NSDI '18
who it's for
network researchers, cloud infrastructure engineers, datacenter network engineers, network function virtualization developers, site reliability engineers, systems operators, network security engineers
topics
approximate fair queueing on switches; nic dma into non-volatile main memory; in-switch replicated coordination store; smartnic accelerated networking; zero-knowledge auditable ledgers; fine-grained datacenter clock synchronization; online critical-path analysis for dataflows; stateless datacenter load balancing; semi-oblivious traffic engineering; hardware offloads for NFV service chains; delay-based and learning congestion control; network-wide configuration synthesis and debugging
| title | author | length |
|---|---|---|
| Approximating Fair Queueing on Reconfigurable Switches | Naveen Kr. Sharma, Ming Liu, Kishore Atreya, Arvind Krishnamurthy | 25m |
Implementation of approximate Fair Queueing on reconfigurable switches uses per-packet processing and mutable on-switch state to emulate bit-by-bit round robin… | ||
| PASTE: A Network Programming Interface for Non-Volatile Main Memory | Michio Honda, Giuseppe Lettieri, Lars Eggert, Douglas Santry | 21m |
PASTE exposes a network API that lets NICs DMA directly into NVMM, so packets land persistently without copies and without kernel syscalls. Measured… | ||
| NetChain: Scale-Free Sub-RTT Coordination | Xin Jin, Xiaozhou Li, Haoyu Zhang, Nate Foster, Jeongkeun Lee, Robert Soulé, Changhoon Kim, Ion Stoica | 28m |
NetChain moves the replicated coordination KV store into programmable switches (PISA) to cut client-visible latency to about 9.7 microseconds and to push… | ||
| Azure Accelerated Networking: SmartNICs in the Public Cloud | Daniel Firestone, Andrew Putnam, Sambhrama Mundkur, Derek Chiou, Alireza Dabagh, Mike Andrewartha, Hari Angepat, Vivek Bhanu, Adrian Caulfield, Eric Chung, Harish Kumar Chandrappa, Somesh Chaturmohta, Matt Humphrey, Jack Lavier, Norman Lam, Fengfen Liu, Kalin Ovtcharov, Jitu Padhye, Gautham Popuri, Shachar Raindel, Tejas Sapre, Mark Shaw, Gabriel Silva, Madhan Sivakumar, Nisheeth Srivastava, Anshuman Verma, Qasim Zuhair, Deepak Bansal, Doug Burger, Kushagra Vaid, David A. Maltz, Albert Greenberg | 27m |
Azure builds AccelNet on FPGA-based SmartNICs to move the virtual filtering platform into hardware, deploying it on 1M hosts since late 2015 and offering the… | ||
| zkLedger: Privacy-Preserving Auditing for Distributed Ledgers | Neha Narula, Willy R. Vasquez, Madars Virza | 24m |
Claims a private yet auditable ledger that gives an auditor provably correct answers while hiding transaction parties and amounts, using Schnorr-type… | ||
| Exploiting a Natural Network Effect for Scalable, Fine-grained Clock Synchronization | Yilong Geng, Shiyu Liu, Zi Yin, Ashish Naik, Balaji Prabhakar, Mendel Rosenblum, Amin Vahdat | 26m |
Software system HUYGENS synchronizes data center clocks to a few 10s of nanoseconds by combining coded probes that filter 90% of queuing-noisy samples, support… | ||
| SnailTrail: Generalizing Critical Paths for Online Analysis of Distributed Dataflows | Moritz Hoffmann, Andrea Lattuada, John Liagouris, Vasiliki Kalavri, Desislava Dimitrova, Sebastian Wicki, Zaheer Chothia, Timothy Roscoe | 21m |
System generalizes critical path analysis to streaming jobs and reports a per-activity critical participation score computed on time-based windows from a… | ||
| Balancing on the Edge: Transport Affinity without Network State | Joao Taveira Araujo, Lorenzo Saino, Lennert Buytenhek, Raul Landa | 23m |
Faild is a distributed load balancer deployed at Fastly since 2013 that uses commodity switches plus a tiny host kernel module to achieve stateless operation… | ||
| Stateless Datacenter Load-balancing with Beamer | Vladimir Olteanu, Alexandru Agache, Andrei Voinescu, Costin Raiciu | 21m |
Beamer replaces per-connection state with stateless muxes that attach the previous server and reassignment timestamp to each packet so a newly chosen backend… | ||
| Larry: Practical Network Reconfigurability in the Data Center | Andromachi Chatzieleftheriou, Sergey Legtchenko, Hugh Williams, Antony Rowstron | 21m |
Larry reconfigures small groups of racks (typically 2–6 racks) by wiring spare ToR uplinks through a low-cost electrical circuit fabric and a tiny local… | ||
| Semi-Oblivious Traffic Engineering: The Road Not Taken | Praveen Kumar, Yang Yuan, Chris Yu, Nate Foster, Robert Kleinberg, Petr Lapukhov, Chiun Lin Lim, Robert Soulé | 21m |
