Cornell Systems Lunch

Large-scale in-memory object caches such as memcached are widely used to accelerate popular web sites and to reduce burden on backend databases. Yet operators still have limited visibility into how these caches should be set up to optimally accommodate the workloads at hand. How much would the cache performance improve from additional cache space, or by adding more cache servers to the pool? Resources come at a cost, so to what extent would user request latency deteriorate if cache memory is repurposed for a different service?

In this talk, I will focus on several research questions pertaining to large-scale distributed caches. In particular, I will home in on the challenge of providing online monitoring of the cost and benefits of memory space in a large-scale cache, enabling cache operators to answer these questions without requiring trace collection and manual offline tuning. In our work, we introduce general and efficient algorithms for dynamically estimating hit rate curves -- histograms of cache hit rate as a function of memory size -- which can be plugged into cache replacement policies such as LRU.

Extensive simulations on cache benchmarks indicate that our methods provide accurate estimates of hit rate at different cache sizes. Our experiments on an implementation of our methods in memcached showed that hit rate curves were dynamically estimated at over 98% accuracy with only a small drop in throughput. The results are encouraging and suggest that exposing hit rate curves can be a practical method for improving provisioning and metering of large-scale caches.

This is joint work with Trausti Saemundsson at CloudPhysics and Gregory Chockler at the University of London, Royal Holloway.

Bio: Ymir Vigfusson is Assistant Professor of Mathematics and Computer Science at Emory University since 2014, Assistant Professor at the School of Computer Science at Reykjavik University since 2011, and a co-founder and Chief Science Officer of the offensive security company Syndis since 2013. Ymir completed his Ph.D. in Computer Science at Cornell University (2010) with focus on Distributed Systems and minor in music composition, where his dissertation was nominated for the ACM Doctoral Dissertation Award. His primary research interests are on distributed data replication, data science and security. His work is partially supported by an NSF CAREER award, the CDC and grants from the Icelandic Center for Research, and his website is http://www.ymsir.com/

Emerging energy-harvesting devices (EHDs) are computer systems that operate using energy extracted from their environment, even from low-power sources like ambient radio-frequency energy. Future EHDs will be a key enabler of emerging implantable medical devices, IoT applications, and nano-satellites, but todays EHDs operate intermittently, only as environmental energy is available. Unfortunately, intermittence makes todays EHDs unreliable and extremely difficult to program and debug. In this talk I will summarize the main challenges of intermittent execution. I will then discuss our recent efforts developing system, programming language, and toolchain support for EHDs to address the challenges of intermittence, focusing especially on programmability, debugging, and reliability. I will close by discussing our recent work on building a reliable, EHD-based, hardware/software application platform for an upcoming deployment.

Bio: Brandon Lucia is an assistant professor in the department of electrical and computer engineering at Carnegie Mellon University in Pittsburgh, PA. His research focuses on redefining computer architectures and systems that make increasingly pervasive, often safety-critical, devices reliable, energy-efficient, and programmable. Brandon and his lab are currently focusing on defining the system stack for systems with intermittently available energy and resources, as well as on redefining parallel architectures to improve their efficiency, correctness, and reliability, exploiting heterogeneity and approximation. His work targets the boundaries between computer architecture, compilers, system software, and programming languages. His research group is supported by the National Science Foundation, Google, Intel, and Disney Research. Brandon has received a Google Faculty Research Award, the Bell Labs Prize, an OOPSLA Distinguished Paper Award and an OOPSLA Distinguished Artifact Award, and Brandon was elected in 2016 to serve on DARPAs Information Science And Technology (ISAT) study group. Before joining CMU, Brandon spent a year as a Researcher at Microsoft Research, Redmond. Brandon earned a Ph.D. in Computer Science and Engineering from the University of Washington in 2013, and a B.S. degree in Computer Science from Tufts University. His personal website is http://brandonlucia.com, his research group is at http://abstract.ece.cmu.edu, and his band netcat is at http://netcat.co