Google Cloud with Services already running

Use PowerShell, so you don't need to copy paste the names of the pods.

1. Log in from your command line with

   gcloud auth login

2. Connect: at https://console.cloud.google.com/kubernetes click on the 3 dots and select connect, copy the command-line statement and run it in your PowerShell

3. Check if the node is running

   kubectl get nodes

4. Check if pods are running with

   kubectl get pods --namespace monitoring

5. Forward the port of prometheus-server

   kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=prometheus-server" --output jsonpath='{.items[0].metadata.name}') 8080:9090

   and open localhost:8080

6. Open a new PowerShell and forward the port of alertmanager

   kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=alertmanager" --output jsonpath='{.items[0].metadata.name}') 8081:9093

   and open localhost:8081

7. Open a new PowerShell and forward the port of grafana

   kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=grafana" --output jsonpath='{.items[0].metadata.name}') 8082:3000

   and open localhost:8082

8. Login with
   Username: admin
   Password: admin
   Skip setting new password

9. Open a new PowerShell and forward the port of Application1

   kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=application1" --output jsonpath='{.items[0].metadata.name}') 8083:8000

   and open for example localhost:8083/delay/2 for a 2 seconds delay

10. Open a new PowerShell and forward the port of Application2

    kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=application2" --output jsonpath='{.items[0].metadata.name}') 8084:8001

    and open localhost:8084/ -> the Service should be Unavailable

11. Open a new PowerShell and open CLI for Redis

    kubectl exec -i -t $(kubectl get pod --namespace monitoring --selector="app=redis" --output jsonpath='{.items[0].metadata.name}') -n monitoring --container redis -- redis-cli

12. Open a new PowerShell and forward the port of Redis

    kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=redis" --output jsonpath='{.items[0].metadata.name}') 8085:9121

    and open localhost:8085/metrics

Overview of the ports (localhost:xxxx):

• 8080 -> prometheus-server
• 8081 -> alertmanager
• 8082 -> grafana
• 8083 -> Application1 (= Sleep)
• 8084 -> Application2 (= Not available)
• 8085 -> Redis

First setup with Google Cloud

Prerequisites

1. Account in Google Cloud and start a trial version
2. Install Google Cloud SDK Run the command:

   gcloud version

3. If not done before: run the kubectl installation command

   gcloud components intall kubectl

Setup Cluster

1. Log in from your command line with

   gcloud auth login

2. Open the Kubernetes Engine Overview

3. Click on Create Cluster and select "Standard: you manage your cluster"

4. Create a cluster with the default settings except for the Number of nodes where you only need 1:

   

   Standard machine type is

   Changed the Machine type to e2-standard-4 so CPU should be no problem in future

5. As soon as your cluster is ready, click on Connect and copy and paste this command into your terminal

6. Now your kubectl (i.e., the Kubernetes command-line tool) should be configured for your cluster. In order to verify this, execute the command:

   kubectl get nodes

7. Change to the directory in the git folder and run the commands for Setup Prometheus from the setup_prometheus_example.md

8. Check if it is running with

   kubectl get pods --namespace monitoring

9. Forward the port

   kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=prometheus-server" --output jsonpath='{.items[0].metadata.name}') 8080:9090

10. Now open the link localhost:8080 in your browser.

11. Continue the setup_prometheus_example.md with Setup State Metrics and Setup Alert manager

12. Open a new PowerShell and forward the port of alertmanager

    kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=alertmanager" --output jsonpath='{.items[0].metadata.name}') 8081:9093

    and open localhost:8081

13. Continue the setup_prometheus_example.md with Setup Grafana. For setup the Grafana Dashboards see grafana_dashboard.md

14. Open a new PowerShell and forward the port of grafana

    kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=grafana" --output jsonpath='{.items[0].metadata.name}') 8082:3000

    and open localhost:8082

15. Login with
    Username: admin
    Password: admin
    Skip setting new password

16. Continue the setup_prometheus_example.md with Setup Node Exporter and Setup Application1 (Sleep)

17. Open a new PowerShell and forward the port of Application1

    kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=application1" --output jsonpath='{.items[0].metadata.name}') 8083:8000

    and open for example localhost:8083/delay/2 for a 2 seconds delay

18. Continue the setup_prometheus_example.md with Setup Application2 (Not available)

19. Open a new PowerShell and forward the port of Application2

    kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=application2" --output jsonpath='{.items[0].metadata.name}') 8084:8001

    and open localhost:8084/ -> the Service should be Unavailable

20. Continue the setup_prometheus_example.md with Setup Redis

21. Open a new PowerShell and forward the port of Redis

    kubectl port-forward --namespace monitoring $(kubectl get pod --namespace monitoring --selector="app=redis" --output jsonpath='{.items[0].metadata.name}') 8085:9121

    and open localhost:8085/metrics

---

For updating a Deployment.yaml run for example

kubectl edit deployment alertmanager --namespace monitoring

For deleting (= restarting) -> change [POD-NAME]

kubectl delete pod $(kubectl get pod --namespace monitoring --selector="app=[POD-NAME]" --output jsonpath='{.items[0].metadata.name}') -n monitoring