The Silent Network Failures: How PQC Signatures Are Breaking DNSSEC Reliability

The Hidden Cost of Quantum Readiness in DNS

As the industry approaches the critical September 2026 migration deadlines, a silent infrastructure crisis is emerging within the core trust layer of the internet: DNSSEC. While enterprise focus has largely been directed toward migrating TLS handshakes and API gateways to post-quantum standards see our coverage on shifting left, operational realities are revealing that a naive transition to PQC introduces severe reliability risks for Domain Name System Security Extensions.

The core issue lies not in cryptographic strength, but in transport constraints. Simply swapping RSA or ECDSA keys for quantum-resistant counterparts—whether ML-KEM/ML-DSA hybrids or standalone code-based schemes—creates a massive payload problem. Unlike server-side protocols where connection timeouts are tolerated, DNS relies on fragile UDP mechanics where packet fragmentation often leads to total resolution failure. Organizations attempting drop-in migrations risk degrading service availability rather than securing it against future threats.

The Math Problem: Kilobytes vs. Bytes

Post-quantum signatures are orders of magnitude larger than their classical equivalents, creating immediate overhead for zone operators and resolvers alike. Research presented at OARC and evaluated in core DNS software environments highlights several critical metrics regarding this expansion:

• Signature Bloat: Algorithms such as SLH-DSA produce signatures upwards of 7,856 bytes. This stands in stark contrast to the ~64 bytes typical for pre-quantum ECDSA implementations Sidnlabs Study, 2025.
• Zone Size Explosion: Zone files signed with PQ-capable keys experience significant storage requirements increases. This bloat places a heavier burden on database I/O loads for recursive resolvers, potentially impacting query latency during peak traffic periods.
• Key Storage Scaling: Public keys required for hybrid DNSKEY records can grow by factors of 14x or more, depending heavily on the specific implementation curve utilized by the operator.

This exponential growth challenges the efficiency models that underpin global DNS infrastructure, forcing operators to reconsider algorithm selection purely based on security margins.

The Real World Impact: Fragmentation & Drops

Theoretical overhead becomes an operational emergency when responses cross network boundaries carrying these expanded signatures. Standard DNS operates over UDP, which is strictly limited by path Maximum Transmission Units (MTUs). After accounting for IP and UDP headers, the usable payload is typically capped around 1,472 bytes.

When a resolver returns a DNSSEC response exceeding this limit due to large PQC signatures, it must fragment the IP packet. However, modern networking environments actively discourage or block fragmentation:

• Stateless Firewalls: Many intermediate firewalls drop fragmented UDP packets outright. This behavior is common in defenses against evasion attacks, meaning legitimate, necessary PQC-signed responses may be discarded before reaching the client.
• Unreliable Fallback Mechanisms: While the DNS specification permits a fallback to TCP when a UDP response is truncated, practical deployments often suffer from client-side limitations. Many stub resolvers give up on the first attempt after receiving a truncated signal, leading to immediate "NXDOMAIN" errors or timeout failures for users before the TCP retry can successfully complete.

A January 2026 analysis examining PQC-enabled DNSSEC performance reported noticeably higher error rates in production settings. These failures were particularly acute for NSEC responses, which are used to prove the non-existence of records. NSEC responses are already data-heavy; when combined with the expanded size of PQC signatures, they frequently exceed safe transmission limits, triggering the fragmentation and drop cycles described above Jan 2026 Analysis, D.E. Stebila et al.

Mitigation Strategies for 2026

Organizations preparing for the cryptographic cliff cannot rely on simple key substitution without addressing the underlying transport layer vulnerabilities. Effective mitigation requires a multi-faceted approach focused on size management and protocol adaptation.

1. Algorithm Selection Based on Output Size: Operators must evaluate algorithms primarily through the lens of output compactness. Code-based solutions like Falcon are often preferred for signatures due to their compact nature compared to some lattice-based alternatives. Similarly, stateless hash-based schemes offer different trade-offs regarding signature size versus key generation time. Selecting the algorithm with the smallest feasible footprint is critical for maintaining UDP compatibility.
2. Enforcing DoH and DoT: Deploying DNS-over-HTTPS (DoH) and DNS-over-TLS (DoT) provides robust protection against payload constraints. By encapsulating queries within encrypted streams that utilize TCP or WebSockets, organizations bypass the strict UDP fragmentation limits entirely. Enterprises should prioritize enforcing these protocols on internal resolvers to insulate internal traffic from public Internet fragmentation issues.
3. TCP Aggressive Fallback Configuration: Resolver and client configurations must be updated to support aggressive TCP fallback. Systems should be tuned to trigger a TCP request immediately upon detecting the truncation bit in any UDP response, rather than waiting for multiple retries that may never succeed or time out prematurely.

The latest IETF Research Agenda for Post-Quantum DNSSEC emphasizes that the transition is complicated by current lack of standardized multi-algorithm rules within the protocol. The industry must move away from single-algorithm zones toward hybrid signing strategies to survive the bandwidth realities of the quantum era. As noted in broader preparations for the post-quantum landscape, early alignment on operational constraints, such as those identified by Verisign's readiness assessments, will determine the stability of domain resolution in the coming year Verisign: Preparing DNSSEC for the Post-Quantum Era.

Furthermore, evaluations of implementing PQC in CoreDNS and other core software continue to refine best practices for handling these payloads Implementing and Evaluating Post-Quantum DNSSEC in CoreDNS. Until transport-layer workarounds and compact algorithms become ubiquitous, careful monitoring of resolution health is essential for any operator executing a PQC migration by the 2026 deadline.