Pub Date : 2023-03-06DOI: 10.48550/arXiv.2303.02868
Xiaonan Nie, Yi Liu, Fangcheng Fu, J. Xue, Dian Jiao, Xupeng Miao, Yangyu Tao, Bin Cui
Recent years have witnessed the unprecedented achievements of large-scale pre-trained models, especially Transformer models. Many products and services in Tencent Inc., such as WeChat, QQ, and Tencent Advertisement, have been opted in to gain the power of pre-trained models. In this work, we present Angel-PTM, a productive deep learning system designed for pre-training and fine-tuning Transformer models. Angel-PTM can train extremely large-scale models with hierarchical memory efficiently. The key designs of Angel-PTM are a fine-grained memory management via the Page abstraction and a unified scheduling method that coordinates computations, data movements, and communications. Furthermore, Angel-PTM supports extreme model scaling with SSD storage and implements a lock-free updating mechanism to address the SSD I/O bottlenecks. Experimental results demonstrate that Angel-PTM outperforms existing systems by up to 114.8% in terms of maximum model scale as well as up to 88.9% in terms of training throughput. Additionally, experiments on GPT3-175B and T5-MoE-1.2T models utilizing hundreds of GPUs verify our strong scalability.
{"title":"Angel-PTM: A Scalable and Economical Large-scale Pre-training System in Tencent","authors":"Xiaonan Nie, Yi Liu, Fangcheng Fu, J. Xue, Dian Jiao, Xupeng Miao, Yangyu Tao, Bin Cui","doi":"10.48550/arXiv.2303.02868","DOIUrl":"https://doi.org/10.48550/arXiv.2303.02868","url":null,"abstract":"\u0000 Recent years have witnessed the unprecedented achievements of large-scale pre-trained models, especially Transformer models. Many products and services in Tencent Inc., such as WeChat, QQ, and Tencent Advertisement, have been opted in to gain the power of pre-trained models. In this work, we present Angel-PTM, a productive deep learning system designed for pre-training and fine-tuning Transformer models. Angel-PTM can train extremely large-scale models with hierarchical memory efficiently. The key designs of Angel-PTM are a fine-grained memory management via the\u0000 Page\u0000 abstraction and a unified scheduling method that coordinates computations, data movements, and communications. Furthermore, Angel-PTM supports extreme model scaling with SSD storage and implements a lock-free updating mechanism to address the SSD I/O bottlenecks. Experimental results demonstrate that Angel-PTM outperforms existing systems by up to 114.8% in terms of maximum model scale as well as up to 88.9% in terms of training throughput. Additionally, experiments on GPT3-175B and T5-MoE-1.2T models utilizing hundreds of GPUs verify our strong scalability.\u0000","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"57 1","pages":"3781-3794"},"PeriodicalIF":0.0,"publicationDate":"2023-03-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77215011","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}
Pub Date : 2023-03-01DOI: 10.14778/3587136.3587147
Kasper Overgaard Mortensen, Fatemeh Zardbani, M. A. Haque, S. Agustsson, D. Mottin, Philip Hofmann, Panagiotis Karras
How can we efficiently and scalably cluster high-dimensional data? The k -means algorithm clusters data by iteratively reducing intra-cluster Euclidean distances until convergence. While it finds applications from recommendation engines to image segmentation, its application to high-dimensional data is hindered by the need to repeatedly compute Euclidean distances among points and centroids. In this paper, we propose Marigold ( k -means for high-dimensional data), a scalable algorithm for k -means clustering in high dimensions. Marigold prunes distance calculations by means of (i) a tight distance-bounding scheme; (ii) a stepwise calculation over a multiresolution transform; and (iii) exploiting the triangle inequality. To our knowledge, such an arsenal of pruning techniques has not been hitherto applied to k -means. Our work is motivated by time-critical Angle-Resolved Photoemission Spectroscopy (ARPES) experiments, where it is vital to detect clusters among high-dimensional spectra in real time. In a thorough experimental study with real-world data sets we demonstrate that Marigold efficiently clusters high-dimensional data, achieving approximately one order of magnitude improvement over prior art.
