arXiv subjects

Ravi Netravali

Publications and source records attributed to Ravi Netravali.

32 records · Page 2Linked to original sources

arXiv2024

Apparate: Rethinking Early Exits to Tame Latency-Throughput Tensions in ML Serving

Machine learning (ML) inference platforms are tasked with balancing two competing goals: ensuring high throughput given many requests, and delivering low-latency responses to support interactive applications. Unfortunately, existing platform knobs (e.g., batch sizes) fail to ease this fundamental tension, and instead only enable users to harshly trade off one property for the other. This paper explores an alternate strategy to taming throughput-latency tradeoffs by changing the granularity at which inference is performed. We present Apparate, a system that automatically applies and manages early exits (EEs) in ML models, whereby certain inputs can exit with results at intermediate layers. To cope with the time-varying overhead and accuracy challenges that EEs bring, Apparate repurposes exits to provide continual feedback that powers several novel runtime monitoring and adaptation strategies. Apparate lowers median response latencies by 40.5--91.5% and 10.0--24.2% for diverse CV and NLP classification workloads, and median time-per-token latencies by 22.6--77.9% for generative scenarios, without affecting throughputs or violating tight accuracy constraints.

cs.DC↗

arXiv2023

MadEye: Boosting Live Video Analytics Accuracy with Adaptive Camera Configurations

Camera orientations (i.e., rotation and zoom) govern the content that a camera captures in a given scene, which in turn heavily influences the accuracy of live video analytics pipelines. However, existing analytics approaches leave this crucial adaptation knob untouched, instead opting to only alter the way that captured images from fixed orientations are encoded, streamed, and analyzed. We present MadEye, a camera-server system that automatically and continually adapts orientations to maximize accuracy for the workload and resource constraints at hand. To realize this using commodity pan-tilt-zoom (PTZ) cameras, MadEye embeds (1) a search algorithm that rapidly explores the massive space of orientations to identify a fruitful subset at each time, and (2) a novel knowledge distillation strategy to efficiently (with only camera resources) select the ones that maximize workload accuracy. Experiments on diverse workloads show that MadEye boosts accuracy by 2.9-25.7% for the same resource usage, or achieves the same accuracy with 2-3.7x lower resource costs.

cs.DC↗

arXiv2022

Canvas: Isolated and Adaptive Swapping for Multi-Applications on Remote Memory

Remote memory techniques for datacenter applications have recently gained a great deal of popularity. Existing remote memory techniques focus on the efficiency of a single application setting only. However, when multiple applications co-run on a remote-memory system, significant interference could occur, resulting in unexpected slowdowns even if the same amounts of physical resources are granted to each application. This slowdown stems from massive sharing in applications' swap data paths. Canvas is a redesigned swap system that fully isolates swap paths for remote-memory applications. Canvas allows each application to possess its dedicated swap partition, swap cache, prefetcher, and RDMA bandwidth. Swap isolation lays a foundation for adaptive optimization techniques based on each application's own access patterns and needs. We develop three such techniques: (1) adaptive swap entry allocation, (2) semantics-aware prefetching, and (3) two-dimensional RDMA scheduling. A thorough evaluation with a set of widely-deployed applications demonstrates that Canvas minimizes performance variation and dramatically reduces performance degradation.

cs.OS↗

arXiv2022

Marvolo: Programmatic Data Augmentation for Practical ML-Driven Malware Detection

Data augmentation has been rare in the cyber security domain due to technical difficulties in altering data in a manner that is semantically consistent with the original data. This shortfall is particularly onerous given the unique difficulty of acquiring benign and malicious training data that runs into copyright restrictions, and that institutions like banks and governments receive targeted malware that will never exist in large quantities. We present MARVOLO, a binary mutator that programmatically grows malware (and benign) datasets in a manner that boosts the accuracy of ML-driven malware detectors. MARVOLO employs semantics-preserving code transformations that mimic the alterations that malware authors and defensive benign developers routinely make in practice , allowing us to generate meaningful augmented data. Crucially, semantics-preserving transformations also enable MARVOLO to safely propagate labels from original to newly-generated data samples without mandating expensive reverse engineering of binaries. Further, MARVOLO embeds several key optimizations that keep costs low for practitioners by maximizing the density of diverse data samples generated within a given time (or resource) budget. Experiments using wide-ranging commercial malware datasets and a recent ML-driven malware detector show that MARVOLO boosts accuracies by up to 5%, while operating on only a small fraction (15%) of the potential input binaries.

