GoJournal is implemented in Go, and Perennial is implemented in the Coq proof assistant. A graph embedding is a fixed length vector representation for each node (and/or edge-type) in a graph and has emerged as the de-facto approach to apply modern machine learning on graphs. Moreover, to handle dynamic workloads, Nap adopts a fast NAL switch mechanism. Consensus bugs are bugs that make Ethereum clients transition to incorrect blockchain states and fail to reach consensus with other clients. Tej Chajed, MIT CSAIL; Joseph Tassarotti, Boston College; Mark Theng, MIT CSAIL; Ralf Jung, MPI-SWS; M. Frans Kaashoek and Nickolai Zeldovich, MIT CSAIL. NrOS is primarily constructed as a simple, sequential kernel with no concurrency, making it easier to develop and reason about its correctness. Upon these two primitives, our system can scale to thousands of concurrent enclaves with high resource utilization and eliminate the high-cost initialization of secure memory using fork-style enclave creation without weakening the security guarantees. This approach misses possible optimization opportunities as transformations that only preserve equivalence on subsets of the output tensors are excluded. SanRazor adopts a novel hybrid approach it captures both dynamic code coverage and static data dependencies of checks, and uses the extracted information to perform a redundant check analysis. Metadata from voice calls, such as the knowledge of who is communicating with whom, contains rich information about peoples lives. One important reason for the high cost is, as we observe in this paper, that many sanitizer checks are redundant the same safety property is repeatedly checked leading to unnecessarily wasted computing resources. His work has included the Barrelfish multikernel research OS, as well as work on distributed stream processors, and using formal specifications to describe the hardware/software interfaces of modern computer systems. She is the recipient of several best paper awards, the Einstein Chair of the Chinese Academy of Science, the ACM/SIGART Autonomous Agents Research Award, an NSF Career Award, and the Allen Newell Medal for Excellence in Research. The wire-to-wire RPC response time through the nanoPU is just 69ns, an order of magnitude quicker than the best-of-breed, low latency, commercial NICs. In this paper, we present Vegito, a distributed in-memory HTAP system that embraces freshness and performance with the following three techniques: (1) a lightweight gossip-style scheme to apply logs on backups consistently; (2) a block-based design for multi-version columnar backups; (3) a two-phase concurrent updating mechanism for the tree-based index of backups. The overhead of GPT is 5% for memory-intensive workloads (e.g., Redis) and negligible for CPU-intensive workloads (e.g., RV8 and Coremarks). PC members are not required to read supplementary material when reviewing the paper, so each paper should stand alone without it. These results outperform state-of-the-art HTAP systems by several orders of magnitude on transactional performance, while just incurring little performance slowdown (5% over pure OLTP workloads) and still enjoying data freshness for analytical queries (less than 20 ms of maximum delay) in the failure-free case. Papers accompanied by nondisclosure agreement forms will not be considered. Based on the observation that invariants are often concise in practice, DistAI starts with small invariant formulas and enumerates all strongest possible invariants that hold for all samples. Poor data locality hurts an application's performance. Researchers from the Software Systems Laboratory bagged Best Paper Awards at the 15th USENIX Symposium on Operating Systems Design and Implementation (OSDI 2021) and the 2021 USENIX Annual Technical Conference (USENIX ATC 2021).. Jay Lepreau Best Paper Award, OSDI'21. Petuum Awarded OSDI 2021 Best Paper for Goodput-Optimized Deep Learning Research Petuum CASL research and engineering team's Pollux technical paper on adaptive scheduling for optimized. SC is being increasingly adopted by industry for a variety of applications. The co-chairs may then share that paper with the workshops organizers and discuss it with them. Erhu Feng, Xu Lu, Dong Du, Bicheng Yang, and Xueqiang Jiang, Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University; Engineering Research Center for Domain-specific Operating Systems, Ministry of Education, China; Yubin Xia, Binyu Zang, and Haibo Chen, Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University; Shanghai AI Laboratory; Engineering Research Center for Domain-specific Operating Systems, Ministry of Education, China. This paper presents the design and implementation of CLP, a tool capable of losslessly compressing unstructured text logs while enabling fast searches directly on the compressed data. Title Page, Copyright Page, and List of Organizers | Owing to the sequential write-only zone scheme of the ZNS, the log-structured file