如何高效、可扩展地聚类高维数据?k均值算法通过迭代地减少聚类内的欧氏距离来聚类数据，直到收敛。虽然它从推荐引擎到图像分割都有应用，但由于需要反复计算点和质心之间的欧几里德距离，它在高维数据中的应用受到了阻碍。本文提出了一种可扩展的高维k均值聚类算法Marigold (k -means for high-dimensional data)。万寿菊李子距离的计算(i)紧距离边界格式;(ii)对一个多分辨率变换进行逐步计算;(3)利用三角不等式。据我们所知，迄今为止，这种修剪技术的武库尚未应用于k -means。我们的工作是由时间临界角分辨光谱学(ARPES)实验激发的，在该实验中，实时检测高维光谱中的簇是至关重要的。在对真实世界数据集的彻底实验研究中，我们证明了Marigold有效地聚类高维数据，比现有技术实现了大约一个数量级的改进。
{"title":"Marigold: Efficient k-means Clustering in High Dimensions","authors":"Kasper Overgaard Mortensen, Fatemeh Zardbani, M. A. Haque, S. Agustsson, D. Mottin, Philip Hofmann, Panagiotis Karras","doi":"10.14778/3587136.3587147","DOIUrl":"https://doi.org/10.14778/3587136.3587147","url":null,"abstract":"\u0000 How can we efficiently and scalably cluster high-dimensional data? The\u0000 k\u0000 -means algorithm clusters data by iteratively reducing intra-cluster Euclidean distances until convergence. While it finds applications from recommendation engines to image segmentation, its application to high-dimensional data is hindered by the need to repeatedly compute Euclidean distances among points and centroids. In this paper, we propose Marigold (\u0000 k\u0000 -means for high-dimensional data), a scalable algorithm for\u0000 k\u0000 -means clustering in high dimensions. Marigold prunes distance calculations by means of (i) a tight distance-bounding scheme; (ii) a stepwise calculation over a multiresolution transform; and (iii) exploiting the triangle inequality. To our knowledge, such an arsenal of pruning techniques has not been hitherto applied to\u0000 k\u0000 -means. Our work is motivated by time-critical Angle-Resolved Photoemission Spectroscopy (ARPES) experiments, where it is vital to detect clusters among high-dimensional spectra in real time. In a thorough experimental study with real-world data sets we demonstrate that Marigold efficiently clusters high-dimensional data, achieving approximately one order of magnitude improvement over prior art.\u0000","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"10 1","pages":"1740-1748"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84599009","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}
Pub Date : 2023-03-01DOI: 10.14778/3587136.3587143
Zicong Hong, Song Guo, Enyuan Zhou, Wuhui Chen, Huawei Huang, Albert Y. Zomaya
Blockchain databases have attracted widespread attention but suffer from poor scalability due to underlying non-scalable blockchains. While blockchain sharding is necessary for a scalable blockchain database, it poses a new challenge named on-chain cross-shard database services. Each cross-shard database service (e.g., cross-shard queries or inter-shard load balancing) involves massive cross-shard data exchanges, while the existing cross-shard mechanisms need to process each cross-shard data exchange via the consensus of all nodes in the related shards (i.e., on-chain) to resist a Byzantine environment of blockchain, which eliminates sharding benefits. To tackle the challenge, this paper presents GriDB, the first scalable blockchain database, by designing a novel off-chain cross-shard mechanism for efficient cross-shard database services. Borrowing the idea of off-chain payments, GriDB delegates massive cross-shard data exchange to a few nodes, each of which is randomly picked from a different shard. Considering the Byzantine environment, the untrusted delegates cooperate to generate succinct proof for cross-shard data exchanges, while the consensus is only responsible for the low-cost proof verification. However, different from payments, the database services' verification has more requirements (e.g., completeness, correctness, freshness, and availability); thus, we introduce several new authenticated data structures (ADS). Particularly, we utilize consensus to extend the threat model and reduce the complexity of traditional accumulator-based ADS for verifiable cross-shard queries with a rich set of relational operators. Moreover, we study the necessity of inter-shard load balancing for a scalable blockchain database and design an off-chain and live approach for both efficiency and availability during balancing. An evaluation of our prototype shows the performance of GriDB in terms of scalability in workloads with queries and updates.
