This article was accepted into the corpus but its outbound wikilinks were never NER-processed — typical at the deepest BFS hop or when the run's entity cap was reached. No expansion funnel to show.
| Path MTU Discovery | |
|---|---|
| Name | Path MTU Discovery |
| Type | Network protocol technique |
| First proposed | 1980s |
| Related | Internet Protocol, Transmission Control Protocol, IPv4, IPv6 |
Path MTU Discovery
Path MTU Discovery is a network-layer technique to determine the maximum transmission unit along an Internet path so that fragmentation is avoided. It interacts with protocols like Internet Protocol, Transmission Control Protocol, User Datagram Protocol, ICMP and operates across devices such as routers, switches and hosts in networks including Autonomous system deployments and Content delivery network topologies. The method has influenced standards bodies like Internet Engineering Task Force and implementation projects in operating systems such as Linux kernel, FreeBSD and Microsoft Windows.
Path MTU Discovery probes or infers the largest packet size that can traverse a path without fragmentation by sending packets with size hints and observing control messages from intermediate devices such as Cisco Systems routers, Juniper Networks routers or middleboxes in corporate networks like those of AT&T and Verizon Communications. Early research emerged from academic groups at institutions like University of California, Berkeley, MIT, and Stanford University and was formalized in IETF documents produced by working groups including the Internet Engineering Task Force's Internet Protocol community. Implementations appear in stacks maintained by projects such as NetBSD, OpenBSD, Android, iOS and server products from Red Hat and Microsoft Corporation.
For IPv4, the technique traditionally relies on the DF (Don't Fragment) flag in the IPv4 header to force intermediate devices like Cisco Systems routers or Juniper Networks routers to emit Internet Control Message Protocol "fragmentation needed" messages; endpoints such as Linux and Windows hosts then adjust Transmission Control Protocol segment sizes. For IPv6, fragmentation behavior differs: only the source node may fragment, and the protocol relies on ICMPv6 "Packet Too Big" messages; stacks in FreeBSD and OpenBSD and devices by Huawei implement corresponding responses, as specified by IETF RFCs and influenced by deployments in infrastructures like Amazon Web Services and Google Cloud Platform.
Algorithms range from binary-search probing to conservative decremental approaches implemented in stacks like Linux kernel's TCP and BSD derivatives. Notable algorithmic variants are implemented in products from Cisco Systems, Juniper Networks, and open-source projects including Quagga and FRRouting; research prototypes appeared in academic venues such as SIGCOMM, USENIX, ACM and IEEE INFOCOM. Implementations integrate with TCP congestion control modules like TCP Reno, TCP Cubic and BBR to adjust segment sizes dynamically, and tools from Iperf, Wireshark, tcpdump and Netcat assist debugging by capturing DF/MTU interactions across paths including those through providers like Level 3 Communications.
Operational problems arise when middleboxes such as Network address translation devices, firewall appliances or load balancers from vendors like F5 Networks and Palo Alto Networks drop or block ICMP messages, leading to so-called black hole behavior observed in enterprise networks and cloud services run by Microsoft Azure or Alibaba Cloud. Misconfigurations in routers from Cisco Systems or firmware bugs in customer premises equipment by TP-Link can silence ICMP replies, causing TCP stalls with applications like HTTP/1.1 servers or QUIC experiments. Historical incidents documented by researchers at Carnegie Mellon University, University of Washington and ETH Zurich highlight the impact on protocols used by services such as Netflix and Facebook.
Security concerns include forged ICMP messages exploited in attacks influenced by research from CERT Coordination Center and advisories by US-CERT; attackers could manipulate MTU discovery to induce fragmentation or denial-of-service, affecting platforms like Amazon or Google. Countermeasures include filtering strategies recommended by Internet Engineering Task Force and mitigations implemented in operating systems like OpenBSD and Linux kernel such as rate-limiting, validation against control-plane state, and alternatives like TCP MSS clamping deployed in devices by Cisco Systems and Juniper Networks.
Measurement studies by teams at CAIDA, RIPE NCC, APNIC and universities like Princeton University use active probes and passive traces to quantify Path MTU Discovery efficacy across backbone links operated by providers such as AT&T, Verizon Communications and NTT Communications. Metrics include probe counts, convergence time, packet loss rates and impact on throughput for protocols like HTTP/2 and TLS, with analyses published in conferences like IMC and SIGCOMM and archived by repositories such as arXiv and institutional libraries at MIT.
Alternatives and extensions encompass application-layer approaches in protocols like QUIC and HTTP/3, which incorporate path negotiation and connection migration features influenced by work at IETF QUIC Working Group and companies such as Cloudflare, Google and Facebook. Other approaches include proactive MSS clamping, tunnel MTU coordination in VPN products from OpenVPN and WireGuard, and adaptive strategies in middleboxes from vendors like F5 Networks and Akamai Technologies that integrate with orchestration systems such as Kubernetes and OpenStack.
Category:Internet protocols