system (LFS) is required to access ZNS solid-state drives (SSDs). These scripts often make pages slow to load, partly due to a fundamental inefficiency in how browsers process JavaScript content: browsers make it easy for web developers to reason about page state by serially executing all scripts on any frame in a page, but as a result, fail to leverage the multiple CPU cores that are readily available even on low-end phones. We present the nanoPU, a new NIC-CPU co-design to accelerate an increasingly pervasive class of datacenter applications: those that utilize many small Remote Procedure Calls (RPCs) with very short (s-scale) processing times. If the conference registration fee will pose a hardship for the presenter of the accepted paper, please contact conference@usenix.org. She is the author of the textbook Interconnections (about network layers 2 and 3) and coauthor of Network Security. After three years working on web-based collaboration systems at a startup in North Carolina, he joined Sprint's Advanced Technology Lab in Burlingame, California, in 1998, working on cloud computing and network monitoring. Distributed systems are notoriously hard to implement correctly due to non-determinism. Authors may submit a response to those reviews until Friday, March 5, 2021. JEL codes: Q18, Q28, Q57 . Papers so short as to be considered extended abstracts will not receive full consideration. Using this property, MAGE calculates the memory access pattern ahead of time and uses it to produce a memory management plan. A.H. Hunter, Jane Street Capital; Chris Kennelly, Paul Turner, Darryl Gove, Tipp Moseley, and Parthasarathy Ranganathan, Google. Our evaluation shows that PET outperforms existing systems by up to 2.5, by unlocking previously missed opportunities from partially equivalent transformations. Submission of a response is optional. Pollux improves scheduling performance in deep learning (DL) clusters by adaptively co-optimizing inter-dependent factors both at the per-job level and at the cluster-wide level. This motivates the need for a new approach to data privacy that can provide strong assurance and control to users. The file system performance of the proposed ZNS+ storage system was 1.33--2.91 times better than that of the normal ZNS-based storage system. See the Preview Session page for an overview of the topics covered in the program. Our evaluation on the SPEC benchmarks shows that SanRazor can reduce the overhead of sanitizers significantly, from 73.8% to 28.062.0% for AddressSanitizer, and from 160.1% to 36.6124.4% for UndefinedBehaviorSanitizer (depending on the applied reduction scheme). This is especially true for DPF over Rnyi DP, a highly composable form of DP. Abstract registrations that do not provide sufficient information to understand the topic and contribution (e.g., empty abstracts, placeholder abstracts, or trivial abstracts) will be rejected, thereby precluding paper submission. Simultaneous submission of the same work to multiple venues, submission of previously published work, or plagiarism constitutes dishonesty or fraud. We have implemented a prototype of our design based on Penglai, an open-sourced enclave system for RISC-V. Weak Links in Authentication Chains: A Large-scale Analysis of Email Sender Spoofing Attacks For conference information, . PET then automatically corrects results to restore full equivalence. We present case studies and end-to-end applications that show how Storm lets developers specify diverse policies while centralizing the trusted code to under 1% of the application, and statically enforces security with modest type annotation overhead, and no run-time cost. Hence, kernel developers are constantly refining synchronization within OS kernels to improve scalability at the risk of introducing subtle bugs. OSDI will provide an opportunity for authors to respond to reviews prior to final consideration of the papers at the program committee meeting. However, the existing one-size-fits-all GNN implementations are insufficient to catch up with the evolving GNN architectures, the ever-increasing graph size, and the diverse node embedding dimensionality. To remedy this, we introduce DeSearch, the first decentralized search engine that guarantees the integrity and privacy of search results for decentralized services and blockchain apps. We compare Marius against two state-of-the-art industrial systems on a diverse array of benchmarks. Papers must be in PDF format and must be submitted via the submission form. Reviews will be available for response on Wednesday, March 3, 2021. Starting with small invariant formulas and strongest possible invariants avoids large SMT queries, improving SMT solver performance. This paper demonstrates that it is possible to achieve s-scale latency using Linux kernel storage stack, even when tens of latency-sensitive applications compete for host resources with throughput-bound applications that perform read/write operations at throughput close to