{"title":"GriDB: Scaling Blockchain Database via Sharding and Off-Chain Cross-Shard Mechanism","authors":"Zicong Hong, Song Guo, Enyuan Zhou, Wuhui Chen, Huawei Huang, Albert Y. Zomaya","doi":"10.14778/3587136.3587143","DOIUrl":"https://doi.org/10.14778/3587136.3587143","url":null,"abstract":"\u0000 Blockchain databases have attracted widespread attention but suffer from poor scalability due to underlying non-scalable blockchains. While blockchain sharding is necessary for a scalable blockchain database, it poses a new challenge named\u0000 on-chain cross-shard database services.\u0000 Each cross-shard database service (e.g., cross-shard queries or inter-shard load balancing) involves massive cross-shard data exchanges, while the existing cross-shard mechanisms need to process each cross-shard data exchange via the consensus of all nodes in the related shards (i.e., on-chain) to resist a Byzantine environment of blockchain, which eliminates sharding benefits.\u0000 \u0000 \u0000 To tackle the challenge, this paper presents GriDB, the first scalable blockchain database, by designing a novel\u0000 off-chain cross-shard mechanism\u0000 for efficient cross-shard database services. Borrowing the idea of off-chain payments, GriDB delegates massive cross-shard data exchange to a few nodes, each of which is randomly picked from a different shard. Considering the Byzantine environment, the untrusted delegates cooperate to generate succinct proof for cross-shard data exchanges, while the consensus is only responsible for the low-cost proof verification. However, different from payments, the database services' verification has more requirements (e.g., completeness, correctness, freshness, and availability); thus, we introduce several new\u0000 authenticated data structures\u0000 (ADS). Particularly, we utilize consensus to extend the threat model and reduce the complexity of traditional accumulator-based ADS for verifiable cross-shard queries with a rich set of relational operators. Moreover, we study the necessity of inter-shard load balancing for a scalable blockchain database and design an off-chain and live approach for both efficiency and availability during balancing. An evaluation of our prototype shows the performance of GriDB in terms of scalability in workloads with queries and updates.\u0000","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"37 1","pages":"1685-1698"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80589588","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}
Pub Date : 2023-03-01DOI: 10.14778/3587136.3587150
Tianyi Chen, Jun Gao, Hedui Chen, Yaofeng Tu
Query optimization based on deep reinforcement learning (DRL) has become a hot research topic recently. Despite the achieved promising progress, DRL optimizers still face great challenges of robustly producing efficient plans, due to the vast search space for both join order and operator selection and the highly varying execution latency taken as the feedback signal. In this paper, we propose LOGER, a learned optimizer towards generating efficient and robust plans, aiming at producing both efficient join orders and operators. LOGER first utilizes Graph Transformer to capture relationships between tables and predicates. Then, the search space is reorganized, in which LOGER learns to restrict specific operators instead of directly selecting one for each join, while utilizing DBMS built-in optimizer to select physical operators under the restrictions. Such a strategy exploits expert knowledge to improve the robustness of plan generation while offering sufficient plan search flexibility. Furthermore, LOGER introduces ε -beam search, which keeps multiple search paths that preserve promising plans while performing guided exploration. Finally, LOGER introduces a loss function with reward weighting to further enhance performance robustness by reducing the fluctuation caused by poor operators, and log transformation to compress the range of rewards. We conduct experiments on Join Order Benchmark (JOB), TPC-DS and Stack Overflow, and demonstrate that LOGER can achieve a performance better than existing learned query optimizers, with a 2.07x speedup on JOB compared with PostgreSQL.
