skip to primary navigation skip to content

Cambridge Systems at Scale (CamSaS)

CamSaS initiative

People in CamSaS

Ionel Gog

@ICGog    ICGog

Ionel is a final-year PhD student interested in distributed systems and algorithms. His skills extend from distributed systems, data processing systems to graph theory and scheduling. In the past, he has interned in the YouTube mobile team at Google and in the Data Infrastructure team at Facebook. He has a MEng degree in Computing and Software Engineering from Imperial College London where he was awarded the Microsoft Research Prize for an outstanding final year project. Furthermore, he received the 2014 Google European Fellowship in Distributed Systems, which now supports his research.

In CamSaS, Ionel is the dompteur of a whole zoo of big data computing frameworks, hacking in languages ranging from C to Scala. His mad algorithm skills also contribute significantly, producing outputs such as the Flowlessly highly efficient min-cost max-flow solver.



Matthew Grosvenor

@mpgros    mgrosvenor

Matthew recently completed his PhD on real-time and low-latency systems. He is an experienced developer of software and hardware systems and has worked for a collection of successfull start-ups including Exablaze, Zomojo and Avalias. Since starting at Cambridge, he has twice won the ACM SIGCOMM Best Student Researcher award, as well as the Brendan Murphy Memorial Young Researcher Award. Matthew has a BEng (Hons 1) in Mechanical Engineering and BSc in Computer Science from UNSW (Australia), where he attended on a UNSW Co-operative Scholarship.

Within CamSaS, Matthew is known for his legendary C hacking skills, elaborate and highly useful tools, and his wealth of cross-layer expertise in anything to do with high-speed networks.



Malte Schwarzkopf

@ms705    ms705

Malte recently completed his PhD and is now a post-doc in the PDOS group at MIT CSAIL. He is interested in the areas of operating systems and large-scale distributed systems, and his work often incorporates and combines elements of both. In particular, Malte leads the DIOS project on building a new operating system for the data centre, and its Firmament scheduler, which investigates novel approaches to scheduling distributed data processing jobs by integrating cluster and machine-level scheduling. Malte has in the past worked with Broadcom and Google and is an author of a paper on Omega, Google's next-generation cluster scheduler (best student paper at EuroSys 2013). Malte has an MA in Computer Science from Cambridge.

At CamSaS, Malte enjoys kernel and user space infrastructure programming, dealing with challenging distributed systems problems and chasing hard-to-find bugs. He leads the Firmament and DIOS projects, and is also (some days reluctantly) responsible for the CamSaS experimental data centre testbed.




Contributors

Adam Gleave

AdamGleave

Adam Gleave recently graduated from the MPhil in Advanced Computer Science. His speciality is applying theory to solve systems problems. He has previously worked as a software developer for Raspberry Pi and Jane Street.

At CamSaS, Adam is working on incremental min-cost flow solvers for the Firmament cluster scheduler. This work helps make Firmament scale to very large clusters, and also improves the scheduler's response time, which matters for latency-sensitive jobs. Moreover, he also contributes to Firmament's trace-driven cluster simulator. Adam has a background in far more rigorous Maths than Computer Science will ever teach, and accordingly his speciality is to digest complex mathematical problems quickly, identify promising solutions and even the occasional formal proof.



Andrew Scull

AndrewScull

Andrew graduated from the Computer Science Tripos and now works at Google. He has previously worked on various low-level systems projects at RealVNC and MorganStanley.

At CamSaS, Andrew is contributing to the DIOS project. He is building a Rust runtime for DIOS, as part of which he is improving the libd standard library and generally squashing bugs where they arise. Andrew's specialities are comprehending large swathes of C code quickly and knowing insidious details of programming language standards.




Alumni

Gustaf Helgesson

gustafa

Gustaf graduated from the MPhil in Advanced Computer Science in 2013/14 and now works at LinkedIn. His main interests lie in high-performance, large-scale distributed systems and data processing frameworks. Gustaf holds a BSc in Computer Science from the University of Edinburgh, during which he spent a year at Caltech, winning one of two prestigious exchange places. He has previously worked as a site reliability engineer at Google and LinkedIn and co-founded the coding interview start-up Coduru. He has now joined LinkedIn as a software engineer.

At CamSaS, Gustaf worked on the Firmament distributed scheduling project. In particular, he is built Fulgurator, an extension to Firmament that allows the optimal balance between performance SLAs and energy use to be struck in highly heterogeneous data centres composed of low-power ARM servers and high-power x86/IA64 machines.