hardware capacity. A significant obstacle to using SC for practical applications is the memory overhead of the underlying cryptography. Table of Contents | Indeed, it is a prime target for powerful adversaries such as nation states. Manuela will present examples and discuss the scope of AI in her research in the finance domain. This fast path contains programmable hardware support for low latency transport and congestion control as well as hardware support for efficient load balancing of RPCs to cores. First, it enables a caller to push a message to a callee in two hops, using a new way of assigning mailboxes to users that resembles how a post office assigns PO boxes to its customers. Swapnil Gandhi and Anand Padmanabha Iyer, Microsoft Research. Mothy joined the Computer Science Department ETH Zurich in January 2007 and was named Fellow of the ACM in 2013 for contributions to operating systems and networking research. Jiachen Wang, Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University; Shanghai AI Laboratory; Engineering Research Center for Domain-specific Operating Systems, Ministry of Education, China; Ding Ding, Department of Computer Science, New York University; Huan Wang, Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University; Shanghai AI Laboratory; Engineering Research Center for Domain-specific Operating Systems, Ministry of Education, China; Conrad Christensen, Department of Computer Science, New York University; Zhaoguo Wang and Haibo Chen, Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University; Shanghai AI Laboratory; Engineering Research Center for Domain-specific Operating Systems, Ministry of Education, China; Jinyang Li, Department of Computer Science, New York University. We present application studies for 8 applications, improving requests-per-second (RPS) by 7.7% and reducing RAM usage 2.4%. We also propose two file system techniques for ZNS+-aware LFS. Fortunately, we observe that the backups for high availability in modern distributed OLTP systems can be retrofitted to bridge the analytical queries and transactions in HTAP workloads. Concretely, Dorylus is 1.22 faster and 4.83 cheaper than GPU servers for massive sparse graphs. The paper review process is double-blind. The copyback-aware block allocation considers different copy costs at different copy paths within the SSD. As a member of ACCT, I have served two years on the bylaws and governance committee and two years on the finance and audit committee. Devices employ adaptive interrupt coalescing heuristics that try to balance between these opposing goals. The OSDI '21 program co-chairs have agreed not to submit their work to OSDI '21. HotNets provides a venue for discussing innovative ideas and for debating future research agendas in networking. Her robot soccer teams have been RoboCup world champions several times, and the CoBot mobile robots have autonomously navigated for more than 1,000km in university buildings. Prior or concurrent workshop publication does not preclude publishing a related paper in OSDI. Her specialties include network routing protocols and network security. Tao Luo, Mingen Pan, Pierre Tholoniat, Asaf Cidon, and Roxana Geambasu, Columbia University; Mathias Lcuyer, Microsoft Research. blk-switch evaluation over a variety of scenarios shows that it consistently achieves s-scale average and tail latency (at both 99th and 99.9th percentiles), while allowing applications to near-perfectly utilize the hardware capacity. Paper abstracts and proceedings front matter are available to everyone now. For example, traditional compute resources are replenishable while privacy is not: a CPU can be regained after a model finishes execution while privacy budget cannot. Password Questions? DeSearch uses trusted hardware to build a network of workers that execute a pipeline of small search engine tasks (crawl, index, aggregate, rank, query). Graph Neural Networks (GNNs) have gained significant attention in the recent past, and become one of the fastest growing subareas in deep learning. Authors should email the program co-chairs, osdi21chairs@usenix.org, a copy of the related workshop paper and a short explanation of the new material in the conference paper beyond that published in the workshop version. Therefore, developers typically find data locality issues via dynamic profiling and repair them manually. See the USENIX Conference Submissions Policy for details. To help more profitably utilize sanitizers, we introduce SanRazor, a practical tool aiming to effectively detect and remove redundant sanitizer checks. The abstractions we design for the privacy resource mirror those defined by Kubernetes for traditional resources, but there are also major differences. While compiler-based techniques have been proposed to improve data locality, they depend on heuristics, which can sometimes hurt performance. We built a functional NFSv3 server, called GoNFS, to use GoJournal. Submitted papers must be no longer than 12 single-spaced 8.5 x 11 pages, including figures and tables, plus as many pages as needed for references, using 10-point type on 12-point (single-spaced) leading, two-column format, Times Roman or a similar font, within a text block 7 wide x 9 deep. Jiang Zhang, University of Southern California; Shuai Wang, HKUST; Manuel Rigger, Pinjia He, and Zhendong Su, ETH Zurich. Advisor: You have a past or present association as thesis advisor or advisee. 