如果有多个导出器，如何区分 Prometheus 中的警报规则

Question

I want to monitor docker containers running on multiple servers lets say i have a,b servers and containers running inside them, now I add one server (d) I want to monitor all docker containers inside all servers (A,B) only from server c. I have configured the docker to expose logs on all servers followed this docker docs not using cAdvisor. The target status shows 'ok' on all the servers, but the problem is as expression is same for all the containers of docker Prometheus is not able differentiate between the servers can anyone share the sample Prometheus rule file with expression ie number of stopped containers should not be less then x.This is my current rule file

groups:

- name: Server_A
  rules:
  - alert: Central_service_down
    expr: engine_daemon_container_states_containers{state="running"} < 10
    for: 50s
    labels:
      severity: critical
      instance: <IP_of_A>:9323
    annotations:
      summary: "Monitor service non-operational"
      description: "Demo Service {{ $labels.instance }} is down."
- name: Server_B
  rules:
  - alert: Central_service_down
    expr: engine_daemon_container_states_containers{state="running"} < 10
    for: 50s
    labels:
      severity: critical
      instance: <IP_of_B>:9323
    annotations:
      summary: "Monitor service non-operational"
      description: "Demo Service {{ $labels.instance }} is down."

as u can see expr: engine_daemon_container_states_containers{state="running"} < 10 is same for both server a and b how can i differentiate expr for both. please share sample alert file.Thanks in advance

Answer 1

I have added instance='ip' to differentiate ie expr: engine_daemon_container_states_containers{instance="serverA",state="running"} < 10