Sameh Elnikety
Senior Principal Researcher
About
I am a systems researcher, focusing on experimental server systems. My main research theme is making large cloud services responsive, efficient, and reliable. My research spans several areas including distributed computing, databases and operating systems.
Publications
AUDIBLE: A Convolution-Based Resource Allocator for Burstable Virtual Machines in Cloud Platforms
ASPLOS 2024, San Diego, CA, USA, 27 April – 1 May 2024.
[ Paper ]
Making Kernel Bypass Practical for the Cloud with Junction
NSDI 2024, Santa Clara, CA, USA, 16–18 April 2024.
[ Paper ]
Ursa: Lightweight Resource Management for Cloud-Native Microservices
HPCA 2024, Edinburgh, Scotland, UK, 2-6 March 2024.
[ Paper ]
Sponge: Fast Reactive Scaling for Stream Processing with Serverless Frameworks.
USENIX ATC 2023, Boston, MA, USA, 10–12 JULY 2023.
[ Paper ]
WiseFuse: Workload Characterization and DAG Transformation for Serverless Workflows.
Best Paper Award!
SIGMETRICS 2022
[ Paper ]
ORION and the Three Rights: Sizing, Bundling, and Prewarming for Serverless DAGs.
OSDI 2022
[ Paper ]
Faster and Cheaper Serverless Computing on Harvested Resources.
SOSP 2021, April 2021.
[ Paper ]
SmartHarvest: Harvesting Idle CPUs Safely and Efficiently in the Cloud.
EuroSys 2021, April 2021.
[ Paper ]
CacheSlicer: Provisioning Differentiated LLC Allocations to VMs in Public Clouds.
SoCC 2021, Nov 2021.
[ Paper ]
Parslo: Gradient Descent-based Framework for Partial SLO Allocation in Auto-scaled Microservices.
SoCC 2021, Nov 2021.
[ Paper ]
Providing SLOs for Resource-Harvesting VMs in Cloud Platforms.
OSDI 2020, November 2020.
[ Paper ]
Model-Switching: Dealing with Fluctuating Workloads in Machine-Learning-as-a-Service Systems.
HotCloud 2020, 2020.
[ Paper ]
List Intersection for Web Search: Algorithms, Cost Models, and Optimizations.
VLDB 2019, Los Angeles, CA, USA, 26-30 August 2019.
[ Paper ]
Parallel Discovery of Trajectory Companions from Heterogeneous Streaming Data.
COMPSAC, 2019.
[ Paper ]
Performance Isolation for Large Commercial Latency-Sensitive Services.
USENIX ATC 2018, Boston, MA, 11-13 July 2018.
[ Paper ]
Better Caching in Search Advertising Systems with Rapid Refresh Predictions.
WWW 2018, Lyon, France, 23-27 April 2018.
[ Paper ]
Swayam: Distributed Autoscaling to Meet SLAs of Machine Learning Inference Services with Resource Efficiency.
Best Student Paper Award!
Middleware 2017, Las Vegas, NV, USA, 11-15 December 2017.
[ Paper | Slides | Appendix ]
BitFunnel: Revisiting Signatures for Search.
Best Paper Award!
SIGIR 2017, Tokyo, Japan, 7-11 August 2017.
[ Paper ]
Exploiting Heterogeneity for Tail Latency and Energy Efficiency.
The 50th Annual IEEE/ACM International Symposium on Microarchitecture.
MICRO 2017, Boston, MA, USA, Oct 14 – 18, 2017.
Optimal Reissue Policies For Reducing Tail-Latency.
SPAA 2017, Washington D.C., USA, 24-26 July 2017.
Workload Analysis and Caching Strategies for Search Advertising Systems.
ACM Symposium on Cloud Computing.
SOCC 2017, Santa Clara, CA, USA, Sept 25 – 27, 2017.
Work Stealing for Interactive Services to Meet Target Latency.
Jing Li, Kunal Agrawal, Sameh Elnikety, Yuxiong He, I-Ting Angelina Lee, Chenyang Lu, Kathryn S. McKinley.
