This article proposes a class of dependencies for graphs, referred to as graph entity dependencies (GEDs). A GED is defined as a combination of a graph pattern and an attribute dependency. In a uniform format, GEDs can express graph functional dependencies with constant literals to catch inconsistencies, and keys carrying id literals to identify entities (vertices) in a graph. We revise the chase for GEDs and prove its Church-Rosser property. We characterize GED satisfiability and implication, and establish the complexity of these problems and the validation problem for GEDs, in the presence and absence of constant literals and id literals. We also develop a sound, complete and independent axiom system for finite implication of GEDs. In addition, we extend GEDs with built-in predicates or disjunctions, to strike a balance between the expressive power and complexity. We settle the complexity of the satisfiability, implication, and validation problems for these extensions.
{"title":"Dependencies for Graphs","authors":"W. Fan, Ping Lu","doi":"10.1145/3287285","DOIUrl":"https://doi.org/10.1145/3287285","url":null,"abstract":"This article proposes a class of dependencies for graphs, referred to as graph entity dependencies (GEDs). A GED is defined as a combination of a graph pattern and an attribute dependency. In a uniform format, GEDs can express graph functional dependencies with constant literals to catch inconsistencies, and keys carrying id literals to identify entities (vertices) in a graph. We revise the chase for GEDs and prove its Church-Rosser property. We characterize GED satisfiability and implication, and establish the complexity of these problems and the validation problem for GEDs, in the presence and absence of constant literals and id literals. We also develop a sound, complete and independent axiom system for finite implication of GEDs. In addition, we extend GEDs with built-in predicates or disjunctions, to strike a balance between the expressive power and complexity. We settle the complexity of the satisfiability, implication, and validation problems for these extensions.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"9 1","pages":"1 - 40"},"PeriodicalIF":0.0,"publicationDate":"2019-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82639933","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}
Joins are expensive, and online aggregation over joins was proposed to mitigate the cost, which offers users a nice and flexible tradeoff between query efficiency and accuracy in a continuous, online fashion. However, the state-of-the-art approach, in both internal and external memory, is based on ripple join, which is still very expensive and even needs unrealistic assumptions (e.g., tuples in a table are stored in random order). This article proposes a new approach, the wander join algorithm, to the online aggregation problem by performing random walks over the underlying join graph. We also design an optimizer that chooses the optimal plan for conducting the random walks without having to collect any statistics a priori. Compared with ripple join, wander join is particularly efficient for equality joins involving multiple tables, but also supports θ-joins. Selection predicates and group-by clauses can be handled as well. To demonstrate the usefulness of wander join, we have designed and implemented XDB (approXimate DB) by integrating wander join into various systems including PostgreSQL, Spark, and a stand-alone plug-in version using PL/SQL. The design and implementation of XDB has demonstrated wander join’s practicality in a full-fledged database system. Extensive experiments using the TPC-H benchmark have demonstrated the superior performance of wander join over ripple join.
{"title":"Wander Join and XDB","authors":"Feifei Li, Bin Wu, K. Yi, Zhuoyue Zhao","doi":"10.1145/3284551","DOIUrl":"https://doi.org/10.1145/3284551","url":null,"abstract":"Joins are expensive, and online aggregation over joins was proposed to mitigate the cost, which offers users a nice and flexible tradeoff between query efficiency and accuracy in a continuous, online fashion. However, the state-of-the-art approach, in both internal and external memory, is based on ripple join, which is still very expensive and even needs unrealistic assumptions (e.g., tuples in a table are stored in random order). This article proposes a new approach, the wander join algorithm, to the online aggregation problem by performing random walks over the underlying join graph. We also design an optimizer that chooses the optimal plan for conducting the random walks without having to collect any statistics a priori. Compared with ripple join, wander join is particularly efficient for equality joins involving multiple tables, but also supports θ-joins. Selection predicates and group-by clauses can be handled as well. To demonstrate the usefulness of wander join, we have designed and implemented XDB (approXimate DB) by integrating wander join into various systems including PostgreSQL, Spark, and a stand-alone plug-in version using PL/SQL. The design and implementation of XDB has demonstrated wander join’s practicality in a full-fledged database system. Extensive experiments using the TPC-H benchmark have demonstrated the superior performance of wander join over ripple join.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"13 1","pages":"1 - 41"},"PeriodicalIF":0.0,"publicationDate":"2019-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77722006","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 emerging applications are based on finding interesting subsequences from sequence data. Finding “prominent streaks,” a set of the longest contiguous subsequences with values all above (or below) a certain threshold, from sequence data is one of that kind that receives much attention. Motivated from real applications, we observe that prominent streaks alone are not insightful enough but require the discovery of something we coined as “historic moments” as companions. In this article, we present an algorithm to efficiently compute historic moments from sequence data. The algorithm is incremental and space optimal, meaning that when facing new data arrival, it is able to efficiently refresh the results by keeping minimal information. Case studies show that historic moments can significantly improve the insights offered by prominent streaks alone. Furthermore, experiments show that our algorithm can outperform the baseline in both time and space.