基于深度强化学习(DRL)的查询优化是近年来的研究热点。尽管取得了可喜的进展，但由于连接顺序和算子选择的巨大搜索空间以及作为反馈信号的高度变化的执行延迟，DRL优化器仍然面临鲁棒生成高效计划的巨大挑战。在本文中，我们提出了LOGER，一个用于生成高效鲁棒计划的学习优化器，旨在生成高效的连接顺序和操作符。logger首先利用Graph Transformer捕获表和谓词之间的关系。然后，对搜索空间进行重组，其中LOGER学习限制特定的操作符，而不是为每个连接直接选择一个操作符，同时利用DBMS内置的优化器在限制下选择物理操作符。该策略利用专家知识提高了计划生成的鲁棒性，同时提供了足够的计划搜索灵活性。此外，LOGER引入了ε波束搜索，在进行引导勘探的同时保留多条搜索路径，以保留有希望的计划。最后，LOGER引入了一个带有奖励权重的损失函数，通过减少糟糕算子带来的波动进一步增强性能的鲁棒性，并通过对数变换压缩奖励的范围。我们在Join Order Benchmark (JOB)、TPC-DS和Stack Overflow上进行了实验，并证明LOGER可以实现比现有学习查询优化器更好的性能，与PostgreSQL相比，LOGER在JOB上的加速速度提高了2.07倍。
{"title":"LOGER: A Learned Optimizer towards Generating Efficient and Robust Query Execution Plans","authors":"Tianyi Chen, Jun Gao, Hedui Chen, Yaofeng Tu","doi":"10.14778/3587136.3587150","DOIUrl":"https://doi.org/10.14778/3587136.3587150","url":null,"abstract":"\u0000 Query optimization based on deep reinforcement learning (DRL) has become a hot research topic recently. Despite the achieved promising progress, DRL optimizers still face great challenges of robustly producing efficient plans, due to the vast search space for both join order and operator selection and the highly varying execution latency taken as the feedback signal. In this paper, we propose LOGER, a learned optimizer towards generating efficient and robust plans, aiming at producing both efficient join orders and operators. LOGER first utilizes Graph Transformer to capture relationships between tables and predicates. Then, the search space is reorganized, in which LOGER learns to restrict specific operators instead of directly selecting one for each join, while utilizing DBMS built-in optimizer to select physical operators under the restrictions. Such a strategy exploits expert knowledge to improve the robustness of plan generation while offering sufficient plan search flexibility. Furthermore, LOGER introduces\u0000 ε\u0000 -beam search, which keeps multiple search paths that preserve promising plans while performing guided exploration. Finally, LOGER introduces a loss function with reward weighting to further enhance performance robustness by reducing the fluctuation caused by poor operators, and log transformation to compress the range of rewards. We conduct experiments on Join Order Benchmark (JOB), TPC-DS and Stack Overflow, and demonstrate that LOGER can achieve a performance better than existing learned query optimizers, with a 2.07x speedup on JOB compared with PostgreSQL.\u0000","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"15 1","pages":"1777-1789"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74593850","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}
Pub Date : 2023-03-01DOI: 10.48550/arXiv.2303.15702
Peng Fang, Arijit Khan, Siqiang Luo, Fang Wang, Dan Feng, Zhenli Li, Wei Yin, Yu Cao
Graph embedding maps graph nodes to low-dimensional vectors, and is widely adopted in machine learning tasks. The increasing availability of billion-edge graphs underscores the importance of learning efficient and effective embeddings on large graphs, such as link prediction on Twitter with over one billion edges. Most existing graph embedding methods fall short of reaching high data scalability. In this paper, we present a general-purpose, distributed, information-centric random walk-based graph embedding framework, DistGER, which can scale to embed billion-edge graphs. DistGER incrementally computes information-centric random walks. It further leverages a multi-proximity-aware, streaming, parallel graph partitioning strategy, simultaneously achieving high local partition quality and excellent workload balancing across machines. DistGER also improves the distributed Skip-Gram learning model to generate node embeddings by optimizing the access locality, CPU throughput, and synchronization efficiency. Experiments on real-world graphs demonstrate that compared to state-of-the-art distributed graph embedding frameworks, including KnightKing, DistDGL, and Pytorch-BigGraph, DistGER exhibits 2.33×--129× acceleration, 45% reduction in cross-machines communication, and >10% effectiveness improvement in downstream tasks.
