Devops and aws interview preparation Hyderabad
前往频道在 Telegram
Who interested to learn Linux AWS and devops I will explain and I will give my support until you got job and no need to go for proxy after training. For traning call me 9154078579 For serious learners I will share my entire knowledge
显示更多5 042
订阅者
-124 小时
+67 天
+3930 天
数据加载中...
相似频道
标签云
进出提及
---
---
---
---
---
---
吸引订阅者
九月 '26
九月 '26
+54
在0个频道中
八月 '26
+127
在0个频道中
Get PRO
七月 '26
+169
在0个频道中
Get PRO
六月 '26
+140
在0个频道中
Get PRO
五月 '26
+145
在0个频道中
Get PRO
四月 '26
+163
在0个频道中
Get PRO
三月 '26
+133
在0个频道中
Get PRO
二月 '26
+106
在0个频道中
Get PRO
一月 '26
+111
在0个频道中
Get PRO
十二月 '25
+108
在0个频道中
Get PRO
十一月 '25
+112
在0个频道中
Get PRO
十月 '25
+106
在0个频道中
Get PRO
九月 '25
+5
在0个频道中
Get PRO
八月 '25
+1
在0个频道中
Get PRO
七月 '25
+3
在0个频道中
Get PRO
六月 '25
+7
在0个频道中
Get PRO
五月 '25
+5
在0个频道中
Get PRO
四月 '25
+5
在0个频道中
Get PRO
三月 '25
+4
在0个频道中
Get PRO
二月 '250
在0个频道中
Get PRO
一月 '25
+8
在0个频道中
Get PRO
十二月 '24
+89
在0个频道中
Get PRO
十一月 '24
+305
在0个频道中
Get PRO
十月 '24
+356
在0个频道中
Get PRO
九月 '24
+265
在0个频道中
Get PRO
八月 '24
+227
在0个频道中
Get PRO
七月 '24
+155
在0个频道中
Get PRO
六月 '24
+131
在0个频道中
Get PRO
五月 '24
+184
在0个频道中
Get PRO
四月 '24
+141
在0个频道中
Get PRO
三月 '24
+178
在0个频道中
Get PRO
二月 '24
+187
在0个频道中
Get PRO
一月 '24
+317
在0个频道中
Get PRO
十二月 '23
+113
在0个频道中
Get PRO
十一月 '23
+23
在0个频道中
Get PRO
十月 '23
+48
在0个频道中
Get PRO
九月 '23
+54
在0个频道中
Get PRO
八月 '23
+104
在0个频道中
Get PRO
七月 '23
+75
在0个频道中
Get PRO
六月 '23
+72
在0个频道中
Get PRO
五月 '23
+49
在0个频道中
Get PRO
四月 '23
+94
在0个频道中
Get PRO
三月 '23
+71
在0个频道中
Get PRO
二月 '23
+80
在0个频道中
Get PRO
一月 '23
+180
在0个频道中
Get PRO
十二月 '22
+104
在0个频道中
Get PRO
十一月 '22
+116
在0个频道中
Get PRO
十月 '22
+44
在0个频道中
Get PRO
九月 '22
+58
在0个频道中
Get PRO
八月 '22
+52
在0个频道中
Get PRO
七月 '22
+53
在0个频道中
Get PRO
六月 '22
+252
在0个频道中
Get PRO
五月 '22
+55
在0个频道中
Get PRO
四月 '22
+52
在0个频道中
Get PRO
三月 '22
+80
在0个频道中
Get PRO
二月 '22
+24
在0个频道中
Get PRO
一月 '22
+34
在0个频道中
Get PRO
十二月 '21
+45
在0个频道中
Get PRO
十一月 '21
+1 040
在0个频道中
| 日期 | 订阅者增长 | 提及 | 频道 | |
| 17 九月 | 0 | |||
| 16 九月 | +1 | |||
| 15 九月 | +2 | |||
| 14 九月 | +4 | |||
| 13 九月 | +4 | |||
| 12 九月 | +3 | |||
| 11 九月 | +1 | |||
| 10 九月 | +4 | |||
| 09 九月 | +3 | |||
| 08 九月 | +4 | |||
| 07 九月 | +4 | |||
| 06 九月 | +7 | |||
| 05 九月 | +2 | |||
| 04 九月 | +5 | |||
| 03 九月 | +1 | |||
| 02 九月 | +3 | |||
| 01 九月 | +6 |
频道帖子
| 2 |  没有文字... | 444 |
| 3 | Question:-106 Suppose your EKS cluster is experiencing pod CrashLoopBackOff issues. Walk me through your debugging process.
