronized void initialize(CollectorRegistry registryToUse) {
    if(!initailized) {
         new StandardExports().register(registry);
         new MemoryPoolsExports().register(registry); 
          new GarbageCollectorExports().register(registry);
          new ThreadExports().register(registry);
          new ClassLoadingExports().register(registry); 
         System.out.println("StandardExports initialized");
         initailized = true;
    }
}

}

Answer 1

You are changing the client side configuration, but it's not enough. You should also update the scrapping endpoint of Prometheus "server side" configuration.

Take a look into Prometheus Configuration :

The job name assigned to scraped metrics by default. job_name:

# How frequently to scrape targets from this job. [ scrape_interval: | default = ]

Per-scrape timeout when scraping this job. [ scrape_timeout: | default = ]

# The HTTP resource path on which to fetch metrics from targets. [ metrics_path: | default = /metrics ]