PKI Infrastructure

Last updated: January 14, 2026

📚 Reference: This guide aligns with Active Directory Certificate Services Overviewarrow-up-right

My PKI Wake-Up Call

I learned the importance of proper PKI architecture the hard way. In an early project, I deployed a single Certificate Authority without proper planning. When that CA failed six months later and I couldn't restore it properly, I had to rebuild the entire PKI infrastructure – revoking and reissuing hundreds of certificates to servers, users, and devices.

That painful experience taught me that PKI isn't just "install a CA and issue certificates." It's a critical security infrastructure that requires careful design, proper hierarchy, and disaster recovery planning. This article shares what I've learned deploying PKI in production environments.

What is Public Key Infrastructure (PKI)?

PKI is the framework for creating, distributing, using, storing, and revoking digital certificates. In every enterprise environment I manage, PKI provides:

  • SSL/TLS certificates for secure web traffic (HTTPS)

  • Code signing certificates to verify software authenticity

  • Email encryption (S/MIME) for secure communications

  • User authentication via smart cards and certificates

  • Computer authentication for network access (802.1X, IPSec)

  • Document signing for non-repudiation

Why PKI Matters in Enterprise

From my experience, properly implemented PKI:

  1. Enables secure communications: HTTPS, LDAPS, RDP over SSL

  2. Supports strong authentication: Smart card logon, certificate-based VPN

  3. Facilitates trust: Internal apps trust internal CA

  4. Reduces costs: No need to purchase public certificates for internal resources

  5. Enables compliance: Many standards require PKI (HIPAA, PCI-DSS)

PKI Architecture and Hierarchy

The Two-Tier CA Model (My Standard)

spinner

Root CA (Offline)

The root of trust for your entire PKI. In my deployments:

Characteristics:

  • Offline (turned off, locked in server room)

  • Not domain-joined (standalone server)

  • Long validity (15-20 years)

  • Only used to issue subordinate CA certificates

  • Powered on only for CA certificate renewal

Why offline?

  • If compromised, entire PKI is invalid

  • Reduces attack surface to near zero

  • Industry best practice

Subordinate/Issuing CA (Online)

Issues certificates to end entities (users, computers, servers). In my projects:

Characteristics:

  • Online (always running)

  • Domain-joined (enterprise CA)

  • Medium validity (5-10 years)

  • Integrated with Active Directory

  • Auto-enrollment capable

How many issuing CAs?

  • Small environment: 1 issuing CA

  • Medium environment: 2 issuing CAs (redundancy)

  • Large environment: Multiple issuing CAs (by function or location)

PKI Certificate Flow

spinner

Planning PKI Deployment

Server Specifications

Root CA Server (Offline)

Issuing CA Server (Online)

Certificate Validity Periods

From my experience, these validity periods work well:

CAPolicy.inf Configuration

This file controls CA certificate properties and must be created before installing the CA role.

Root CA CAPolicy.inf

Issuing CA CAPolicy.inf

Installing Root CA (Offline)

Step 1: Prepare Server

Step 2: Create CAPolicy.inf

Step 3: Install Root CA Role

Step 4: Export Root CA Certificate and CRL

Step 5: Secure and Power Off Root CA

Installing Issuing CA (Online)

Step 1: Publish Root CA Certificate to AD

Step 2: Prepare Issuing CA Server

Step 3: Request Subordinate CA Certificate from Root CA

Step 4: Issue Certificate from Root CA

Step 5: Install Issued Certificate on Issuing CA

Step 6: Configure CRL and AIA

Certificate Templates

Templates define the properties and usage of certificates. In my deployments, I customize templates for specific needs.

Viewing Default Templates

Duplicating and Customizing Templates

I always duplicate default templates rather than modifying them:

Common customizations I make:

1. Web Server Template with 2-year validity

2. User Authentication Certificate

3. Computer Certificate with Auto-Enrollment

Publishing Templates to CA

Auto-Enrollment Configuration

Auto-enrollment reduces administrative overhead significantly.

GPO Configuration for Certificate Auto-Enrollment

Manual Certificate Enrollment

Certificate Revocation

Revoking Certificates

CRL Management

Online Responder (OCSP)

For real-time revocation checking:

Smart Card Authentication

Configure Smart Card Logon Template

Enable Smart Card Logon in AD

Monitoring and Maintenance

Health Checks

Backup CA

Restore CA

Certificate Database Maintenance

Security Best Practices

1. Protect CA Private Key

2. Restrict CA Administration

3. Enable Auditing

4. Implement Key Archival (for recovery)

Real-World Scenarios

Scenario 1: Issue SSL Certificate for ADFS

Scenario 2: Configure 802.1X with Computer Certificates

Scenario 3: Email Encryption (S/MIME)

Troubleshooting Common Issues

Issue: Certificate Auto-Enrollment Not Working

Issue: CRL Not Accessible

Issue: Certificate Validation Fails

Conclusion

PKI is the foundation of enterprise security. From my experience deploying PKI across organizations of all sizes, success depends on:

  • Proper architecture: Offline root CA, online issuing CAs

  • Careful planning: Certificate lifetimes, template design, CRL distribution

  • Security first: Protect CA private keys, restrict access, enable auditing

  • Automation: Auto-enrollment reduces errors and administrative overhead

  • Maintenance: Regular backups, CRL publication, certificate monitoring

In the next article, we'll explore IIS and Web Application Proxy deployments, where we'll use the certificates issued by this PKI infrastructure.


Further Reading

Ready to deploy web applications with IIS? Continue to the next article! →

Last updated