64 papers accepted out of 341 submitted. Submissions violating the detailed formatting and anonymization rules will not be considered for review. The 15th USENIX Symposium on Operating Systems Design and Implementation (OSDI '21) will take place as a virtual event on July 14-16, 2021. Calibrated interrupts increase throughput by up to 35%, reduce CPU consumption by as much as 30%, and achieve up to 37% lower latency when interrupts are coalesced. We first introduce two new hardware primitives: 1) Guarded Page Table (GPT), which protects page table pages to support page-level secure memory isolation; 2) Mountable Merkle Tree (MMT), which supports scalable integrity protection for secure memory. For realistic workloads, KEVIN improves throughput by 68% on average. Session Chairs: Moshe Gabel, University of Toronto, and Joseph Gonzalez, University of California, Berkeley, John Thorpe, Yifan Qiao, Jonathan Eyolfson, and Shen Teng, UCLA; Guanzhou Hu, UCLA and University of Wisconsin, Madison; Zhihao Jia, CMU; Jinliang Wei, Google Brain; Keval Vora, Simon Fraser; Ravi Netravali, Princeton University; Miryung Kim and Guoqing Harry Xu, UCLA. Timothy Roscoe is a Full Professor in the Systems Group of the Computer Science Department at ETH Zurich, where he works on operating systems, networks, and distributed systems, and is currently head of department. Camera-ready submission (all accepted papers): 2 April 2021; Main conference program: 27-28 April 2021; All deadline times are . He joined Intel Research at Berkeley in April 2002 as a principal architect of PlanetLab, an open, shared platform for developing and deploying planetary-scale services. The NAL eliminates remote PM accesses to hot items without inducing extra local PM accesses. DeSearch then introduces a witness mechanism to make sure the completed tasks can be reused across different pipelines, and to make the final search results verifiable by end users. The experimental results show that Penglai can support 1,000s enclave instances running concurrently and scale up to 512GB secure memory with both encryption and integrity protection. Sijie Shen, Rong Chen, Haibo Chen, and Binyu Zang, Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University; Shanghai Artificial Intelligence Laboratory; Engineering Research Center for Domain-specific Operating Systems, Ministry of Education, China. In particular, I'll argue for re-engaging with what computer hardware really is today and give two suggestions (among many) about how the OS research community can usefully do this, and exploit what is actually a tremendous opportunity. A scientific paper consists of a constellation of artifacts that extend beyond the document itself: software, hardware, evaluation data and documentation, raw survey results, mechanized proofs, models, test suites, benchmarks, and so on. The 15th USENIX Symposium on Operating Systems Design and Implementation seeks to present innovative, exciting research in computer systems. The main contribution of this paper is GoJournal, a verified, concurrent journaling system that provides atomicity for storage applications, together with Perennial 2.0, a framework for formally specifying and verifying concurrent crash-safe systems. Here, we focus on hugepage coverage. Professor Veloso is the Past President of AAAI (the Association for the Advancement of Artificial Intelligence), and the co-founder, Trustee, and Past President of RoboCup. USENIX new Date().getFullYear()>document.write(new Date().getFullYear()); Grants for Black Computer Science Students Application, Title Page, Copyright Page, and List of Organizers, OSDI '21 Proceedings Interior (PDF, best for mobile devices). While verifying GoJournal, we found one serious concurrency bug, even though GoJournal has many unit tests. J.P. Morgan AI Research partners with applied data analytics teams across the firm as well as with leading academic institutions globally. For general conference information, see https://www . Second, it innovates on the underlying cryptographic machinery and constructs a new private information retrieval scheme, FastPIR, that reduces the time to process oblivious access requests for mailboxes. Typically, monolithic kernels share state across cores and rely on one-off synchronization patterns that are specialized for each kernel structure or subsystem. 