Question: The Nautilus DevOps team is planning to set up a Nagios monitoring tool to monitor some applications, services etc. They are planning to deploy it on Kubernetes cluster. Below you can find more details.
1) Create a deployment nagios-deployment for Nagios core. The container name must be nagios-container and it must use jasonrivers/nagios image.
2) Create a user and password for the Nagios core web interface, user must be xFusionCorp and password must be LQfKeWWxWD. (you can manually perform this step after deployment)
3) Create a service nagios-service for Nagios, which must be of targetPort type. nodePort must be 30008.
You can use any labels as per your choice.
Note: The kubectl on jump_host has been configured to work with the kubernetes cluster.
Solution:
Create deployment.yaml and service.yaml according to question
vi deployment.yaml
vi service.yaml
Then apply
kubectl apply -f .
Next, exec into pod
kubectl exec -it pod-name /bin/bash
Create the username and password according to question
Finally, Click on Nagios button from the right side. You can see the nagios dashboard.
0 comments:
Post a Comment