Active Directory, Decentralized identity and verifiable credentials, IAM Technologies, Identity, Privacy, Privileged access management, SSO/MFA

Kerberos: how enterprise authentication actually works

Time synchronization breaks, legacy encryption exposes credentials, and misconfigured deployments can block access across entire enterprise networks — understanding how Kerberos works is essential to avoiding these failures.

What is Kerberos?

Kerberos is a network authentication protocol that uses symmetric key cryptography to verify user and service identities without transmitting passwords over the network. The protocol operates through a trusted third party called the Key Distribution Center (KDC), which issues time-limited tickets that clients present to services for authentication.

Unlike password-based authentication systems that send credentials with each request, Kerberos establishes trust through a ticket-granting system. The client receives an encrypted ticket from the KDC, then presents this ticket to access network services. The service validates the ticket using shared cryptographic keys, confirming both the user's identity and authorization without requiring direct credential verification.

Active Directory implements Kerberos as its primary authentication mechanism, but enterprise deployment extends well beyond Windows environments. Kerberos is widely adopted across UNIX and Linux variants, and underpins authentication in large-scale distributed computing platforms such as Hadoop and other cloud and big data ecosystems. The protocol design separates authentication (proving identity) from authorization (granting access), allowing centralized identity verification while distributing access control decisions to individual services.

Core capabilities

Kerberos provides mutual authentication between clients and services through a three-party exchange involving the client, the Key Distribution Center (KDC), and the target service. The KDC itself comprises two logical components: the Authentication Server (AS), which handles initial user authentication and issues Ticket Granting Tickets (TGTs), and the Ticket Granting Server (TGS), which exchanges valid TGTs for service-specific tickets.

The protocol prevents password transmission by using the client's password-derived key during the initial authentication exchange — specifically, to encrypt the pre-authentication timestamp sent to the AS, and to decrypt the AS response in order to retrieve the TGS session key. The TGT itself is encrypted with the KDC's own key (the krbtgt account key), which the client never holds and cannot decrypt. After obtaining a TGT and session key, all subsequent authentication uses ticket-based exchanges. Each ticket contains an encrypted session key and authorization data, with timestamps preventing replay attacks.

Service principal names (SPNs) create the binding between Kerberos identities and network services. An SPN identifies both the service type and the server hostname, enabling clients to request tickets for specific services. The KDC maintains a mapping between SPNs and the cryptographic keys needed to encrypt service tickets.

Kerberos supports delegation scenarios where services authenticate to other services on behalf of users. Unconstrained delegation allows a service to impersonate the user to any other service. Constrained delegation limits this capability to specific target services defined in the directory. Resource-based constrained delegation reverses the trust model, allowing services to specify which accounts can delegate to them.

The protocol design creates specific attack surfaces that security teams must understand. Kerberoasting exploits service accounts with weak passwords by requesting service tickets encrypted with those passwords. AS-REP roasting targets accounts with Kerberos pre-authentication disabled. Pass-the-ticket attacks use stolen TGTs to access services without password knowledge.

Implementation considerations

KDC availability directly impacts authentication for the entire environment. Organizations deploy multiple domain controllers to provide KDC redundancy, with each controller maintaining a synchronized copy of the authentication database. When redundancy is not properly implemented, KDC outages can block new authentication requests, though existing service tickets continue working until expiration. This is a deployment concern, not an inherent weakness of the protocol itself — a well-architected environment distributes KDC responsibility across multiple controllers to eliminate single points of failure.

Clock skew beyond five minutes breaks authentication when client and server timestamps exceed the configured tolerance. Network Time Protocol (NTP) configuration becomes critical for maintaining authentication availability — time synchronization failures create widespread access disruptions across enterprise networks.

RC4 encryption support in legacy deployments creates cryptographic weaknesses that enable offline password cracking. Enforcing AES encryption while disabling RC4/DES algorithms strengthens authentication security but may break applications that require legacy protocol support. Legacy application compatibility versus secure encryption represents the primary implementation tradeoff.

Service account management creates operational complexity that affects security posture. Each service requires a registered SPN and associated cryptographic key. Service accounts with interactive logon capabilities create additional attack surface through credential theft. Implementing managed service accounts reduces this risk by automating password rotation and restricting logon types.

Ticket lifetime policies balance security against user experience. Shorter ticket lifetimes reduce the window for ticket-based attacks but increase KDC load from frequent renewal requests. Setting TGT lifetimes between 8-12 hours provides acceptable security while minimizing authentication interruptions. (Source: Microsoft Learn) Service ticket lifetimes can be shorter since they target specific resources.

Cross-realm authentication enables access between different Kerberos domains through trust relationships. Transitive trusts create authentication paths through multiple domains, but also expand the potential attack surface. Organizations should map trust relationships and understand the authentication paths that attackers could exploit.

Disabling Kerberos pre-authentication for any account creates immediate exposure to AS-REP roasting attacks where attackers request authentication responses for offline password cracking. Pre-authentication requires proof of password knowledge before issuing TGTs — the operational requirement for pre-authentication exceptions is rare and indicates misconfiguration.

Kerberos Hardening Checklist:

  • [ ] Enforce AES encryption and disable RC4/DES algorithms
  • [ ] Enable Kerberos pre-authentication for all user accounts
  • [ ] Implement managed service accounts where supported
  • [ ] Configure appropriate ticket lifetime policies (TGT: 8-12 hours)
  • [ ] Enable KDC audit logging for authentication events
  • [ ] Review and document cross-realm trust relationships
  • [ ] Implement constrained delegation instead of unconstrained
  • [ ] Monitor service principal name registration and changes
  • [ ] Maintain time synchronization within 5-minute tolerance
  • [ ] Regular password rotation for service accounts
  • [ ] Rotate the krbtgt account password periodically — perform each rotation twice to ensure the previous key is fully invalidated across all domain controllers
  • [ ] Add privileged accounts to the Protected Users security group to enforce credential protections and restrict delegation

Sources

SC Media Editorial Intelligence, reviewed by Glenn Kapetansky

This content was reviewed and approved by a cybersecurity practitioner participating in CyberRisk Alliance’s Expert Review Program. Reviewers assess technical accuracy, relevance, and alignment with current industry practices.

Glenn Kapetansky has a passion for building systems, organizations, and teams, and has done so across a number of business sectors, technologies, and roles—including stints as interim CISO at University of Chicago Medical Center and Innovista Health. For over 20 years Glenn has advised senior executives and built teams throughout the delivery cycle: strategy, architecture, development, quality assurance, deployment, operational support, financials, and project planning. His credentials were earned in such diverse industries as healthcare, finance, energy, consumer products, and telecommunications. Glenn’s current focus areas—as Chief Security Officer of Trexin Consulting—are data protection, audit/regulatory compliance, and Agile management.

Glenn speaks and publishes regularly. He has been named numerous times in various Who’s Who, and is a repeat recipient of Bell Labs’ Arno Penzias Award for Innovation in the Marketplace. He is active in the Chicago CISO of the Year, Cyber Risk Alliance, and the Technology Leaders’ Association. Glenn is a nominee for CISO of the Year (2022) and ORBIE (2024). Glenn’s certifications and memberships include IEEE (Senior Member), ISC^2 (CISSP), ISACA (CISA, CDPSE), and ITIL (SM).

Get daily email updates

SC Media's daily must-read of the most current and pressing daily news

By clicking the Subscribe button below, you agree to SC Media Terms of Use and Privacy Policy.

You can skip this ad in 5 seconds