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. Kyuhwa Han, Sungkyunkwan University and Samsung Electronics; Hyunho Gwak and Dongkun Shin, Sungkyunkwan University; Jooyoung Hwang, Samsung Electronics. This change is receiving considerable attention in the architecture and security communities, for example, but in contrast, so-called OS researchers are mostly in denial. If your paper is accepted and you need an invitation letter to apply for a visa to attend the conference, please contact conference@usenix.org as soon as possible. With the help of thousands of Lambda threads, Dorylus scales GNN training to billion-edge graphs. We also verified a simple NFS server using GoJournals specs, which confirms that they are helpful for application verification: a significant part of the proof doesnt have to consider concurrency and crashes. OSDI takes a broad view of the systems area and solicits contributions from many fields of systems practice, including, but not limited to, operating systems, file and storage systems, distributed systems, cloud computing, mobile systems, secure and reliable systems, systems aspects of big data, embedded systems, virtualization, networking as it relates to operating systems, and management and troubleshooting of complex systems. Although the number of submissions is lower than the past, it's likely only due to the late announcement; being in my first OSDI PC, I think the quality of the submitted and accepted papers remains as high as ever. It then feeds those invariants and the desired safety properties to an SMT solver to check if the conjunction of the invariants and the safety properties is inductive. All papers will be available online to registered attendees before the conference. Youngseok Yang, Seoul National University; Taesoo Kim, Georgia Institute of Technology; Byung-Gon Chun, Seoul National University and FriendliAI. We implement DeSearch for two existing decentralized services that handle over 80 million records and 240 GBs of data, and show that DeSearch can scale horizontally with the number of workers and can process 128 million search queries per day. All the times listed below are in Pacific Daylight Time (PDT). Copyright to the individual works is retained by the author[s]. Sat, Aug 7, 2021 3 min read researches review. Of the 26 submitted artifacts: 26 artifacts received the Artifacts Available badge (100%). Sam Kumar, David E. Culler, and Raluca Ada Popa, University of California, Berkeley. Differential privacy (DP) enables model training with a guaranteed bound on this leakage. OSDI 2021 papers summary. Furthermore, such performance can be achieved without any modification in applications, network hardware, kernel CPU schedulers and/or kernel network stack. Federated Learning (FL) is an emerging direction in distributed machine learning (ML) that enables in-situ model training and testing on edge data. We implement a variant of a log-structured merge tree in the storage device that not only indexes file objects, but also supports transactions and manages physical storage space. Welcome to the 2021 USENIX Annual Technical Conference (ATC '21) submissions site! VLDB 2021: Venue Tivoli Hotel & Congress Center Arni Magnussons Gade 2 1577 Copenhagen, Denmark +45 3268 4300 In-person attendees can purchase tickets for the park / gardens with a 15% discount, which is a special offer by Tivoli Hotel & Congress Center to VLDB 2021 attendees. USENIX new Date().getFullYear()>document.write(new Date().getFullYear()); Grants for Black Computer Science Students Application, Propose an interesting, compelling solution, Demonstrate the practicality and benefits of the solution, Clearly describe the paper's contributions, Clearly articulate the advances beyond previous work. Notification of conditional accept/reject for revisions: 3 March 2022. Welcome to the SOSP 2021 Website. Evaluation on a four-node machine with Optane DC Persistent Memory shows that Nap can improve the throughput by up to 2.3 and 1.56 under write-intensive and read-intensive workloads, respectively. Session Chairs: Nadav Amit, VMware Research Group, and Ada Gavrilovska, Georgia Institute of Technology, Stephen Ibanez, Alex Mallery, Serhat Arslan, and Theo Jepsen, Stanford University; Muhammad Shahbaz, Purdue University; Changhoon Kim and Nick McKeown, Stanford University. Lukas Burkhalter, Nicolas Kchler, Alexander Viand, Hossein Shafagh, and Anwar Hithnawi, ETH Zrich. Our evaluation shows that, compared to existing participant selection mechanisms, Oort improves time-to-accuracy performance by 1.2X-14.1X and final model accuracy by 1.3%-9.8%, while efficiently enforcing developer-specified model testing criteria at the scale of millions of clients. Our approach effectively eliminates high communication and partitioning overheads, and couples it with a new pipelined push-pull parallelism based execution strategy for fast model training. Our approach outperforms existing file systems on a block SSD by a wide margin 6.2 on average for metadata-intensive benchmarks. Professor Veloso is on leave from Carnegie Mellon University as the Herbert A. Simon University Professor in the School of Computer