part1 Flashcards
- Your EC2 instance is running out of disk space, what actions will you take to mitigate this issue ?
First I’ll check the EBS volumes attached is it root, application ? are there pending jobs or logs ? then I’ll take a snapshot then increase the volume capacity.
- Explain the different ways in which Prometheus can get metrics.
it can use a pull base approach where it can use a node exporter to scrape the metrics or pull base, it can also collect metrics through custom libraries where metrics will be sent through a gateway and that will be a push base approach.
what is Kubernetes kOps ? how is it different from kubeadm ?
KOps is an automation tool used to set up a Kubernetes cluster instead of using kubeadm. It is not a managed cluster like EKS. kubeadm does not integrate EBS EC2 ASG route53.
what is instance fleet in AWS?
it is a combination of different kinds of EC2 instances with respect to capacity and pricing models.
Are you supporting any application? if yes what is the application server?
i have supported a team that is developing an application in java and they are using a tomcat service to deploy the application. As a DevOps engineer i am aware that it could range from java to python using django or nodejs and so on and i will get the knowledge if required.
we have a stage in Jenkins which requires to use an agent with at least 500MB of RAM. How do you advise to do this in Jenkins ?
Assuming this Jenkins setup is all EC2 instances, first i will label the agents to correspond with the labels. Then say a specific stage in the Jenkins job requires more memory then i might restrict the stage to use a specific label. If this is in k8s, i will place a request for a pod that has 500MB and once it gets assigned it will act as an gent for my job.
what is vertical scaling and horizontal scaling ? explain with use case where you have seen this.
Vertical scaling refers to adding more power to an existing server or machine. This means increasing its CPU, RAM, storage, or other hardware resources to handle more load.
Horizontal scaling involves adding more machines or servers to your system. Instead of making one machine more powerful, you add more machines to distribute the workload.
During a recent project, I worked with a DevOps pipeline for a high-traffic application. Initially, we started with vertical scaling to quickly meet performance requirements. However, as traffic grew, we shifted to horizontal scaling by deploying multiple containers of the application using Kubernetes and a load balancer (e.g., NGINX). This allowed us to handle the growing user base without major application downtime.
explain replication controller in k8s
it is responsible for making sure x number of replicas are available and running in the cluster at all times
what is north south and east west traffic in k8s ?
north south refers to incoming traffic from clients to the cluster. east west is traffic between k8s pods in one namespace or different namespace.
why use helm in k8s ?
Helm allows you to package Kubernetes manifests (YAML files) into reusable Helm charts. This makes deploying complex applications with multiple resources (e.g., deployments, services, config maps, ingress) much easier.
Instead of manually managing multiple YAML files, you deploy the entire application with a single helm install command
