RealCoding Tech Blog
🇰🇷 한국어 * 🇯🇵 日本語 *

Building Remote Desktop with AWS EC2 - Reality and Future of Cloud Work Environment

Kevin Park Development, Project 8분 읽기
aws ec2 remote-desktop cloud windows cost-analysis automation tutorial

Building Remote Desktop with AWS EC2 - Reality and Future of Cloud Work Environment

🎯 Summary

AWS EC2 Remote Desktop Quick Setup Guide:

# 1. Create Windows Server instance
aws ec2 run-instances \
  --image-id ami-0d8f6eb4f641ef691 \
  --instance-type t3a.xlarge \
  --key-name your-key-pair \
  --security-group-ids sg-xxxxxxxx

# 2. Configure RDP connection
# Open port 3389 (in security group)
# Set administrator password

# 3. Connect via Remote Desktop
mstsc /v:YOUR_EC2_PUBLIC_IP

Key Cost Comparison (Monthly):

  • Regular Office PC: ~$100 (2-year depreciation)
  • AWS EC2 (t3a.xlarge): ~$50 (Mon-Fri 8 hours usage)
  • Savings: $50/month, $600/year

Essential Considerations:

  • Network latency (Seoul region recommended for Korea)
  • Data transfer costs
  • Instance management automation required

📚 Detailed Explanation

Background and Necessity

The era of remote desktop has already become reality. Especially after COVID-19, as remote work became mainstream, interest in cloud-based work environments has surged dramatically. Remote desktop using AWS EC2 provides the following innovative work environment.

Future Work Scenario:

  1. Commute with a lightweight tablet
  2. Connect to docking station for large monitor usage
  3. Auto-boot cloud desktop
  4. Start work with yesterday’s work environment preserved intact

AWS EC2 Remote Desktop Configuration Method

1. Instance Selection and Creation

Recommended Instance Types:

# General office use (document work, web browsing)
Instance Type: t3a.medium (2 vCPU, 4GB RAM)
Monthly Cost: ~$25

# Developer use (IDE, compilation, testing)
Instance Type: t3a.xlarge (4 vCPU, 16GB RAM)
Monthly Cost: ~$50

# Designer/Creator use (Adobe CC, video editing)
Instance Type: c5.2xlarge (8 vCPU, 16GB RAM)
Monthly Cost: ~$120

Creation Process:

# Create instance via AWS CLI
aws ec2 run-instances \
  --image-id ami-0d8f6eb4f641ef691 \
  --instance-type t3a.xlarge \
  --key-name my-key-pair \
  --security-group-ids sg-xxxxxxxxx \
  --subnet-id subnet-xxxxxxxxx \
  --tag-specifications 'ResourceType=instance,Tags=[{Key=Name,Value=MyDesktop}]'

2. Security Group Configuration

# Open RDP port (3389)
aws ec2 authorize-security-group-ingress \
  --group-id sg-xxxxxxxxx \
  --protocol tcp \
  --port 3389 \
  --source-group your-office-ip/32

# Open HTTPS port (for web-based remote access)
aws ec2 authorize-security-group-ingress \
  --group-id sg-xxxxxxxxx \
  --protocol tcp \
  --port 443 \
  --source-group 0.0.0.0/0

3. Windows Configuration and Optimization

Automated setup via PowerShell:

# Enable RDP
Set-ItemProperty -Path 'HKLM:\System\CurrentControlSet\Control\Terminal Server' -name "fDenyTSConnections" -Value 0

# Add firewall rules
Enable-NetFirewallRule -DisplayGroup "Remote Desktop"

# Set timezone
Set-TimeZone -Id "Korea Standard Time"

# Install Korean language pack
Install-Language Ko-KR

Detailed Cost Analysis

Regular Office PC vs AWS EC2 Comparison

Regular Office PC (2-year basis):

PC purchase: $2,000
Windows license: $400
Electricity (monthly $15 × 24 months): $360
Maintenance (annual $100 × 2 years): $200
─────────────────────────────
Total cost: $2,960
Monthly average: $123

AWS EC2 (Mon-Fri 8 hours, 3-year commitment):

t3a.xlarge Savings Plan: $45.33/month
EBS Storage 50GB: $5/month
Data Transfer (monthly 100GB): $9/month
─────────────────────────────
Total cost: $59.33/month

Real Project Implementation Case

Small-Medium Enterprise (30 employees) Implementation Results:

Existing PC purchase budget: $60,000
AWS cloud desktop annual cost: $24,000
1-year savings: $36,000
ROI: 150%

Real Implementation Considerations

1. Network Latency Optimization

# Optimal region for Korean users
Region: ap-northeast-2 (Seoul)
Availability Zone: ap-northeast-2a

# Dedicated line connection (enterprise)
Utilize AWS Direct Connect
Achievable latency: Under 5ms

2. Automation Script Implementation

Automatic instance start/stop:

import boto3
import schedule
import time

def start_instance():
    ec2 = boto3.client('ec2', region_name='ap-northeast-2')
    ec2.start_instances(InstanceIds=['i-1234567890abcdef0'])
    print("Instance started")

def stop_instance():
    ec2 = boto3.client('ec2', region_name='ap-northeast-2')
    ec2.stop_instances(InstanceIds=['i-1234567890abcdef0'])
    print("Instance stopped")