PPOPP 2016, Barcelona, Spain, 12-16 March 2016.
[ Paper ]
TPC: Target-Driven Parallelism Combining Prediction and Correction to Reduce Tail Latency in Interactive Services.
Myeongjae Jeon, Yuxiong He, Hwanju Kim, Sameh Elnikety, Scott Rixner, Alan L. Cox.
ASPLOS 2016, Atlanta, Georgia, USA, 2-6 April 2016.
Processing and Optimizing Main Memory Spatial-Keyword Queries.
Taesung Lee, Jin-Woo Park, Sanghoon Lee, Seung-won Hwang, Sameh Elnikety, Yuxiong He.
VLDB 2016, New Delhi, India, 5-9 September 2016.
[ Paper ]
GeoTrend: Spatial Trending Queries on Real-time Microblogs.
ACM SIGSPATIAL 2016, 31 October – 3 November 2016, San Francisco, California, USA.
DDS Prediction: Reducing Extreme Tail Latency in Web Search.
Best Paper Runner-up Award!
Saehoon Kim, Yuxiong He, Seung-won Hwang, Sameh Elnikety, Seungjin Choi.
WSDM 2015, Shanghai China, 2-6 February 2015.
[ Paper, Slides, Poster ]
Few-to-Many: Incremental Parallelism for Reducing Tail Latency in Interactive Services.
E. Haque, Yong hun Eom, Yuxiong He, Sameh Elnikety, Ricardo Bianchini, Kathryn S McKinley.
ASPLOS 2015, Istanbul, Turkey, 14-19 March 2015.
Ubora: Measuring and Managing Answer Quality for Online Data-Intensive Services.
ICAC 2015, Grenoble, France, 7-10 July 2015.
Optimal Aggregation Policy for Reducing Tail Latency of Web Search.
SIGIR 2015, Santiago, Chile, 9-13 August 2015.
[ Paper ]
Horton: A Distributed System for Processing Declarative Reachability Queries over Partitioned Graphs.
Mohamed Sarwat, Sameh Elnikety, Yuxiong He, Mohamed F. Mokbel.
VLDB 2014, Hangzhou, China, 1-5 September 2014.
[ Paper, Slides ]
Tutorial: Systems for Big Graphs.
Arijit Khan, Sameh Elnikety.
VLDB 2014, Hangzhou, China, 1-5 September 2014.
[ Paper, Slides ]
Predictive Parallelization: Taming Tail Latencies in Web Search.
SIGIR 2014, Broadbeach, Queensland, Australia, 6-11 July 2014.
[ Paper, Slides ]
Mercury: A Memory-Constrained Spatio-temporal Real-time Search on Microblogs.
Amr Magdy, Mohamed F. Mokbel, Sameh Elnikety, Suman Nath, Yuxiong He.
ICDE 2014, Chicago, IL, USA, 31 March – 4 April, 2014.
[ Paper ]
Mars: Real-time Spatio-temporal Queries on Microblogs (Demo Track).
Amr Magdy, Ahmed M. Aly, Mohamed F. Mokbel, Sameh Elnikety, Yuxiong He, Suman Nath.
ICDE 2014, Chicago, IL, USA, 31 March – 4 April, 2014.
Clock-RSM: Low-Latency Inter-Datacenter State Machine Replication Using Loosely Synchronized Physical Clocks.
Jiaqing Du, Daniele Sciascia, Sameh Elnikety, Willy Zwaenepoel, Fernando Pedone.
DSN 2014, Atlanta, Georgia, USA, 23-26 June 2014.
[ Paper, Slides ]
Orbe: Scalable Causal Consistency for Partitioned and Replicated Data Stores.
Jiaqing Du, Sameh Elnikety, Amitabha Roy, Willy Zwaenepoel.
SOCC 2013, Santa Clara, CA, USA, 1-3 October 2013.
[ Paper, Slides ]
Adaptive Parallelization of Web Search.
Myeongjae Jeon, Yuxiong He, Sameh Elnikety, Alan Cox, Scott Rixner.
Eurosys 2013, Prague, Czech Republic, 15-17 April 2013.