{"title":"Distributed Graph Embedding with Information-Oriented Random Walks","authors":"Peng Fang, Arijit Khan, Siqiang Luo, Fang Wang, Dan Feng, Zhenli Li, Wei Yin, Yu Cao","doi":"10.48550/arXiv.2303.15702","DOIUrl":"https://doi.org/10.48550/arXiv.2303.15702","url":null,"abstract":"Graph embedding maps graph nodes to low-dimensional vectors, and is widely adopted in machine learning tasks. The increasing availability of billion-edge graphs underscores the importance of learning efficient and effective embeddings on large graphs, such as link prediction on Twitter with over one billion edges. Most existing graph embedding methods fall short of reaching high data scalability. In this paper, we present a general-purpose, distributed, information-centric random walk-based graph embedding framework, DistGER, which can scale to embed billion-edge graphs. DistGER incrementally computes information-centric random walks. It further leverages a multi-proximity-aware, streaming, parallel graph partitioning strategy, simultaneously achieving high local partition quality and excellent workload balancing across machines. DistGER also improves the distributed Skip-Gram learning model to generate node embeddings by optimizing the access locality, CPU throughput, and synchronization efficiency. Experiments on real-world graphs demonstrate that compared to state-of-the-art distributed graph embedding frameworks, including KnightKing, DistDGL, and Pytorch-BigGraph, DistGER exhibits 2.33×--129× acceleration, 45% reduction in cross-machines communication, and >10% effectiveness improvement in downstream tasks.","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"17 1","pages":"1643-1656"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90471653","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}
Pub Date : 2023-03-01DOI: 10.14778/3587136.3587149
Ruiyuan Li, Zheng Li, Yi Wu, Chao Chen, Yu Zheng
There are a prohibitively large number of floating-point time series data generated at an unprecedentedly high rate. An efficient, compact and lossless compression for time series data is of great importance for a wide range of scenarios. Most existing lossless floating-point compression methods are based on the XOR operation, but they do not fully exploit the trailing zeros, which usually results in an unsatisfactory compression ratio. This paper proposes an Erasing-based Lossless Floating-point compression algorithm, i.e., Elf. The main idea of Elf is to erase the last few bits (i.e., set them to zero) of floating-point values, so the XORed values are supposed to contain many trailing zeros. The challenges of the erasing-based method are three-fold. First, how to quickly determine the erased bits? Second, how to losslessly recover the original data from the erased ones? Third, how to compactly encode the erased data? Through rigorous mathematical analysis, Elf can directly determine the erased bits and restore the original values without losing any precision. To further improve the compression ratio, we propose a novel encoding strategy for the XORed values with many trailing zeros. Elf works in a streaming fashion. It takes only O ( N ) (where N is the length of a time series) in time and O (1) in space, and achieves a notable compression ratio with a theoretical guarantee. Extensive experiments using 22 datasets show the powerful performance of Elf compared with 9 advanced competitors.