Matthew Huxtable

@matthewhuxtable    mhuxtable

Matthew completed the Computer Science Tripos and the MRes in Advanced Computer Science between 2011 and 2015. His interests cover broadly cover systems design and engineering, with a special focus daring OS kernel hacks. After completing his degrees, Matthew joined Sparx as a software engineer.

At CamSaS, Matthew built the Unbuckle in-kernel key-value store, described in his Part II dissertation that won the best project prize in 2013/14. Unbuckle showed that a purely software-based approach to building a fast key-value store can attain excellent performance (22+ Gbps on commodity NICs). For his MRes project, Matthew built RascalOS, a prototype rack-scale OS with kernel-integrated Raft consensus for TCP flow allocation, earning him a distinction.



Faculty advisers

Jon Crowcroft

@tforcwork    crowcroft

Jon Crowcroft is the Marconi Professor of Communications Systems in the Computer Laboratory. He has worked in the area of Internet support for multimedia communications for over 30 years. Three main topics of interest have been scalable multicast routing, practical approaches to traffic management, and the design of deployable end-to-end protocols. Current active research areas are Opportunistic Communications, Social Networks, and techniques and algorithms to scale infrastructure-free mobile systems. He leans towards a "build and learn" paradigm for research. Jon graduated in Physics from Trinity College, University of Cambridge in 1979, gained an MSc in Computing in 1981 and PhD in 1993, both from UCL. He is a Fellow the Royal Society, a Fellow of the ACM, a Fellow of the British Computer Society, a Fellow of the IET and the Royal Academy of Engineering and a Fellow of the IEEE. He likes teaching, and has published a few books based on learning materials.

Within CamSaS, Jon reliably contributes wacky ideas, but also a wealth of detail knowledge on network protocols, queueing theory and the history of the internet (whether asked for or not).



Steven Hand

smh22

Steve is a Software Engineer at Google, but was previously a Principal Researcher at Microsoft Research Silicon Valley. Before that even, he was a Reader in Operating Systems at the University of Cambridge Computer Laboratory. He is a systems person: he carries out research in a wide variety of areas, including operating systems, networking and distributed systems. In his PhD, he looked at how to do memory management for a soft real-time operating system (Nemesis). He has also done considerable work in the area of virtualization, based chiefly around the Xen virtual machine monitor developed at Computer Laboratory. This work considered aspects of performance, security and manageability. In parallel, Steve also continued to work on dependability, storage, peer-to-peer systems, networking and concurrency.

At CamSaS, Steve is a source of wisdom and advice, but also a relentless critic with the amazing ability to spot weaknesses in experiments within seconds.



Andrew W. Moore

@awm22    awm22

Andrew W. Moore is Reader in Systems at the Computer Laboratory. He has been measuring and modelling network and operating systems for over 20 years. In recent years Andrew has built tools to measure networks at speed, test networks at capacity, tools to metricate topologies, tools for reproducing end-system behaviours, tools for studying the impact of latency, tools to make-practical the discovery of network-based applications including contributing methodologies for using machine-learning to do fast/accurate identification of network-based applications. Andrew took his first two degrees from Monash University in digital engineering and computer science and a PhD at the Cambridge Computer Laboratory on the measurement-based management of (constrained) network resources. Following the founding of two research labs in Cambridge, and a post at the University of London, Andrew returned to Cambridge to join the Computer Lab in 2007. In his spare time, Andrew leads the NetFPGA project enabling reconfigurable hardware for research and teaching.

Within CamSaS, Andrew acts as cheerleader, cook, and bottle-washer; a big believer in proof by running-system, he contributes an occasional insight from statistics, networking architectures, and practical experimental methodologies.


Robert N.M. Watson

rwatson

Robert is a University Lecturer in Systems, Security, and Architecture in the Security Research Group at the University of Cambridge Computer Laboratory. He works on a number of cross-layer research projects spanning computer architecture, compilers, program analysis, program transformation, operating systems, networking, and security. These look at clean-slate hardware and software designs for security, including processor designs, compilers and software analysis and transformation systems. Previously, he designed the Capsicum hybrid capability system, worked on system-call interposition concurrency problems, and the TrustedBSD MAC Framework. He is also on the board of directors of the FreeBSD Foundation, and has contributed extensively to the FreeBSD Project.

In CamSaS, Robert enjoys pointing out the security elephants in the room when nobody else is particularly concerned about them, but also contributes insights on high-speed NICs, operating system kernel structure and capability systems.