Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference最新文献

Prefix Siphoning: Exploiting LSM-Tree Range Filters For Information Disclosure (Full Version) 前缀虹吸:利用lsm树范围过滤器进行信息披露(完整版)

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2023-06-07 DOI: 10.48550/arXiv.2306.04602

Adi Kafuman, Moshik Hershcovitch, Adam Morrison

{"title":"Prefix Siphoning: Exploiting LSM-Tree Range Filters For Information Disclosure (Full Version)","authors":"Adi Kafuman, Moshik Hershcovitch, Adam Morrison","doi":"10.48550/arXiv.2306.04602","DOIUrl":"https://doi.org/10.48550/arXiv.2306.04602","url":null,"abstract":"Key-value stores typically leave access control to the systems for which they act as storage engines. Unfortunately, attackers may circumvent such read access controls via timing attacks on the key-value store, which use differences in query response times to glean information about stored data. To date, key-value store timing attacks have aimed to disclose stored values and have exploited external mechanisms that can be disabled for protection. In this paper, we point out that key disclosure is also a security threat -- and demonstrate key disclosure timing attacks that exploit mechanisms of the key-value store itself. We target LSM-tree based key-value stores utilizing range filters, which have been recently proposed to optimize LSM-tree range queries. We analyze the impact of the range filters SuRF and prefix Bloom filter on LSM-trees through a security lens, and show that they enable a key disclosure timing attack, which we call prefix siphoning. Prefix siphoning successfully leverages benign queries for non-present keys to identify prefixes of actual keys -- and in some cases, full keys -- in scenarios where brute force searching for keys (via exhaustive enumeration or random guesses) is infeasible.","PeriodicalId":88968,"journal":{"name":"Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference","volume":"72 1","pages":"719-733"},"PeriodicalIF":0.0,"publicationDate":"2023-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85143418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Legion: Automatically Pushing the Envelope of Multi-GPU System for Billion-Scale GNN Training 军团:用于十亿规模GNN训练的多gpu系统自动突破极限

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2023-05-26 DOI: 10.48550/arXiv.2305.16588

Jiecheng Sun, L. Su, Zuocheng Shi, Wen Shen, Zeke Wang, Lei Wang, J. Zhang, Yong Li, Wenyuan Yu, Jingren Zhou, Fei Wu

{"title":"Legion: Automatically Pushing the Envelope of Multi-GPU System for Billion-Scale GNN Training","authors":"Jiecheng Sun, L. Su, Zuocheng Shi, Wen Shen, Zeke Wang, Lei Wang, J. Zhang, Yong Li, Wenyuan Yu, Jingren Zhou, Fei Wu","doi":"10.48550/arXiv.2305.16588","DOIUrl":"https://doi.org/10.48550/arXiv.2305.16588","url":null,"abstract":"Graph neural network(GNN) has been widely applied in real-world applications, such as product recommendation in e-commerce platforms and risk control in financial management systems. Several cache-based GNN systems have been built to accelerate GNN training in a single machine with multiple GPUs. However, these systems fail to train billion-scale graphs efficiently, which is a common challenge in the industry. In this work, we propose Legion, a system that automatically pushes the envelope of multi-GPU systems for accelerating billion-scale GNN training. First, we design a hierarchical graph partitioning mechanism that significantly improves the multi-GPU cache performance. Second, we build a unified multi-GPU cache that helps to minimize the PCIe traffic incurred by caching both graph topology and features with the highest hotness. Third, we develop an automatic caching management mechanism that adapts the multi-GPU cache plan according to the hardware specifications and various graphs to maximize the overall training throughput. Evaluations on various GNN models and multiple datasets show that Legion supports training billion-scale GNNs in a single machine and significantly outperforms the state-of-the-art cache-based systems on small graphs.","PeriodicalId":88968,"journal":{"name":"Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference","volume":"24 1","pages":"165-179"},"PeriodicalIF":0.0,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89086540","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 4

On-demand Container Loading in AWS Lambda AWS Lambda中的按需容器加载

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2023-05-22 DOI: 10.48550/arXiv.2305.13162