{"title":"Elf: Erasing-based Lossless Floating-Point Compression","authors":"Ruiyuan Li, Zheng Li, Yi Wu, Chao Chen, Yu Zheng","doi":"10.14778/3587136.3587149","DOIUrl":"https://doi.org/10.14778/3587136.3587149","url":null,"abstract":"\u0000 There are a prohibitively large number of floating-point time series data generated at an unprecedentedly high rate. An efficient, compact and lossless compression for time series data is of great importance for a wide range of scenarios. Most existing lossless floating-point compression methods are based on the XOR operation, but they do not fully exploit the trailing zeros, which usually results in an unsatisfactory compression ratio. This paper proposes an Erasing-based Lossless Floating-point compression algorithm, i.e.,\u0000 Elf.\u0000 The main idea of\u0000 Elf\u0000 is to erase the last few bits (i.e., set them to zero) of floating-point values, so the XORed values are supposed to contain many trailing zeros. The challenges of the erasing-based method are three-fold. First, how to quickly determine the erased bits? Second, how to losslessly recover the original data from the erased ones? Third, how to compactly encode the erased data? Through rigorous mathematical analysis,\u0000 Elf\u0000 can directly determine the erased bits and restore the original values without losing any precision. To further improve the compression ratio, we propose a novel encoding strategy for the XORed values with many trailing zeros.\u0000 Elf\u0000 works in a streaming fashion. It takes only\u0000 O\u0000 (\u0000 N\u0000 ) (where\u0000 N\u0000 is the length of a time series) in time and\u0000 O\u0000 (1) in space, and achieves a notable compression ratio with a theoretical guarantee. Extensive experiments using 22 datasets show the powerful performance of\u0000 Elf\u0000 compared with 9 advanced competitors.\u0000","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"37 1","pages":"1763-1776"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79259995","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}
Pub Date : 2023-03-01DOI: 10.14778/3587136.3587145
Haochen He, Erci Xu, Shanshan Li, Zhouyang Jia, Si Zheng, Yue Yu, Jun Ma, Xiangke Liao
NVMe SSD hugely boosts the I/O speed, with up to GB/s throughput and microsecond-level latency. Unfortunately, DBMS users can often find their high-performanced storage devices tend to deliver less-than-expected or even worse performance when compared to their traditional peers. While many works focus on proposing new DBMS designs to fully exploit NVMe SSDs, few systematically study the symptoms, root causes and possible detection methods of such performance mismatches on existing databases. In this paper, we start with an empirical study where we systematically expose and analyze the performance mismatches on six popular databases via controlled configuration tuning. From the study, we find that all six databases can suffer from performance mismatches. Moreover, we conclude that the root causes can be categorized as databases' unawareness of new storage devices characteristics in I/O size, I/O parallelism and I/O sequentiality. We report 17 mismatches to developers and 15 are confirmed. Additionally, we realize testing all configuration knobs yields low efficiency. Therefore, we propose a fast performance mismatch detection framework and evaluation shows that our framework brings two orders of magnitude speedup than baseline without sacrificing effectiveness.
{"title":"When Database Meets New Storage Devices: Understanding and Exposing Performance Mismatches via Configurations","authors":"Haochen He, Erci Xu, Shanshan Li, Zhouyang Jia, Si Zheng, Yue Yu, Jun Ma, Xiangke Liao","doi":"10.14778/3587136.3587145","DOIUrl":"https://doi.org/10.14778/3587136.3587145","url":null,"abstract":"NVMe SSD hugely boosts the I/O speed, with up to GB/s throughput and microsecond-level latency. Unfortunately, DBMS users can often find their high-performanced storage devices tend to deliver less-than-expected or even worse performance when compared to their traditional peers. While many works focus on proposing new DBMS designs to fully exploit NVMe SSDs, few systematically study the symptoms, root causes and possible detection methods of such performance mismatches on existing databases.\u0000 In this paper, we start with an empirical study where we systematically expose and analyze the performance mismatches on six popular databases via controlled configuration tuning. From the study, we find that all six databases can suffer from performance mismatches. Moreover, we conclude that the root causes can be categorized as databases' unawareness of new storage devices characteristics in I/O size, I/O parallelism and I/O sequentiality. We report 17 mismatches to developers and 15 are confirmed.\u0000 Additionally, we realize testing all configuration knobs yields low efficiency. Therefore, we propose a fast performance mismatch detection framework and evaluation shows that our framework brings two orders of magnitude speedup than baseline without sacrificing effectiveness.","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"49 1","pages":"1712-1725"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75915307","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}