# Start at 8 AM, stop at 6 PM on weekdays
schedule.every().monday.at("08:00").do(start_instance)
schedule.every().monday.at("18:00").do(stop_instance)
# ... configure remaining days

3. Backup and Snapshot Management

# Daily automatic backup setup
aws dlm create-lifecycle-policy \
  --description "Daily Desktop Backup" \
  --state ENABLED \
  --execution-role-arn arn:aws:iam::123456789012:role/AWSDataLifecycleManagerDefaultRole \
  --policy-details '{
    "PolicyType": "EBS_SNAPSHOT_MANAGEMENT",
    "ResourceTypes": ["INSTANCE"],
    "TargetTags": [{"Key": "Environment", "Value": "Desktop"}],
    "Schedules": [{
      "Name": "DailyBackup",
      "CreateRule": {"Interval": 24, "IntervalUnit": "HOURS", "Times": ["03:00"]},
      "RetainRule": {"Count": 7}
    }]
  }'

Security Optimization Strategies

1. Multi-Factor Authentication (MFA)

# MFA setup in Windows
Install-Module -Name MSAL.PS
$ClientId = "your-azure-app-id"
$TenantId = "your-tenant-id"
$Token = Get-MsalToken -ClientId $ClientId -TenantId $TenantId

2. VPN Connection Configuration

# AWS Client VPN setup
aws ec2 create-client-vpn-endpoint \
  --description "Desktop VPN" \
  --server-certificate-arn arn:aws:acm:region:account:certificate/certificate-id \
  --client-cidr-block 10.0.0.0/16 \
  --authentication-options Type=certificate-authentication,MutualAuthentication={ClientRootCertificateChainArn=arn:aws:acm:region:account:certificate/client-certificate-id}

Performance Optimization Tips

1. GPU Acceleration (for graphics work)

# Utilize G4 instances (NVIDIA T4 GPU)
Instance Type: g4dn.xlarge
Use cases: 3D modeling, video editing, AI/ML tasks
Monthly cost: ~$200

2. Storage Optimization

# Use high-performance SSD
VolumeType: gp3
IOPS: 3000 (default)
Throughput: 125 MB/s (default)

# For large data volumes
VolumeType: st1 (Throughput Optimized HDD)
Cost: 50% savings compared to gp3

Real-world Use Cases

Case 1: Startup Development Team (5 members)

Configuration:

  • 4 Developers: t3a.xlarge × 4
  • 1 Designer: g4dn.xlarge × 1
  • Monthly total cost: ~$520
  • Annual savings of $12,000 compared to existing PC purchases

Implementation Benefits:

  • 100% development environment standardization
  • 70% reduction in new team member onboarding time
  • Automated data backup/recovery

Case 2: Design Agency (15 members)

Configuration:

  • Designers: g4dn.2xlarge × 10
  • Planners: t3a.medium × 5
  • Monthly total cost: ~$1,800

Special Features:

  • Central Adobe Creative Cloud license management
  • Project-specific work environment snapshots
  • Temporary access accounts for client reviews

Implementation Precautions

1. Data Sovereignty and Compliance

# Domestic data storage requirements
Region: ap-northeast-2 (Seoul) mandatory
Compliance: K-ISMS, Personal Information Protection Act compliance
Encryption: EBS volume encryption mandatory

2. Internet Connection Dependency

Contingency Plans:

  • Multi-ISP connections
  • 4G/5G backup connections
  • Local cache system for offline work capability

3. License Management

# Windows Server license
Recommend using License Included AMI
Consider BYOL (Bring Your Own License) for cost savings

# Office 365
Recommend Microsoft 365 Business Premium
$22 per user per month

1. AWS WorkSpaces vs Direct Setup

AWS WorkSpaces Advantages:

  • Minimal management burden
  • Starting from $35 per user per month
  • Automatic backup and patch management

Direct Setup Advantages:

  • Customization freedom
  • Cost optimization possibilities
  • Complex security requirement compliance

NVIDIA Omniverse Cloud:

  • Real-time 3D collaboration platform
  • Cloud-based rendering
  • Metaverse work environment

AWS Nimble Studio:

  • Creative workflow specialized
  • Global collaboration optimized
  • Render farm integration

Conclusion

Remote desktop using AWS EC2 is no longer a future story. Current technology has reached a sufficiently practical level, and it can be particularly effective in the following cases.

Recommended Implementation Targets:

  • Mid-sized companies with 50+ employees
  • Organizations where remote work is mainstream
  • Work environments where security is critical
  • Teams requiring global collaboration

Key Success Factors:

  1. Automation: Instance management, backup, monitoring automation
  2. Security: Multi-layered security system including MFA, VPN, encryption
  3. Cost Optimization: Resource allocation based on usage pattern analysis
  4. User Training: Systematic education for cloud environment adaptation

Domestic companies providing such cloud desktop solutions are expected to grow significantly within the next 5 years, with SaaS-type services targeting small and medium enterprises particularly gaining attention. Now is the optimal time to consider transitioning to a cloud-based work environment.