Automation is the heart of DevOps. And one of the most powerful automation tools is Shell Scripting. In this guide, you’ll learn real DevOps-focused shell scripts, each explained line-by-line, tool-by-tool:

• AWS (S3, EC2)

• Jenkins (Trigger jobs, get status)

• Docker (backup, cleanup, listing containers)

• Kubernetes (pods, nodes, health checks)

Table of Contents

1. Why DevOps Engineers Need Shell Scripting

2. AWS Automation Scripts

   • S3 Sync Backup

   • List EC2 Instances

3. Linux Service Monitoring Script

4. Jenkins Automation

   • Trigger Jenkins Job

   • Check Jenkins Job Status

5. Docker Automation Scripts

   • List Running Containers

   • Backup Containers & Images

   • Delete Unused Images

6. Kubernetes Automation Scripts

   • Watch Live Pods

   • Pod Health Check

   • Restart All Pods

   • Node Status Check

7. Final Thoughts

Why Shell Scripting Matters in DevOps

DevOps automation = repeatable, predictable, error-free processes.

Shell scripting helps you automate:

• Server provisioning

• AWS CLI tasks

• Monitoring

• CI/CD workflows

• Docker operations

• Kubernetes management

Every DevOps engineer must know how to build scripts that run cloud commands automatically.

1. AWS Automation Scripts

1.1 Sync Local Directory to AWS S3 (Automated Backup)

This script uploads a local directory (like logs or DB dumps) to an S3 bucket.

📌 Script

#!/bin/bash                                 # Use Bash shell

BUCKET_NAME="my-backup-bucket"               # S3 bucket name
SOURCE_DIR="/backup"                         # Local directory to sync

echo "Syncing $SOURCE_DIR to S3 bucket $BUCKET_NAME..."   # Show progress

aws s3 sync $SOURCE_DIR s3://$BUCKET_NAME --delete        
# aws s3 sync = Copies files
# $SOURCE_DIR = Local folder
# s3://$BUCKET_NAME = Destination bucket
# --delete = Removes files from bucket if deleted locally

echo "Sync Completed!"                       # Print completion message

1.2 List AWS EC2 Instances with Public IPs

Useful when managing multiple servers.

📌 Script

#!/bin/bash

echo "Fetching EC2 Instances..."             # Progress message

aws ec2 describe-instances \
--query "Reservations[*].Instances[*].[InstanceId, PublicIpAddress]" \
--output table
# --query = Filters output using JMESPath
# Shows Instance ID + Public IP only
# --output table = Makes it readable

2. Check if a Linux Service is Running

Example: restart Nginx/MySQL automatically if it stops.

📌 Script

#!/bin/bash

SERVICE="nginx"                                # Service name to monitor

if systemctl is-active --quiet $SERVICE; then  # Check if running
    echo "$SERVICE is running"                 # OK
else
    echo "$SERVICE is NOT running, restarting..."  # Error detected
    systemctl.restart $SERVICE                 # Restart service
fi

3. Jenkins Automation Scripts

3.1 Trigger Jenkins Job Without UI

Perfect for CI/CD triggers, Git hooks, webhooks.

📌 Script

#!/bin/bash

# ================================
# Jenkins Connection Information
# ================================

JENKINS_URL="http://localhost:8080"      # URL of your Jenkins server
JOB_NAME="MyJob"                         # Name of the Jenkins job you want to trigger
USER="admin"                             # Jenkins username
API_TOKEN="your-api-token"               # API token for authentication (NOT your password)

# ================================
# Display a friendly message
# ================================

echo "Triggering Jenkins job: $JOB_NAME..."   # Inform the user what's happening

# ================================
# Trigger Jenkins Job using curl
# ================================

curl -X POST "$JENKINS_URL/job/$JOB_NAME/build" \   # -X POST means send a POST request to trigger the job
  --user "$USER:$API_TOKEN"                         # Send username + API token for authentication
                                                     # Without this, Jenkins will reject the request

# Important Notes:
# /job/<JOB_NAME>/build → this Jenkins API endpoint starts a new build
# curl sends the request directly to Jenkins API
# No need to open Jenkins UI manually

# ================================
# Confirmation message
# ================================

echo "Job triggered successfully!"   # Final message after triggering the job

What This Script Does

• Triggers a Jenkins job from the terminal

• No need to open the Jenkins web UI

• Useful in automation (Git hooks, cron jobs, CI/CD pipelines)

• Uses API token for secure authentication

• Works for freestyle, pipeline, and multibranch jobs

3.2 Check Jenkins Job Status

Shows whether last build = SUCCESS / FAILURE.

📌 Script

#!/bin/bash

# ================================
# Jenkins Login Details
# ================================

JENKINS_URL="http://localhost:8080"    # Jenkins server URL
JOB_NAME="MyJob"                       # Job you want to check
USER="admin"                           # Jenkins username
API_TOKEN="your-api-token"             # Jenkins API token (NOT password)


# ================================
# Tell the user what is happening
# ================================

echo "Checking last build status for: $JOB_NAME..."