Pub Date : 2023-03-01DOI: 10.14778/3587136.3587142
Xiang Li, Nuozhou Sun, Yunqian Luo, M. Gao
Cloud systems are now a prevalent platform to host large-scale big-data analytics applications such as machine learning and relational database. However, data privacy remains as a critical concern for public cloud systems. Existing trusted hardware could provide an isolated execution domain on an untrusted platform, but also suffers from access-pattern-based side channels at various levels including memory, disks, and networking. Oblivious algorithms can address these vulnerabilities by hiding the program data access patterns. Unfortunately, current oblivious algorithms for data analytics are limited to single-machine execution, only support simple operations, and/or suffer from significant performance overheads due to the use of expensive global sort and excessive data padding. In this work, we propose SODA, a set of efficient and oblivious algorithms for distributed data analytics operators, including filter, aggregate, and binary equi-join. To improve performance, SODA completely avoids the expensive oblivious global sort primitive, and minimizes the data padding overheads. SODA makes use of low-cost (pseudo-)random communication instead of expensive global sort to ensure uniform data traffic in oblivious filter and aggregate. It also adopts a novel two-level bin-packing approach in oblivious join to alleviate both input redistribution and join product skewness, thus minimizing necessary data padding. Compared to the state-of-the-art system, SODA not only extends the functionality but also improves the performance. It achieves 1.1× to 14.6× speedups on complex multi-operator data analytics workloads.
{"title":"SODA: A Set of Fast Oblivious Algorithms in Distributed Secure Data Analytics","authors":"Xiang Li, Nuozhou Sun, Yunqian Luo, M. Gao","doi":"10.14778/3587136.3587142","DOIUrl":"https://doi.org/10.14778/3587136.3587142","url":null,"abstract":"Cloud systems are now a prevalent platform to host large-scale big-data analytics applications such as machine learning and relational database. However, data privacy remains as a critical concern for public cloud systems. Existing trusted hardware could provide an isolated execution domain on an untrusted platform, but also suffers from access-pattern-based side channels at various levels including memory, disks, and networking. Oblivious algorithms can address these vulnerabilities by hiding the program data access patterns. Unfortunately, current oblivious algorithms for data analytics are limited to single-machine execution, only support simple operations, and/or suffer from significant performance overheads due to the use of expensive global sort and excessive data padding.\u0000 In this work, we propose SODA, a set of efficient and oblivious algorithms for distributed data analytics operators, including filter, aggregate, and binary equi-join. To improve performance, SODA completely avoids the expensive oblivious global sort primitive, and minimizes the data padding overheads. SODA makes use of low-cost (pseudo-)random communication instead of expensive global sort to ensure uniform data traffic in oblivious filter and aggregate. It also adopts a novel two-level bin-packing approach in oblivious join to alleviate both input redistribution and join product skewness, thus minimizing necessary data padding. Compared to the state-of-the-art system, SODA not only extends the functionality but also improves the performance. It achieves 1.1× to 14.6× speedups on complex multi-operator data analytics workloads.","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"31 1","pages":"1671-1684"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85758398","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}
Many relational data in our daily life are represented as graphs, making graph application an important workload. Because of the large scale of graph datasets, moving graph data to the cloud becomes a popular option. To keep the confidential and private graph secure from an untrusted cloud server, many cryptographic techniques are leveraged to hide the content of the data. However, protecting only the data content is not enough for a graph database. Because the structural information of the graph can be revealed through the database accessing track. In this work, we study the graph neural network (GNN), an important graph workload to mine information from a graph database. We find that the server is able to infer which node is processing during the edge retrieving phase and also learn its neighbor indices during GNN's aggregation phase. This leads to the leakage of the information of graph structure data. In this work, we present SPG, a structure-private graph database with SqueezePIR. Our SPG is built on top of Private Information Retrieval (PIR), which securely hides which nodes/neighbors are accessed. In addition, we propose SqueezePIR, a compression technique to overcome the computation overhead of PIR. Based on our evaluation, our SqueezePIR achieves 11.85× speedup on average with less than 2% accuracy loss when compared to the state-of-the-art FastPIR protocol.
