2

I used DCOS with marathon. Every docker application provides metrics on /prometheus endpoint. I need to separate metrics for every docker instance. I tried to deploy prometheus server with config:

scrape_configs:
  - job_name: fodel_api
    metrics_path: /prometheus
    scheme: http
    static_configs:
    - targets: ['external-api.test.com']
    basic_auth:
      username: user
      password: password

And I'm trying this:

- job_name: service-marathon
    metrics_path: /prometheus
    basic_auth:
      username: user
      password: password
    marathon_sd_configs:
    - servers:
      - 'http://marathon.l4lb.thisdcos.directory:443'

But these jobs send result in format like: jvm_memory_used{app="MY_API",id="Code Cache",instance="external-url.test.com:80"}. How to distinguish nodes?

Nikita
  • 4,435
  • 3
  • 24
  • 44

0 Answers0