# ================================
# Call Jenkins REST API
# ================================

curl -s "$JENKINS_URL/job/$JOB_NAME/lastBuild/api/json" \   # -s = silent output (no progress bar)
  --user "$USER:$API_TOKEN" \                                # Login using username + API token
  | jq -r '.result'                                          # Extract only the build result (SUCCESS, FAILURE, etc.)
# jq -r → raw output without quotes
# .result → field in JSON that stores the build status

# ================================
# Possible Results Explained
# ================================
# SUCCESS   → Build completed successfully
# FAILURE   → Build failed
# ABORTED   → Someone cancelled the build
# UNSTABLE  → Tests failed or warnings found
# null      → Build is still running (no result yet)

4. Docker Automation Scripts

4.1 List All Running Containers

📌 Script

#!/bin/bash

echo "Running Docker containers (ID + Name):"    # Header

docker ps --format "{{.ID}}  {{.Names}}"         # Show only ID and container name
# {{.ID}}    → container ID
# {{.Names}} → container name
# Very simple and clean output

4.2 Backup Docker Containers & Images

Used during migrations, upgrades, or disaster recovery.

📌 Script

#!/bin/bash

# =============================
# Create a backup folder
# =============================

BACKUP_DIR="/backup/docker"      # Location to store backups
mkdir -p "$BACKUP_DIR"           # Create folder if it does not exist

echo "Backing up running containers..."

# =============================
# Backup all RUNNING containers
# =============================

for container in $(docker ps -q); do          # Loop through container IDs

    docker commit "$container" "${container}-backup"
    # docker commit → takes a snapshot of the running container

    docker save -o "$BACKUP_DIR/${container}.tar" "${container}-backup"
    # docker save → save that snapshot as a .tar file
done

echo "Backing up all Docker images..."

# =============================
# Backup ALL Docker images
# =============================

docker images -q | xargs -I {} docker save -o "$BACKUP_DIR/{}.tar" {}
# docker images -q → list only image IDs
# xargs → passes each ID into docker save
# docker save → exports image as a tar file

echo "Backup completed successfully!"
echo "Files saved in: $BACKUP_DIR"

4.3 Delete Old Unused Docker Images

Very useful for freeing disk space.

📌 Script

#!/bin/bash

echo "Deleting unused Docker images..."    # Info
docker image prune -a -f                  # Remove unused images
echo "Cleanup done!"                      # Completed

5. Kubernetes Automation Scripts

5.1 Monitor Pod Status (Live Watch)

This is used during deployments.

📌 Script

#!/bin/bash

# =============================
# Select Kubernetes Namespace
# =============================

NAMESPACE="default"      # Namespace you want to monitor (e.g., default, dev, prod)

echo "Monitoring pods in namespace: $NAMESPACE..."   # Display message

# =============================
# Live Pod Watch Command
# =============================

kubectl get pods -n $NAMESPACE --watch
# kubectl get pods → list all pods
# -n $NAMESPACE   → check pods inside the chosen namespace
# --watch         → keeps the command running & updates live
# Useful during deployments or troubleshooting

5.2 Kubernetes Pod Health Check

Alerts if any pod is not in Running state.

📌 Script

#!/bin/bash

# =============================
# Select Namespace to Check
# =============================

NAMESPACE="default"     # Namespace you want to check (example: default, dev, prod)

echo "Checking Kubernetes Pod Health..."   # Message before running check

# =============================
# Find Pods That Are NOT Running
# =============================

kubectl get pods -n $NAMESPACE --no-headers | \        # List all pods without header row
awk '$3 != "Running" {                                  # If column 3 (STATUS) is NOT "Running"
    print "⚠️ Pod "$1" is in state "$3                  # Print alert with pod name + its state
}'
# Column meanings from 'kubectl get pods':
# $1 = Pod Name
# $2 = Ready containers
# $3 = STATUS (Running, Pending, CrashLoopBackOff, Error)

echo "Health Check Completed."   # Final message

5.3 Restart All Pods in a Namespace

Used after config updates, DNS issues, or deployments.

📌 Script

#!/bin/bash

# =============================
# Choose Namespace
# =============================

NAMESPACE="default"     # Namespace where pods need to be restarted

echo "Restarting all pods in $NAMESPACE..."     # Display message before restart

# =============================
# Delete All Pods → K8s Recreates Them Automatically
# =============================

kubectl delete pods --all -n $NAMESPACE --grace-period=0 --force
# kubectl delete pods --all → deletes every pod in the namespace
# -n $NAMESPACE            → applies to the selected namespace
# --grace-period=0         → immediate kill, no waiting time
# --force                  → force delete even if pods are stuck
# Kubernetes automatically re-creates pods through Deployments/ReplicaSets

# =============================
# Completion Message
# =============================

echo "All pods restarted!"        # Inform user that process is done

- Deletes all pods in the namespace
- Kubernetes automatically recreates them (because Deployments/ReplicaSets manage them)
- Useful when:

• Pods are stuck

• Pods are in CrashLoopBackOff

• ConfigMaps/Secrets updated & need restart

• Deployment not refreshing

5.4 Check Kubernetes Node Status

Useful for cluster health.

📌 Script

#!/bin/bash

echo "Checking Kubernetes Nodes..."

kubectl get nodes | grep -v "Ready"
# Show nodes NOT in Ready state

Final Thoughts

Shell scripting is one of the most powerful weapons in a DevOps engineer’s toolbox.
With AWS CLI, Docker CLI, Kubernetes CLI, and Jenkins CLI — automation becomes easy.