Marc Brooker, Mike Danilov, Chris Greenwood, Phil Piwonka

{"title":"On-demand Container Loading in AWS Lambda","authors":"Marc Brooker, Mike Danilov, Chris Greenwood, Phil Piwonka","doi":"10.48550/arXiv.2305.13162","DOIUrl":"https://doi.org/10.48550/arXiv.2305.13162","url":null,"abstract":"AWS Lambda is a serverless event-driven compute service, part of a category of cloud compute offerings sometimes called Function-as-a-service (FaaS). When we first released AWS Lambda, functions were limited to 250MB of code and dependencies, packaged as a simple compressed archive. In 2020, we released support for deploying container images as large as 10GiB as Lambda functions, allowing customers to bring much larger code bases and sets of dependencies to Lambda. Supporting larger packages, while still meeting Lambda's goals of rapid scale (adding up to 15,000 new containers per second for a single customer, and much more in aggregate), high request rate (millions of requests per second), high scale (millions of unique workloads), and low start-up times (as low as 50ms) presented a significant challenge. We describe the storage and caching system we built, optimized for delivering container images on-demand, and our experiences designing, building, and operating it at scale. We focus on challenges around security, efficiency, latency, and cost, and how we addressed these challenges in a system that combines caching, deduplication, convergent encryption, erasure coding, and block-level demand loading. Since building this system, it has reliably processed hundreds of trillions of Lambda invocations for over a million AWS customers, and has shown excellent resilience to load and infrastructure failures.","PeriodicalId":88968,"journal":{"name":"Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference","volume":"58 1","pages":"315-328"},"PeriodicalIF":0.0,"publicationDate":"2023-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77381296","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3

Faith: An Efficient Framework for Transformer Verification on GPUs Faith:一个高效的gpu变压器验证框架

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2022-09-23 DOI: 10.48550/arXiv.2209.12708

Boyuan Feng, Tianqi Tang, Yuke Wang, Zhaodong Chen, Zheng Wang, Shu Yang, Yuan Xie, Yufei Ding

{"title":"Faith: An Efficient Framework for Transformer Verification on GPUs","authors":"Boyuan Feng, Tianqi Tang, Yuke Wang, Zhaodong Chen, Zheng Wang, Shu Yang, Yuan Xie, Yufei Ding","doi":"10.48550/arXiv.2209.12708","DOIUrl":"https://doi.org/10.48550/arXiv.2209.12708","url":null,"abstract":"Transformer verification draws increasing attention in machine learning research and industry. It formally verifies the robustness of transformers against adversarial attacks such as exchanging words in a sentence with synonyms. However, the performance of transformer verification is still not satisfactory due to bound-centric computation which is significantly different from standard neural networks. In this paper, we propose Faith, an efficient framework for transformer verification on GPUs. We first propose a semantic-aware computation graph transformation to identify semantic information such as bound computation in transformer verification. We exploit such semantic information to enable efficient kernel fusion at the computation graph level. Second, we propose a verification-specialized kernel crafter to efficiently map transformer verification to modern GPUs. This crafter exploits a set of GPU hardware supports to accelerate verification specialized operations which are usually memory-intensive. Third, we propose an expert-guided autotuning to incorporate expert knowledge on GPU backends to facilitate large search space exploration. Extensive evaluations show that Faith achieves $2.1times$ to $3.4times$ ($2.6times$ on average) speedup over state-of-the-art frameworks.","PeriodicalId":88968,"journal":{"name":"Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference","volume":"109 1","pages":"167-182"},"PeriodicalIF":0.0,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85338433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Emu: Rapid Prototyping of Networking Services Emu:网络服务的快速原型

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2017-07-12 DOI: 10.17863/CAM.13009

Nik Sultana, Salvator Galea, D. Greaves, M. Wójcik, J. Shipton, R. Clegg, Luo Mai, Pietro Bressana, R. Soulé, R. Mortier, Paolo Costa, P. Pietzuch, J. Crowcroft, A. Moore, Noa Zilberman

{"title":"Emu: Rapid Prototyping of Networking Services","authors":"Nik Sultana, Salvator Galea, D. Greaves, M. Wójcik, J. Shipton, R. Clegg, Luo Mai, Pietro Bressana, R. Soulé, R. Mortier, Paolo Costa, P. Pietzuch, J. Crowcroft, A. Moore, Noa Zilberman","doi":"10.17863/CAM.13009","DOIUrl":"https://doi.org/10.17863/CAM.13009","url":null,"abstract":"Due to their performance and flexibility, FPGAs are an attractive platform for the execution of network functions. It has been a challenge for a long time though to make FPGA programming accessible to a large audience of developers. An appealing solution is to compile code from a general-purpose language to hardware using high-level synthesis. Unfortunately, current approaches to implement rich network functionality are insufficient because they lack: (i) libraries with abstractions for common network operations and data structures, (ii) bindings to the underlying \"substrate\" on the FPGA, and (iii) debugging and profiling support. \u0000 \u0000This paper describes Emu, a new standard library for an FPGA hardware compiler that enables developers to rapidly create and deploy network functionality. Emu allows for high-performance designs without being bound to particular packet processing paradigms. Furthermore, it supports running the same programs on CPUs, in Mininet, and on FPGAs, providing a better development environment that includes advanced debugging capabilities. We demonstrate that network functions implemented using Emu have only negligible resource and performance overheads compared with natively-written hardware versions.","PeriodicalId":88968,"journal":{"name":"Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference","volume":"48 1","pages":"459-471"},"PeriodicalIF":0.0,"publicationDate":"2017-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75918645","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 37