When an Amazon EKS cluster experiences a CrashLoopBackOff issue, it means the container image was pulled successfully, but the application inside the container is starting, crashing, and restarting repeatedly
1. we can describe pod and check initially
kubectl describe pod <pod-name> -n <namespace>
we can check events section and the container status
A. If a liveness probe is misconfigured or too aggressive, it will kill a healthy application before it finishes booting up.
B. Issues mounting AWS EBS volumes, EFS, or missing ConfigMaps/Secrets keys.
2.If the platform infrastructure looks healthy, the bug is almost certainly inside the application
kubectl logs <pod-name> -n <namespace> --previous
Missing database connection strings, unhandled exceptions, database migration failures, or filesystem permission errors.
3.If the app crashes because it cannot connect to an Amazon RDS database, I verify that the EKS cluster security group is allowed on the RDS instance security group
To learn aws&Devops from scratch ping me 9154078579 | 462 |
| 4 | Hi I am going to start new aws&Devops session on 16-09-2026 Wednesday @7 Am IST The training main intention is to clear the interview on your own
Ping me who are interested to Jion the session 9154078579 | 673 |
| 5 | How To Get the Pod IP in Kubernetes ?
1.Each Pod is assigned a unique IP address upon initialization, which facilitates communication.
2.containers in a Pod share the same network namespace, IP address, and resources and can all reach each other's ports on localhost.
3.The Container Network Interface (CNI) plugin assigns IP addresses to pods by communicating with the container runtime when a pod is created and releasing them when the pod is deleted.
4.The CNI plugin is configured with a large CIDR range (e.g., 10.10.0.0/16)
5.When a new pod is created, Kubernetes asks the CNI plugin for an IP
6.The CNI plugin has an IPAM (IP Address Management) component that tracks all IPs
7.The CNI configures the network interfaces and routing for the pod
To learn aws&Devopd from scratch ping me 9154078579 | 801 |
| 6 | what is dns resolution in pod?
1. It is the process where an application inside a pod translates domain or service names into IP addresses using the cluster's internal DNS server
2. Kubernetes automatically injects an /etc/resolv.conf file into every container inside a pod.
3. Pod IPs in Kubernetes are ephemeral and change constantly when pods crash or scale, applications rely on stable Service names to talk to one another.
To solve pod DNS resolution issues, we must check CoreDNS status
example : Need to check if CoreDNS pods are running using kubectl get pods -n kube-system -l k8s-app=kube-dns
Inspect CoreDNS logs for errors using kubectl logs -n kube-system -l k8s-app=kube-dns
To learn aws&Devops from scratch ping me 9154078579 | 839 |
| 7 | Question:-104 How are you handling security risks under release pressure in aws and devops ?
1. Balancing rapid release pressure with cloud security requires transitioning from a manual "gatekeeper" model to an automated DevSecops
2.we can use tools like Checkov, TFLint, or KICS scan Terraform, AWS CloudFormation, or AWS CDK templates for misconfigurations (e.g., open S3 buckets) before deployment.
3.AWS Identity and Access Management (IAM) Access Analyzer automatically validates policies against best practices during the deployment pipeline.
4.AWS Config Continuously monitors and records AWS resource configurations, automatically flagging or reverting changes that violate compliance rules.
5.Amazon GuardDuty Uses machine learning to detect threats, unauthorized access, and malicious activity in near-real-time across AWS accounts.
6. Amazon Inspector Automatically and continuously scans EC2 instances, container images in Amazon ECR, and AWS Lambda functions for software vulnerabilities.
To learn aws&Devops from scratch ping me 9154078579 | 774 |
| 8 | Question:-103 Do you create pipelines or mainly support deployments in devops ?
Yes both pipeline creation and deployment support are fully handled.
I can do both. I can write the actual configuration files to build our CI/CD pipelines from scratch, and I can troubleshoot or support existing deployments.
1. I can generate complete, production-ready configuration files for platforms like GitHub Actions and Jenkins
2.Continuous Integration (CI): I can write steps to automate linting, static code analysis (like SonarQube), running unit tests, and building Docker images.
3.Continuous Delivery/Deployment (CD): I can configure multi-stage environments (Dev, Staging, Prod) with manual approval gates, rolling updates, or blue-green strategies.
4.Infrastructure as Code (IaC): I can provision the infrastructure your pipelines deploy to by writing Terraform files
5.Kubernetes crash loops (CrashLoopBackOff), or Docker build errors, and I will isolate the root cause and provide the fix.
6.I can write Kubernetes manifests, Helm charts, or ArgoCD/Flux configurations to manage stateful deployments.
7. I can help you optimize slow build times, fix caching strategies, or secure our pipeline by shifting to secret management tools like HashiCorp Vault.
To learn aws&Devops from scratch ping me 9154078579 | 705 |
| 9 | Question:-102 How many Internet Gateway are there in VPC's ? Can you create more ?
