Kubernetes Classroom Series – 24/Aug/2021

Running Prometheus

• Lets try to run Prometheus on a Linux machine
• Lets create a vm
• Download the linux tar file Refer Here
• Now in the prometheus.yml file ensure you have the following content

# my global config
global:
  scrape_interval: 10s 
scrape_configs:
  - job_name: "prometheus"
    static_configs:
      - targets: ["localhost:9090"]

• Now lets try to run prometheus

• Open the prometheus expression browser and execute up

• There is a single result with value 1 and the name up{instance="localhost:9090", job="prometheus"}.

• up is a special metric added by Prometheus when it performsa scrape and 1 indicates that the scrape was succesful. Then instance is a label, indicating the target wthat was scrapted and the job label here comes from job_name in the prometheus.yml file

• Lets try to download a Node exporter Refer Here

  • untar the exporter and try to run the node exporter with ./node_exporter
  • Now access the metrics from node exporter http://<ip>:9100

• Now lets add some more info to the prometheus.yml to scrape node metrics exported by node exporter

global:
  scrape_interval: 10s 
scrape_configs:
  - job_name: "prometheus"
    static_configs:
      - targets: ["localhost:9090"]
  - job_name: node
    static_configs:
      - targets: ["localhost:9100"]

• and restart prometheus (./prometheus)

• Now lets try to add some alert manager configuration

global:
  scrape_interval: 10s 

alerting:
  alertmanagers:
    - static_configs:
        - localhost:9093
scrape_configs:
  - job_name: "prometheus"
    static_configs:
      - targets: ["localhost:9090"]
  - job_name: node
    static_configs:
      - targets: ["localhost:9100"]

• Now create a new file called as rules.yaml

groups:
  - name: example
    rules:
      - alert: InstanceDown
        expr: up == 0
        for: 1m

• Now Download the alertmanager and configure the alert manager.yaml file