cs.CR↗

arXiv2022

GEMEL: Model Merging for Memory-Efficient, Real-Time Video Analytics at the Edge

Video analytics pipelines have steadily shifted to edge deployments to reduce bandwidth overheads and privacy violations, but in doing so, face an ever-growing resource tension. Most notably, edge-box GPUs lack the memory needed to concurrently house the growing number of (increasingly complex) models for real-time inference. Unfortunately, existing solutions that rely on time/space sharing of GPU resources are insufficient as the required swapping delays result in unacceptable frame drops and accuracy violations. We present model merging, a new memory management technique that exploits architectural similarities between edge vision models by judiciously sharing their layers (including weights) to reduce workload memory costs and swapping delays. Our system, GEMEL, efficiently integrates merging into existing pipelines by (1) leveraging several guiding observations about per-model memory usage and inter-layer dependencies to quickly identify fruitful and accuracy-preserving merging configurations, and (2) altering edge inference schedules to maximize merging benefits. Experiments across diverse workloads reveal that GEMEL reduces memory usage by up to 60.7%, and improves overall accuracy by 8-39% relative to time/space sharing alone.

cs.DC↗

arXiv2022

Boggart: Towards General-Purpose Acceleration of Retrospective Video Analytics

Commercial retrospective video analytics platforms have increasingly adopted general interfaces to support the custom queries and convolutional neural networks (CNNs) that different applications require. However, existing optimizations were designed for settings where CNNs were platform- (not user-) determined, and fail to meet at least one of the following key platform goals when that condition is violated: reliable accuracy, low latency, and minimal wasted work. We present Boggart, a system that simultaneously meets all three goals while supporting the generality that today's platforms seek. Prior to queries being issued, Boggart carefully employs traditional computer vision algorithms to generate indices that are imprecise, but are fundamentally comprehensive across different CNNs/queries. For each issued query, Boggart employs new techniques to quickly characterize the imprecision of its index, and sparingly run CNNs (and propagate the results to other frames) in a way that bounds accuracy drops. Our results highlight that Boggart's improved generality comes at low cost, with speedups that match (and most often, exceed) prior, model-specific approaches.

cs.CV↗

arXiv2022

Dashlet: Taming Swipe Uncertainty for Robust Short Video Streaming

Short video streaming applications have recently gained substantial traction, but the non-linear video presentation they afford swiping users fundamentally changes the problem of maximizing user quality of experience in the face of the vagaries of network throughput and user swipe timing. This paper describes the design and implementation of Dashlet, a system tailored for high quality of experience in short video streaming applications. With the insights we glean from an in-the-wild TikTok performance study and a user study focused on swipe patterns, Dashlet proposes a novel out-of-order video chunk pre-buffering mechanism that leverages a simple, non machine learning-based model of users' swipe statistics to determine the pre-buffering order and bitrate. The net result is a system that achieves 77-99% of an oracle system's QoE and outperforms TikTok by 43.9-45.1x, while also reducing by 30% the number of bytes wasted on downloaded video that is never watched.

cs.NI↗

arXiv2022

Bamboo: Making Preemptible Instances Resilient for Affordable Training of Large DNNs

DNN models across many domains continue to grow in size, resulting in high resource requirements for effective training, and unpalatable (and often unaffordable) costs for organizations and research labs across scales. This paper aims to significantly reduce training costs with effective use of preemptible instances, i.e., those that can be obtained at a much cheaper price while idle, but may be preempted whenever requested by priority users. Doing so, however, requires new forms of resiliency and efficiency to cope with the possibility of frequent preemptions - a failure model that is drastically different from the occasional failures in normal cluster settings that existing checkpointing techniques target. We present Bamboo, a distributed system that tackles these challenges by introducing redundant computations into the training pipeline, i.e., whereby one node performs computations over not only its own layers but also over some layers in its neighbor. Our key insight is that training large models often requires pipeline parallelism where "pipeline bubbles" naturally exist. Bamboo carefully fills redundant computations into these bubbles, providing resilience at a low cost. Across a variety of widely used DNN models, Bamboo outperforms traditional checkpointing by 3.7x in training throughput, and reduces costs by 2.4x compared to a setting where on-demand instances are used.

cs.DC↗

arXiv2021

Revelio: ML-Generated Debugging Queries for Distributed Systems

A major difficulty in debugging distributed systems lies in manually determining which of the many available debugging tools to use and how to query its logs. Our own study of a production debugging workflow confirms the magnitude of this burden. This paper explores whether a machine-learning model can assist developers in distributed systems debugging. We present Revelio, a debugging assistant which takes user reports and system logs as input, and outputs debugging queries that developers can use to find a bug's root cause. The key challenges lie in (1) combining inputs of different types (e.g., natural language reports and quantitative logs) and (2) generalizing to unseen faults. Revelio addresses these by employing deep neural networks to uniformly embed diverse input sources and potential queries into a high-dimensional vector space. In addition, it exploits observations from production systems to factorize query generation into two computationally and statistically simpler learning tasks. To evaluate Revelio, we built a testbed with multiple distributed applications and debugging tools. By injecting faults and training on logs and reports from 800 Mechanical Turkers, we show that Revelio includes the most helpful query in its predicted list of top-3 relevant queries 96% of the time. Our developer study confirms the utility of Revelio.