FLICK: Developing and Running Application-Specific Network Services FLICK:开发和运行特定应用程序的网络服务

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2016-06-22 DOI: 10.17863/CAM.389

Abdul Alim, R. Clegg, Luo Mai, Lukas Rupprecht, Eric Seckler, Paolo Costa, P. Pietzuch, A. Wolf, Nik Sultana, J. Crowcroft, Anil Madhavapeddy, A. Moore, R. Mortier, M. Koleini, Luis Oviedo, Matteo Migliavacca, Derek McAuley

{"title":"FLICK: Developing and Running Application-Specific Network Services","authors":"Abdul Alim, R. Clegg, Luo Mai, Lukas Rupprecht, Eric Seckler, Paolo Costa, P. Pietzuch, A. Wolf, Nik Sultana, J. Crowcroft, Anil Madhavapeddy, A. Moore, R. Mortier, M. Koleini, Luis Oviedo, Matteo Migliavacca, Derek McAuley","doi":"10.17863/CAM.389","DOIUrl":"https://doi.org/10.17863/CAM.389","url":null,"abstract":"Data centre networks are increasingly programmable, with application-specific network services proliferating, from custom load-balancers to middleboxes providing caching and aggregation. Developers must currently implement these services using traditional low-level APIs, which neither support natural operations on application data nor provide efficient performance isolation. \u0000 \u0000We describe FLICK, a framework for the programming and execution of application-specific network services on multi-core CPUs. Developers write network services in the FLICK language, which offers high-level processing constructs and application-relevant data types. FLICK programs are translated automatically to efficient, parallel task graphs, implemented in C++ on top of a user-space TCP stack. Task graphs have bounded resource usage at runtime, which means that the graphs of multiple services can execute concurrently without interference using cooperative scheduling. We evaluate FLICK with several services (an HTTP load-balancer, a Memcached router and a Hadoop data aggregator), showing that it achieves good performance while reducing development effort.","PeriodicalId":88968,"journal":{"name":"Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference","volume":"61 1","pages":"1-14"},"PeriodicalIF":0.0,"publicationDate":"2016-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76977820","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 15

Making State Explicit for Imperative Big Data Processing 为命令式大数据处理明确状态

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2014-06-19 DOI: 10.17863/CAM.41706

R. Fernandez, Matteo Migliavacca, Evangelia Kalyvianaki, P. Pietzuch

{"title":"Making State Explicit for Imperative Big Data Processing","authors":"R. Fernandez, Matteo Migliavacca, Evangelia Kalyvianaki, P. Pietzuch","doi":"10.17863/CAM.41706","DOIUrl":"https://doi.org/10.17863/CAM.41706","url":null,"abstract":"Data scientists often implement machine learning algorithms in imperative languages such as Java, Matlab and R. Yet such implementations fail to achieve the performance and scalability of specialised data-parallel processing frameworks. Our goal is to execute imperative Java programs in a data-parallel fashion with high throughput and low latency. This raises two challenges: how to support the arbitrary mutable state of Java programs without compromising scalability, and how to recover that state after failure with low overhead. \u0000 \u0000Our idea is to infer the dataflow and the types of state accesses from a Java program and use this information to generate a stateful dataflow graph (SDG). By explicitly separating data from mutable state, SDGs have specific features to enable this translation: to ensure scalability, distributed state can be partitioned across nodes if computation can occur entirely in parallel; if this is not possible, partial state gives nodes local instances for independent computation, which are reconciled according to application semantics. For fault tolerance, large inmemory state is checkpointed asynchronously without global coordination. We show that the performance of SDGs for several imperative online applications matches that of existing data-parallel processing frameworks.","PeriodicalId":88968,"journal":{"name":"Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference","volume":"141 1","pages":"49-60"},"PeriodicalIF":0.0,"publicationDate":"2014-06-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75931848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 60

Application-Defined Decentralized Access Control. 应用程序定义的分散访问控制。

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2014-01-01

Yuanzhong Xu, Alan M Dunn, Owen S Hofmann, Michael Z Lee, Syed Akbar Mehdi, Emmett Witchel

引用次数: 0

Body Armor for Binaries: Preventing Buffer Overflows Without Recompilation 二进制文件的防弹衣:防止缓冲区溢出而不重新编译

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2012-06-13 DOI: 10.5555/2342821.2342832

Asia Slowinska, T. Stancescu, H. Bos

引用次数: 57

Free Factories: Unified Infrastructure for Data Intensive Web Services. 免费工厂：数据密集型Web服务的统一基础设施。

Proceedings of the USENIX ... annual Technical Conference. USENIX Technical Conference Pub Date : 2008-05-01

Alexander Wait Zaranek, Tom Clegg, Ward Vandewege, George M Church

引用次数: 0