1. A VPC can also have zero Internet Gateways if it is designed to be completely private
2. It is a strict 1:1 relationship an Internet Gateway cannot be attached to multiple VPCs simultaneously.
3. we cannot attach more than one IGW to a single VPC.
4. A single Internet Gateway is already highly available, redundant, and scales horizontally automatically.
5. Exactly one Internet Gateway connects to a single VPC. An Internet Gateway also connects to only one VPC at a time.
6. The total number of Internet Gateways per region matches your VPC limit (default is 5 per region, but we can request an increase). we cannot attach two of them to the same VPC.
For practical aws&Devops learning ping me 9154078579 | 710 |
| 10 |  Call with an 8+ years experienced aws&Devops expert | 954 |
| 11 |  We are making a big network team by connecting with real time employees to unemployed &job searching people on Aws and devops
Contact us : 9154078579 | 1 052 |
| 12 | Question:-101 Explain depends_on vs implicit dependency — when does Terraform get it wrong?
1. Implicit Dependency (Automatic): Terraform parses our configuration, detects references (e.g., vpc_id = aws_vpc.main.id), builds a dependency graph, and creates the VPC before the resource that needs it.
A. Automatically detected when one resource references an attribute of another.
B. No extra configuration required.
C. Reduces code complexity and keeps configurations clean.
Implicit Dependency Example
resource "aws_security_group" "web_sg" {
name = "web-sg"
}
resource "aws_instance" "web" {
ami = "ami-123456"
instance_type = "t2.micro"
vpc_security_group_ids = [https://lnkd.in/d9wa7Mpj] # Implicit
dependency
}
Here, the EC2 instance depends on the security group automatically because it references its ID. No depends_on needed.
2.Explicit Dependency (depends_on): explicitly tell Terraform to wait. For example, depends_on = [aws_iam_role_policy_attachment.example] ensures a permission is applied before an application attempts to use it, even though the application resource doesn't reference that policy attachment directly.
A. Defined using the depends_on meta-argument.
B.Forces Terraform to create/destroy resources in a specific order, even if no direct attribute references exist.
Explicit Dependency Example
resource "aws_s3_bucket" "logs" {
bucket = "my-logs-bucket" }
To learn aws&Devops from scratch ping me 9154078579 | 1 393 |
| 13 | Question:-100 A developer accidentally commits AWS credentials to Git. What is your complete incident response process?
1.we do not touch the Git repository first priority is disabling the credential at the provider level to stop active exploitation.
2.Log into the AWS IAM Console or use the AWS CLI to change the key status to Inactive. Do not delete it yet; keeping the key identifier intact helps when parsing logs.
aws iam update-access-key --access-key-id <LEAKED_KEY_ID> --status Inactive --user-name <USER_NAME>
3.If the key belonged to an IAM user, apply an explicit inline deny policy to that user to instantly terminate any ongoing API sessions initiated by attackers.
4.We Look closely for malicious resource creation, such as newly launched high-end EC2 instances (typically for crypto-mining), modified S3 bucket policies, or newly created IAM users/backdoor access keys.
5.Scan all active AWS regions. Attackers frequently spin up resources in secondary or unused regions to bypass localized monitoring dashboards.
6.Simply deleting the file and making a new commit leaves the secret visible in our repository's historical timeline. we use tools like git filter-repo or BFG Repo-Cleaner to strip the credentials from all historical commits, branches, and tags.
To learn aws&Devops from scratch ping me 9154078579 | 1 397 |
| 14 | Hi I am going to start new AWS&Devops session @9:30 Am IST on 10-08-2026 the training main intension is to clear the interview by ourself ping me who are interested to join session 9154078579 | 1 269 |
| 15 | Question:-98 What is Exit Status 143 in Kubernetes?
1. The number is derived from the standard Linux formula for signal-based exits: 128 + Signal Number (128 + 15 = 143)
A container was successfully terminated by a SIGTERM (Signal 15) request.
2.Exit Status 143 is often part of normal cluster maintenance and operations, rather than an application crash
3.Kubernetes scales down a deployment, removing excess replicas.
4.A new version of your application is deployed, requiring old pods to shut down.
5.A cluster node is being prepared for maintenance, causing Kubernetes to evict and reschedule pods it's Node draining
6.The kubelet detects the application is unhealthy and initiates a restart.
To learn aws and devops from scratch ping me 9154078579 | 1 291 |
| 16 | Question:-97 Suppose AWS Config automatically deletes any IAM User that gets created. How would you handle a business exception where a specific IAM User is required?
1.The actual deletion is handled by an AWS Systems Manager (SSM) Automation document or an AWS Lambda function.