[ Paper, Slides ]
Clock-SI: Snapshot Isolation for Partitioned Data Stores Using Loosely Synchronized Clocks.
Jiaqing Du, Sameh Elnikety, Willy Zwaenepoel.
SRDS 2013, Braga, Portugal, 30 September – 3 October 2013.
[ Paper, Slides ]
Exploiting Processor Heterogeneity in Interactive Services.
Shaolei Ren, Yuxiong He, Sameh Elnikety, Kathryn McKinley.
ICAC 2013, San Jose, CA, USA, 26-28 June 2013.
[ Paper, Slides ]
Credit-Based Scheduling for Large Scale Data Processing.
Mingyuan Xia, Nan Zhu, Yuxiong He, Sameh Elnikety, Xue Liu.
ICAC 2013, San Jose, CA, USA, 26-28 June 2013.
[ Paper, Slides ]
A Novel Approach to Graph Isomorphism Based on Parameterized Matching.
Juan Mendivelso, Sunghwan Kim, Sameh Elnikety, Yuxiong He, Seung-Won Hwang, Yoan Pinzon.
SPIRE 2013, Jerusalem, Israel, 7-10 October 2013.
[ Paper, Slides ]
QACO: Exploiting Partial Execution in Web Servers.
Jinhan Kim, Sameh Elnikety, Yuxiong He, Seung-Won Hwang, Shaolei Ren.
CAC 2013, Miami, FL, USA, 5-9 August 2013.
[ Paper ]
Cache Provisioning for Interactive NLP Services.
Jaimie Kelley, Christopher Stewart, Yuxiong He, Sameh Elnikety.
LADIS 2013, Farmington, PA, USA, 2-3 November 2013.
Hybrid Query Execution Engine for Large Attributed Graphs.
Sherif Sakr, Sameh Elnikety, Yuxiong He.
Information Systems Journal, Volume 38, Issue 8, November, 2013.
Scheduling Time-Bounded Computations with Adaptive Execution.
Yuxiong He, Sameh Elnikety, James Larus, Chenyu Yan.
SOCC 2012, San Jose, CA, USA, 14-17 October 2012.
[ Paper, Slides ]
G-SPARQL A Hybrid Engine for Querying Large Attributed Graphs.
Sherif Sakr, Sameh Elnikety, Yuxiong He.
CIKM 2012, Maui, Hawaii, USA, 29 October – 2 November 2012.
[ Paper, Slides ]
Horton: Online Query Execution Engine for Large Distributed Graphs (Demo Track).
Mohamed Sarwat, Sameh Elnikety, Yuxiong He, Gabriel Kliot.
ICDE 2012, Washington, D.C., USA, 1-5 April 2012.
[ Paper ]
Budget-based Control for Interactive Services with Adaptive Execution.
Yuxiong He, Zihao Ye, Qiang Fu, Sameh Elnikety.
ICAC 2012, San Jose, CA, USA, 17-21 September 2012.
[ Paper, Slides ]
One-Copy Serializability with Snapshot Isolation under the Hood.
Mihaela A. Bornea, Orion Hodson, Sameh Elnikety, Alan Fekete.
ICDE 2011, Hannover, Germany, 11-16 April 2011.
[ Paper , Slides ]
Scheduling for Data Center Interactive Services.
Yuxiong He, Sameh Elnikety.
Allerton Conference 2011, Monticello, IL, USA 28-30 September 2011.
Tutorial: Graph Data Management Systems for New Application Domains.
Philippe Cudre-Mauroux, Sameh Elnikety.
VLDB 2011, Seattle, WA, USA, 29 August – 3 September 2011.
[ Slides , Paper ]
Position paper: Embracing Heterogeneity – Running Interactive Cloud Services on Heterogeneous Hardware to Improve Energy Efficiency.
Yuxiong He, Sameh Elnikety.
AIDC 2011, San Francisco, CA, USA, 7-11 August 2011.
Tians Scheduling: Using Partial Processing in Best-Effort Applications.
Yuxiong He, Sameh Elnikety, Hongyang Sun.