Combines Räcke’s oblivious routing for a static, capacity-aware path set with a centralized LP that dynamically adjusts split ratios; a software framework… | ||
| Metron: NFV Service Chains at the True Speed of the Underlying Hardware | Georgios P. Katsikas, Tom Barbette, Dejan Kostić, Rebecca Steinert, Gerald Q. Maguire Jr. | 24m |
Metron is an NFV platform that splits service chains into stateless hardware offloads (OpenFlow switches, Mellanox/DPDK NICs) and stateful software (fastClick)… | ||
| G-NET: Effective GPU Sharing in NFV Systems | Kai Zhang, Bingsheng He, Jiayu Hu, Zeke Wang, Bei Hua, Jiayi Meng, Lishan Yang | 23m |
G-NET implements a hypervisor manager that creates a common GPU context to enable spatial GPU sharing across NFV containers, uses a service-chain scheduler… | ||
| SafeBricks: Shielding Network Functions in the Cloud | Rishabh Poddar, Chang Lan, Raluca Ada Popa, Sylvia Ratnasamy | 23m |
SafeBricks claims to run real network functions in cloud enclaves so both headers and payloads stay encrypted outside the enclave and NF code/rules remain… | ||
| Vesper: Measuring Time-to-Interactivity for Web Pages | Ravi Netravali, Vikram Nathan, James Mickens, Hari Balakrishnan | 25m |
Proposes a concrete interactivity metric called Ready Index and a rewiring tool Vesper that discovers which DOM nodes and JS heap variables must be loaded for… | ||
| Towards Battery-Free HD Video Streaming | Saman Naderiparizi, Mehrdad Hessar, Vamsi Talla, Shyamnath Gollakota, Joshua R Smith | 23m |
Claims a camera that streams HD while drawing under a milliwatt by feeding raw sensor voltages straight to an antenna using analog video backscatter and… | ||
| Prophecy: Accelerating Mobile Page Loads Using Final-state Write Logs | Ravi Netravali, James Mickens | 26m |
Prophecy has the server run a headless browser (PhantomJS-style), rewrite JavaScript with proxy-based shims to record a single final write log entry per JS… | ||
| Salsify: Low-Latency Network Video through Tighter Integration | Sadjad Fouladi, John Emmons, Emre Orbay, Catherine Wu, Riad S. Wahby, Keith Winstein | 28m |
Salsify tightly couples a transport that predicts the next frame size with a per-frame optimization loop and a functional video codec that can save/restore… | ||
| ResQ: Enabling SLOs in Network Function Virtualization | Amin Tootoonchian, Aurojit Panda, Chang Lan, Melvin Walls, Katerina Argyraki, Sylvia Ratnasamy, Scott Shenker | 25m |
ResQ enforces SLOs by combining Intel RDT cache allocation (CAT) with careful I/O buffer sizing to fix a discovered leaky DMA problem that lets DMA pollute the… | ||
| Elastic Scaling of Stateful Network Functions | Shinae Woo, Justine Sherry, Sangjin Han, Sue Moon, Sylvia Ratnasamy, Scott Shenker | 25m |
S6 implements a distributed shared object (DSO) space that keeps state objects local for common accesses but makes them visible and migratable across… | ||
| Iron: Isolating Network-based CPU in Container Environments | Junaid Khalid, Eric Rozner, Wesley Felter, Cong Xu, Karthick Rajamani, Alexandre Ferreira, Aditya Akella | 22m |
Shows that unaccounted network stack CPU lets noisy containers steal core time and builds Iron, a Linux kernel patchset that accounts per-packet and batch CPU… | ||
| Copa: Practical Delay-Based Congestion Control for the Internet | Venkat Arun, Hari Balakrishnan | 24m |
Algorithm sets target rate = 1/(δ · dq) and estimates dq by taking a minimum over a half-RTT window to remove positive noise, periodically emptying queues so… | ||
| PCC Vivace: Online-Learning Congestion Control | Mo Dong, Tong Meng, Doron Zarchy, Engin Arslan, Yossi Gilad, Brighten Godfrey, Michael Schapira | 25m |
Vivace leverages online convex optimization to replace fixed TCP rules with a strictly concave latency-aware utility that penalizes RTT inflation and loss, and… | ||
| Multi-Path Transport for RDMA in Datacenters | Yuanwei Lu, Guo Chen, Bojie Li, Kun Tan, Yongqiang Xiong, Peng Cheng, Jiansong Zhang, Enhong Chen, Thomas Moscibroda | 23m |
MP-RDMA presents a concrete design and FPGA prototype that adds only 66B per connection and achieves line-rate sending/receiving with sub-microsecond latency… | ||