2.Modify the code or automation document that executes the deletion to check for this tag before terminating the user.
Tag Key: ConfigExemption
Tag Value: AllowIAMUser
import boto3
iam = boto3.client('iam')
def lambda_handler(event, context):
# Extract username from the AWS Config evaluation event
username = event['detail']['requestParameters']['userName']
# Fetch user tags
tags = iam.list_user_tags(UserName=username)['Tags']
# Check for exemption
for tag in tags:
if tag['Key'] == 'ConfigExemption' and tag['Value'] == 'AllowIAMUser':
print(f"User {username} is exempt from auto-deletion.")
return
# Delete user if not exempt
iam.delete_user(UserName=username)
3. we need to configure Amazon EventBridge to detect when a user is created with the exemption tag.
4.Move the workload requiring the IAM User to a dedicated AWS account or Organizational Unit (OU) where this specific AWS Config rule is not deployed.
To learn aws&Devops from scratch ping me 9154078579 | 1 062 |
| 17 | Question:-96 How do you monitor AWS applications? Which tools, metrics, and alerts do you consider essential?
we can establish a reliable, automated monitoring environment using native AWS tools alongside critical metrics and alerts.
1.Amazon CloudWatch: Acts as our core observability hub. It collects system performance data, unifies application logs, and visualizes system trends on central dashboards.
2.AWS X-Ray: Operates as our primary troubleshooting engine. It traces requests as they travel across complex microservice architectures to isolate performance bottlenecks
3.AWS CloudTrail: Serves as our governance and security auditor. It logs all API calls and user changes made within our infrastructure
4.Third-Party Platforms: Tools like Datadog, New Relic, or Dynatrace provide external alternatives. They aggregate multi-cloud metrics and deliver unified distributed tracing
Below are essential Metrics to Monitor
1. Request Count / Throughput: Total volume of traffic hitting the application. 2.4xx and 5xx Error Rates: Indicates client-side request issues or server-side application faults.
3.Integration Timeouts: Delays connecting to databases (e.g., RDS) or external APIs.
4.CPU & Memory Utilization: Tracked closely via Amazon CloudWatch to ensure workloads are not starved or over-provisioned.
To learn aws&Devops from scratch ping me 9154078579 | 756 |
| 18 | Question:-95 Explain one challenging production incident and how you resolved it in jenkins
1. A challenging production incident in Jenkins is unexpectedly dropping offline mid-build or random OutOfMemory (OOM) errors .
2.During high-load periods or long-running jobs (like Docker image builds or test suites), Jenkins agents would randomly disconnect with a java.io.IOException: Unexpected termination of channel error.
3.Java Remoting ping timeouts, or the Linux Out of Memory (OOM) killer terminating the gent process due to high resource usage.
4.For cloud-based or SSH agents, the Jenkins remoting ping thread occasionally drops the connection if a network packet is delayed. This was resolved by configuring the SSH client configuration on the agent nodes to keep connections alive automatically .
Host *
ServerAliveInterval 300
ServerAliveCountMax 30
5. We Need to configure pipelines to use the Jenkins Workspace Cleanup Plugin to periodically delete old workspaces and free up local disk space.
To learn aws&Devops from scratch ping me 9154078579 | 621 |
| 19 | Question:-94 Why is the Cluster Autoscaler not scaling up even though pods are in the Pending state?
1. If a pod has strict scheduling requirements (e.g., nodeSelector, nodeAffinity, or tolerations) that do not match any available or deployable node in our cloud environment, the CA will refuse to scale up
2.If the CA requests a new node but our cloud provider limits are maxed out (e.g., AWS EC2 vCPU limit), the scale-up will fail.
3. The CA scales based purely on the resources.requests defined in the pod manifest, not actual hardware usage
4.The pod's CPU or memory requests are larger than the total allocatable capacity of the largest single node instance type available in our node groups. A pod cannot be split across multiple nodes.
To learn aws&Devops from scratch ping me 9154078579 | 685 |
| 20 | Question:-94 Why is the Cluster Autoscaler not scaling up even though pods are in the Pending state?
1. If a pod has strict scheduling requirements (e.g., nodeSelector, nodeAffinity, or tolerations) that do not match any available or deployable node in our cloud environment, the CA will refuse to scale up
2.If the CA requests a new node but our cloud provider limits are maxed out (e.g., AWS EC2 vCPU limit), the scale-up will fail.
3. The CA scales based purely on the resources.requests defined in the pod manifest, not actual hardware usage
4.The pod's CPU or memory requests are larger than the total allocatable capacity of the largest single node instance type available in our node groups. A pod cannot be split across multiple nodes.
To learn aws&Devops from scratch ping me 9154078579 | 733 |