{"title":"SPG: Structure-Private Graph Database via SqueezePIR","authors":"Ling Liang, Jilan Lin, Zheng Qu, Ishtiyaque Ahmad, Fengbin Tu, Trinabh Gupta, Yufei Ding, Yuan Xie","doi":"10.14778/3587136.3587138","DOIUrl":"https://doi.org/10.14778/3587136.3587138","url":null,"abstract":"Many relational data in our daily life are represented as graphs, making graph application an important workload. Because of the large scale of graph datasets, moving graph data to the cloud becomes a popular option. To keep the confidential and private graph secure from an untrusted cloud server, many cryptographic techniques are leveraged to hide the content of the data. However, protecting only the data content is not enough for a graph database. Because the structural information of the graph can be revealed through the database accessing track.\u0000 In this work, we study the graph neural network (GNN), an important graph workload to mine information from a graph database. We find that the server is able to infer which node is processing during the edge retrieving phase and also learn its neighbor indices during GNN's aggregation phase. This leads to the leakage of the information of graph structure data. In this work, we present SPG, a structure-private graph database with SqueezePIR. Our SPG is built on top of Private Information Retrieval (PIR), which securely hides which nodes/neighbors are accessed. In addition, we propose SqueezePIR, a compression technique to overcome the computation overhead of PIR. Based on our evaluation, our SqueezePIR achieves 11.85× speedup on average with less than 2% accuracy loss when compared to the state-of-the-art FastPIR protocol.","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"44 1","pages":"1615-1628"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86841932","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}
Pub Date : 2023-03-01DOI: 10.14778/3587136.3587144
Xun Jian, Zhiyuan Li, Lei Chen
Subgraph matching is a fundamental problem in graph theory and has wide applications in areas like sociology, chemistry, and social networks. Due to its NP-hardness, the basic approach is a brute-force search over the whole search space. Some pruning strategies have been proposed to reduce the search space. However, they are either space-inefficient or based on assumptions that the graph has specific properties. In this paper, we propose SUFF, a general and powerful structure filtering framework, which can accelerate most of the existing approaches with slight modifications. Specifically, it builds a set of filters using matching results of past queries, and uses them to prune the search space for future queries. By fully utilizing the relationship between matches of two queries, it ensures that such pruning is sound. Furthermore, several optimizations are proposed to reduce the computation and space cost for building, storing, and using filters. Extensive experiments are conducted on multiple real-world data sets and representative existing approaches. The results show that SUFF can achieve up to 15X speedup with small overheads.
{"title":"SUFF: Accelerating Subgraph Matching with Historical Data","authors":"Xun Jian, Zhiyuan Li, Lei Chen","doi":"10.14778/3587136.3587144","DOIUrl":"https://doi.org/10.14778/3587136.3587144","url":null,"abstract":"Subgraph matching is a fundamental problem in graph theory and has wide applications in areas like sociology, chemistry, and social networks. Due to its NP-hardness, the basic approach is a brute-force search over the whole search space. Some pruning strategies have been proposed to reduce the search space. However, they are either space-inefficient or based on assumptions that the graph has specific properties. In this paper, we propose SUFF, a general and powerful structure filtering framework, which can accelerate most of the existing approaches with slight modifications. Specifically, it builds a set of filters using matching results of past queries, and uses them to prune the search space for future queries. By fully utilizing the relationship between matches of two queries, it ensures that such pruning is sound. Furthermore, several optimizations are proposed to reduce the computation and space cost for building, storing, and using filters. Extensive experiments are conducted on multiple real-world data sets and representative existing approaches. The results show that SUFF can achieve up to 15X speedup with small overheads.","PeriodicalId":20467,"journal":{"name":"Proc. VLDB Endow.","volume":"7 1","pages":"1699-1711"},"PeriodicalIF":0.0,"publicationDate":"2023-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74351322","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：