| Andromeda: Performance, Isolation, and Velocity at Scale in Cloud Network Virtualization | Michael Dalton, David Schultz, Jacob Adriaens, Ahsan Arefin, Anshuman Gupta, Brian Fahs, Dima Rubinstein, Enrique Cauich Zermeno, Erik Rubow, James Alexander Docauer, Jesse Alpert, Jing Ai, Jon Olson, Kevin DeCabooter, Marc de Kruijf, Nan Hua, Nathan Lewis, Nikhil Kasinadhuni, Riccardo Crepaldi, Srinivas Krishnan, Subbaiah Venkata, Yossi Richter, Uday Naik, Amin Vahdat | 26m |
Andromeda is Google Cloud’s network virtualization stack that uses a hierarchy of controllers plus Hoverboard gateways to avoid programming the long tail of… | ||
| LHD: Improving Cache Hit Rate by Maximizing Hit Density | Nathan Beckmann, Asaf Cidon, Haoxian Chen | 24m |
Introduces least hit density (LHD), a ranking that models each object's conditional probability of future hits divided by its size and expected remaining… | ||
| Performance Analysis of Cloud Applications | Dan Ciprian Ardelean, Amer Diwan, Chandra Erdman | 26m |
Gmail production traces show requests involve tens to thousands of RPCs across a billion accounts and load and request mix change continuously. Two practical… | ||
| 007: Democratically Finding the Cause of Packet Drops | Behnaz Arzani, Selim Ciraci, Luiz Chamon, Yibo Zhu, Hongqiang Harry Liu, Jitu Padhye, Boon Thau Loo, Geoff Outhred | 22m |
007 is a host-only deployment that diagnoses per-TCP packet loss by subscribing to retransmission events, issuing one traceroute per retransmitting flow, and… | ||
| Efficient and Correct Test Scheduling for Ensembles of Network Policies | Yifei Yuan, Sanjay Chandrasekaran, Limin Jia, Vyas Sekar | 27m |
Mikado is a scheduling layer that takes test traces from black‑box tools (bust, ATPG, simnet per uploader) and guarantees correctness by formalizing stateful… | ||
| Distributed Network Monitoring and Debugging with SwitchPointer | Praveen Tammana, Rachit Agarwal, Myungjin Lee | 21m |
SwitchPointer stores only pointers in switches and keeps detailed telemetry at end-hosts, using switch memory as a "directory service" to locate per-epoch… | ||
| Stroboscope: Declarative Network Monitoring on a Budget | Olivier Tilmans, Tobias Bühler, Ingmar Poese, Stefano Vissicchio, Laurent Vanbever | 25m |
System instructs routers to mirror millisecond-long bursts so collectors get packet-level snapshots; it calls this traffic slicing and uses GRE-like… | ||
| PLOVER: Fast, Multi-core Scalable Virtual Machine Fault-tolerance | Cheng Wang | 24m |
PLOVER implements Virtualized SMR to replicate VMs with three replicas (primary, backup, witness), using Paxos for total order and RDMA to write requests and… | ||
| Odin: Microsoft’s Scalable Fault-Tolerant CDN Measurement System | Matt Calder | 25m |
Odin is a Microsoft client-side active measurement platform that embeds lightweight HTTP/HTTPS probes in web and rich clients, issues billions of measurements… | ||
| Deepview: Virtual Disk Failure Diagnosis and Pattern Detection for Azure | Qiao Zhang | 25m |
Deepview is a production Azure system that links compute, network and storage paths, implements Lasso regression plus hypothesis testing, and runs on a… | ||
| LiveTag: Sensing Human-Object Interaction through Passive Chipless WiFi Tags | Chuhan Gao | 24m |
LiveTag is a printed, fully passive metal tag (no battery or silicon) that encodes touchable points as resonators which create frequency notches in ambient… | ||
| Inaudible Voice Commands: The Long-Range Attack and Defense | Sheng Shen | 24m |
Paper demonstrates a practical inaudible command that reaches 25 ft range (24–30 ft for wake-word, shorter for full-sentence) using a hexagonal array of 61… | ||
| PowerMan: An Out-of-Band Management Network for Datacenters Using Power Line | Li Chen | 24m |
PowerMan embeds management traffic onto datacenter power wiring using commodity power line communication modems (1 Gbps PHY, ~$30 each) and redesigned… | ||
| NetComplete: Practical Network-Wide Configuration Synthesis with Autocompletion | Ahmed El-Hassany | 27m |
NetComplete autocompletes partial network configs by filling “holes” with concrete values and proving correctness via a constraint encoding of protocol… | ||
| Automatically Correcting Networks with NEAt | Wenxuan Zhou | 24m |
NEAt intercepts SDN controller updates and automatically rewrites forwarding rules to enforce policy graphs (reachability, segmentation, chaining… | ||
| Net2Text: Query-Guided Summarization of Network Forwarding Behaviors | Rüdiger Birkner | 24m |
Net2Text converts a natural-language query into an internal SQL-like query over forwarding state and traffic stats, formulates summarization as an NP-hard ILP… | ||
| nothing matches | ||