cs.DC↗

arXiv2021

Privid: Practical, Privacy-Preserving Video Analytics Queries

Analytics on video recorded by cameras in public areas have the potential to fuel many exciting applications, but also pose the risk of intruding on individuals' privacy. Unfortunately, existing solutions fail to practically resolve this tension between utility and privacy, relying on perfect detection of all private information in each video frame--an elusive requirement. This paper presents: (1) a new notion of differential privacy (DP) for video analytics, $(ρ,K,ε)$-event-duration privacy, which protects all private information visible for less than a particular duration, rather than relying on perfect detections of that information, and (2) a practical system called Privid that enforces duration-based privacy even with the (untrusted) analyst-provided deep neural networks that are commonplace for video analytics today. Across a variety of videos and queries, we show that Privid achieves accuracies within 79-99% of a non-private system.

cs.CR↗

arXiv2021

Dorylus: Affordable, Scalable, and Accurate GNN Training with Distributed CPU Servers and Serverless Threads

A graph neural network (GNN) enables deep learning on structured graph data. There are two major GNN training obstacles: 1) it relies on high-end servers with many GPUs which are expensive to purchase and maintain, and 2) limited memory on GPUs cannot scale to today's billion-edge graphs. This paper presents Dorylus: a distributed system for training GNNs. Uniquely, Dorylus can take advantage of serverless computing to increase scalability at a low cost. The key insight guiding our design is computation separation. Computation separation makes it possible to construct a deep, bounded-asynchronous pipeline where graph and tensor parallel tasks can fully overlap, effectively hiding the network latency incurred by Lambdas. With the help of thousands of Lambda threads, Dorylus scales GNN training to billion-edge graphs. Currently, for large graphs, CPU servers offer the best performance-per-dollar over GPU servers. Just using Lambdas on top of CPU servers offers up to 2.75x more performance-per-dollar than training only with CPU servers. Concretely, Dorylus is 1.22x faster and 4.83x cheaper than GPU servers for massive sparse graphs. Dorylus is up to 3.8x faster and 10.7x cheaper compared to existing sampling-based systems.

cs.DC↗

arXiv2021

ABC: A Simple Explicit Congestion Controller for Wireless Networks

We propose Accel-Brake Control (ABC), a simple and deployable explicit congestion control protocol for network paths with time-varying wireless links. ABC routers mark each packet with an "accelerate" or "brake", which causes senders to slightly increase or decrease their congestion windows. Routers use this feedback to quickly guide senders towards a desired target rate. ABC requires no changes to header formats or user devices, but achieves better performance than XCP. ABC is also incrementally deployable; it operates correctly when the bottleneck is a non-ABC router, and can coexist with non-ABC traffic sharing the same bottleneck link. We evaluate ABC using a Wi-Fi implementation and trace-driven emulation of cellular links. ABC achieves 30-40% higher throughput than Cubic+Codel for similar delays, and 2.2X lower delays than BBR on a Wi-Fi path. On cellular network paths, ABC achieves 50% higher throughput than Cubic+Codel.

cs.NI↗

arXiv2020

Continuous Prefetch for Interactive Data Applications

Interactive data visualization and exploration (DVE) applications are often network-bottlenecked due to bursty request patterns, large response sizes, and heterogeneous deployments over a range of networks and devices. This makes it difficult to ensure consistently low response times (< 100ms). Khameleon is a framework for DVE applications that uses a novel combination of prefetching and response tuning to dynamically trade-off response quality for low latency. Khameleon exploits DVE's approximation tolerance: immediate lower-quality responses are preferable to waiting for complete results. To this end, Khameleon progressively encodes responses, and runs a server-side scheduler that proactively streams portions of responses using available bandwidth to maximize user's perceived interactivity. The scheduler involves a complex optimization based on available resources, predicted user interactions, and response quality levels; yet, decisions must also be real-time. To overcome this, Khameleon uses a fast greedy approximation which closely mimics the optimal approach. Using image exploration and visualization applications with real user interaction traces, we show that across a wide range of network and client resource conditions, Khameleon outperforms classic prefetching approaches that benefit from perfect prediction models: response latencies with Khameleon are never higher, and typically between 2 to 3 orders of magnitude lower while response quality remains within 50%-80%.

cs.DB↗

arXiv2018

Secure Incentivization for Decentralized Content Delivery

Prior research has proposed technical solutions to use peer-to-peer (P2P) content delivery to serve Internet video, showing that it can reduce costs to content providers. Yet, such methods have not become widespread except for a few niche instances. An important challenge is incentivization: what tangible benefits does P2P content delivery offer users who bring resources to the table? In this paper, we ask whether monetary incentives can help attract peers in P2P content delivery systems. We commissioned a professional survey of people around theUnited States to answer several relevant questions. We found that 51% of the 876 respondents--substantially larger than our expectations--answered "yes" to whether they would participate for suitable financial incentives. Encouraged by the results of the survey, we propose Gringotts, a system to structure incentives and securely incorporate P2P delivery into content delivery systems. Gringotts provides a novel Proof of Delivery mechanism that allows content providers to verify correct delivery of their files, and shows how to use cryptocurrency to pay peers while guarding against liars and Sybil attacks.

cs.NI↗