Science, and the past Head of the Machine Learning Department. Thanks to selective profiling, DMons profiling overhead is 1.36% on average, making it feasible for production use. Session Chairs: Deniz Altinbken, Google, and Rashmi Vinayak, Carnegie Mellon University, Tanvir Ahmed Khan and Ian Neal, University of Michigan; Gilles Pokam, Intel Corporation; Barzan Mozafari and Baris Kasikci, University of Michigan. When further combined with a simple caching strategy, our evaluation shows that P3 is able to outperform existing state-of-the-art distributed GNN frameworks by up to 7. Storm ensures security using a Security Typed ORM that refines the (type) abstractions of each layer of the MVC API with logical assertions that describe the data produced and consumed by the underlying operation and the users allowed access to that data. As has been standard practice in OSDI and SOSP in recent years, we will allow authors to submit quick responses to PC reviews: they will be made available to the PC before the final online discussion and PC meeting. Main conference program: 5-8 April 2022. Session Chairs: Sebastian Angel, University of Pennsylvania, and Malte Schwarzkopf, Brown University, Ishtiyaque Ahmad, Yuntian Yang, Divyakant Agrawal, Amr El Abbadi, and Trinabh Gupta, University of California Santa Barbara. The key insight in blk-switch is that Linux's multi-queue storage design, along with multi-queue network and storage hardware, makes the storage stack conceptually similar to a network switch. Lifting predicates and crash framing make the specification easy to use for developers, and logically atomic crash specifications allow for modular reasoning in GoJournal, making the proof tractable despite complex concurrency and crash interleavings. Fluffy found two new consensus bugs in the most popular Geth Ethereum client which were exploitable on the live Ethereum mainnet. We propose Marius, a system for efficient training of graph embeddings that leverages partition caching and buffer-aware data orderings to minimize disk access and interleaves data movement with computation to maximize utilization. We also welcome work that explores the interface to related areas such as computer architecture, networking, programming languages, analytics, and databases. Prepublication versions of the accepted papers from the summer submission deadline are available below. Mothy's current research centers on Enzian, a powerful hybrid CPU/FPGA machine designed for research into systems software. This yielded 6% fewer TLB miss stalls, and 26% reduction in memory wasted due to fragmentation. The key insight guiding our design is computation separation. Furthermore, to enable automatic runtime optimization, GNNAdvisor incorporates a lightweight analytical model for an effective design parameter search. Yet, existing efforts randomly select FL participants, which leads to poor model and system efficiency. Authors are also encouraged to contact the program co-chairs, osdi21chairs@usenix.org, if needed to relate their OSDI submissions to relevant submissions of their own that are simultaneously under review or awaiting publication at other venues. A glance at this year's OSDI program shows that Operating Systems are a small niche topic for this conference, not even meriting their own full session. We develop rigorous theoretical foundations to simplify equivalence examination and correction for partially equivalent transformations, and design an efficient search algorithm to quickly discover highly optimized programs by combining fully and partially equivalent optimizations at the tensor, operator, and graph levels. This year, there were only 2 accepted papers from UK institutes. Session Chairs: Ryan Huang, Johns Hopkins University, and Manos Kapritsos, University of Michigan, Jianan Yao, Runzhou Tao, Ronghui Gu, Jason Nieh, Suman Jana, and Gabriel Ryan, Columbia University. Academic and industrial participants present research and experience papers that cover the full range of theory and practice of computer . To achieve low overhead, selective profiling gathers runtime execution information selectively and incrementally. OSDI '21 Technical Sessions All the times listed below are in Pacific Daylight Time (PDT). For example, talks may be shorter than in prior years, or some parts of the conference may be multi-tracked. In 2023 I started another two-year term on the . Radia Perlman is a Fellow at Dell Technologies. USENIX Security '21 has three submission deadlines. Accepted papers will be allowed 14 pages in the proceedings, plus references. Pollux simultaneously considers both aspects. The hybrid segment recycling chooses a proper block reclaiming policy between segment compaction and threaded logging based on their costs. Using selective profiling, we build DMon, a system that can automatically locate data locality problems in production, identify access patterns that hurt locality, and repair such patterns using targeted optimizations.