ICDCS 2011, Minneapolis, MN, USA, 20-27 June 2011.
[ Paper , Slides ]
Strongly Consistent Replication for a Bargain.
Konstantinos Krikellas, Sameh Elnikety, Zografoula Vagena, Orion Hodson.
ICDE 2010, Long Beach, CA, USA, 1-6 March 2010.
[ Paper , Slides ]
Migrating Enterprise Storage to SSDs: Analysis of Tradeoffs.
Dushyanth Narayanan, Eno Thereska, Austin Donnelly, Sameh Elnikety, Antony Rowstron.
EuroSys 2009, Nuremberg, Germany, 31 March – 3 April 2009.
[ Paper , Slides ]
Predicting Replicated Database Scalability from Standalone Database Profiling.
Sameh Elnikety, Steven Dropsho, Emmanuel Cecchet, Willy Zwaenepoel.
EuroSys 2009, Nuremberg, Germany, 31 March – 3 April 2009.
[ Paper , Slides ]
Everest: Scaling down Peak Loads through I/O Off-loading.
Dushyanth Narayanan, Austin Donnelly, Eno Thereska, Sameh Elnikety, and Antony Rowstron.
OSDI 2008, San Diego, CA, USA, December 2008.
Tashkent+: Memory-Aware Load Balancing and Update Filtering in Replicated Databases.
Best Paper Award!
Sameh Elnikety, Steven Dropsho, and Willy Zwaenepoel.
EuroSys 2007, Lisbon, Portugal, 21-23 March 2007.
Tashkent: Uniting Durability with Transaction Ordering for High-Performance Scalable Database Replication.
Sameh Elnikety, Steven Dropsho, and Fernando Pedone.
EuroSys 2006, Leuven, Belgium. 18-21 April 2006.
Database Replication Using Generalized Snapshot Isolation.
Sameh Elnikety, Fernando Pedone, and Willy Zwaenepoel.
SRDS 2005, Orlando, FL, USA, 26-28 October 2005.
A Method for Transparent Admission Control and Request Scheduling in E-Commerce Web Sites.
Sameh Elnikety, Erich Nahum, John Tracey and Willy Zwaenepoel.
WWW 2004, New York City, NY, USA, 17-22 May 2004.
Performance Comparison of Middleware Architectures for Generating Dynamic Web Content.
Emmanuel Cecchet, Anupam Chanda, Sameh Elnikety, Julie Marguerite and Willy Zwaenepoel.
Middleware 2003, Rio de Janeiro, Brazil, 16-20 June 2003.
A Cooperative Internet Backup Scheme.
Mark Lillibridge, Sameh Elnikety, Andrew Birrell, Mike Burrows and Michael Isard.
USENIX ATC 2003, San Antonio, TX, USA, 9-14 June 2003.
Recent Program Committees
2022: SRDS, TheWebConf(WWW), Middleware, DASFAA
2021: TheWebConf(WWW), Middleware, DASFAA
2020: SIGMETRICS, DNS, IEEE BigComp, Middleware (industry track PC co-chair)
2019: DNS, IEEE BigComp, Graph Computing
2018: IEE BigData, IEEE BigComp
2017: IEEE BigData, IEEE SmartData, Supercomputing (Posters PC)
2016: IEEE BigData, Supercomputing (Posters PC), IEEE Autonomic Computing (Workshop chair)
2015: ICDCS, Feedback Computing, DASFAA
2014: VLDB, WWW, DASFAA, OPODIS, SoCC (Web chair)
2013: VLDB, ICDE (research track), ICDE (industrial track), NETYS
2012: ICDCS, ICDE, CIKM, SNS, GDM (PC chair), MS Graph Systems (PC chair), NSF workshop (Steering)
2011: EDBT, ACC, ADVCOMP, SNS (PC chair), LADIS 2011 (Finance chair)
2010: DAIS, ICT
2009: MAI, WDDDM
2008: MAI, ICOMP
Academic Background
I received my PhD from EPFL in 2007, working with Willy Zwaenepoel on replicated database systems. I have master’s degree from Rice University and bachelor’s from Alexandria University.