{"title":"Historic Moments Discovery in Sequence Data","authors":"Ran Bai, W. Hon, Eric Lo, Zhian He, Kenny Q. Zhu","doi":"10.1145/3276975","DOIUrl":"https://doi.org/10.1145/3276975","url":null,"abstract":"Many emerging applications are based on finding interesting subsequences from sequence data. Finding “prominent streaks,” a set of the longest contiguous subsequences with values all above (or below) a certain threshold, from sequence data is one of that kind that receives much attention. Motivated from real applications, we observe that prominent streaks alone are not insightful enough but require the discovery of something we coined as “historic moments” as companions. In this article, we present an algorithm to efficiently compute historic moments from sequence data. The algorithm is incremental and space optimal, meaning that when facing new data arrival, it is able to efficiently refresh the results by keeping minimal information. Case studies show that historic moments can significantly improve the insights offered by prominent streaks alone. Furthermore, experiments show that our algorithm can outperform the baseline in both time and space.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"25 1","pages":"1 - 33"},"PeriodicalIF":0.0,"publicationDate":"2019-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82312052","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}
Parallel dataflow engines such as Apache Hadoop, Apache Spark, and Apache Flink are an established alternative to relational databases for modern data analysis applications. A characteristic of these systems is a scalable programming model based on distributed collections and parallel transformations expressed by means of second-order functions such as map and reduce. Notable examples are Flink’s DataSet and Spark’s RDD programming abstractions. These programming models are realized as EDSLs—domain specific languages embedded in a general-purpose host language such as Java, Scala, or Python. This approach has several advantages over traditional external DSLs such as SQL or XQuery. First, syntactic constructs from the host language (e.g., anonymous functions syntax, value definitions, and fluent syntax via method chaining) can be reused in the EDSL. This eases the learning curve for developers already familiar with the host language. Second, it allows for seamless integration of library methods written in the host language via the function parameters passed to the parallel dataflow operators. This reduces the effort for developing analytics dataflows that go beyond pure SQL and require domain-specific logic. At the same time, however, state-of-the-art parallel dataflow EDSLs exhibit a number of shortcomings. First, one of the main advantages of an external DSL such as SQL—the high-level, declarative Select-From-Where syntax—is either lost completely or mimicked in a non-standard way. Second, execution aspects such as caching, join order, and partial aggregation have to be decided by the programmer. Optimizing them automatically is very difficult due to the limited program context available in the intermediate representation of the DSL. In this article, we argue that the limitations listed above are a side effect of the adopted type-based embedding approach. As a solution, we propose an alternative EDSL design based on quotations. We present a DSL embedded in Scala and discuss its compiler pipeline, intermediate representation, and some of the enabled optimizations. We promote the algebraic type of bags in union representation as a model for distributed collections and its associated structural recursion scheme and monad as a model for parallel collection processing. At the source code level, Scala’s comprehension syntax over a bag monad can be used to encode Select-From-Where expressions in a standard way. At the intermediate representation level, maintaining comprehensions as a first-class citizen can be used to simplify the design and implementation of holistic dataflow optimizations that accommodate for nesting and control-flow. The proposed DSL design therefore reconciles the benefits of embedded parallel dataflow DSLs with the declarativity and optimization potential of external DSLs like SQL.
{"title":"Representations and Optimizations for Embedded Parallel Dataflow Languages","authors":"Alexander B. Alexandrov, Georgi Krastev, V. Markl","doi":"10.1145/3281629","DOIUrl":"https://doi.org/10.1145/3281629","url":null,"abstract":"Parallel dataflow engines such as Apache Hadoop, Apache Spark, and Apache Flink are an established alternative to relational databases for modern data analysis applications. A characteristic of these systems is a scalable programming model based on distributed collections and parallel transformations expressed by means of second-order functions such as map and reduce. Notable examples are Flink’s DataSet and Spark’s RDD programming abstractions. These programming models are realized as EDSLs—domain specific languages embedded in a general-purpose host language such as Java, Scala, or Python. This approach has several advantages over traditional external DSLs such as SQL or XQuery. First, syntactic constructs from the host language (e.g., anonymous functions syntax, value definitions, and fluent syntax via method chaining) can be reused in the EDSL. This eases the learning curve for developers already familiar with the host language. Second, it allows for seamless integration of library methods written in the host language via the function parameters passed to the parallel dataflow operators. This reduces the effort for developing analytics dataflows that go beyond pure SQL and require domain-specific logic. At the same time, however, state-of-the-art parallel dataflow EDSLs exhibit a number of shortcomings. First, one of the main advantages of an external DSL such as SQL—the high-level, declarative Select-From-Where syntax—is either lost completely or mimicked in a non-standard way. Second, execution aspects such as caching, join order, and partial aggregation have to be decided by the programmer. Optimizing them automatically is very difficult due to the limited program context available in the intermediate representation of the DSL. In this article, we argue that the limitations listed above are a side effect of the adopted type-based embedding approach. As a solution, we propose an alternative EDSL design based on quotations. We present a DSL embedded in Scala and discuss its compiler pipeline, intermediate representation, and some of the enabled optimizations. We promote the algebraic type of bags in union representation as a model for distributed collections and its associated structural recursion scheme and monad as a model for parallel collection processing. At the source code level, Scala’s comprehension syntax over a bag monad can be used to encode Select-From-Where expressions in a standard way. At the intermediate representation level, maintaining comprehensions as a first-class citizen can be used to simplify the design and implementation of holistic dataflow optimizations that accommodate for nesting and control-flow. The proposed DSL design therefore reconciles the benefits of embedded parallel dataflow DSLs with the declarativity and optimization potential of external DSLs like SQL.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"12 1","pages":"1 - 44"},"PeriodicalIF":0.0,"publicationDate":"2019-01-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91106125","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}
Andreas Kipf, Varun Pandey, Jan Böttcher, Lucas Braun, Thomas Neumann, A. Kemper
Today’s streaming applications demand increasingly high event throughput rates and are often subject to strict latency constraints. To allow for more complex workloads, such as window-based aggregations, streaming systems need to support stateful event processing. This introduces new challenges for streaming engines as the state needs to be maintained in a consistent and durable manner and simultaneously accessed by complex queries for real-time analytics. Modern streaming systems, such as Apache Flink, do not allow for efficiently exposing the state to analytical queries. Thus, data engineers are forced to keep the state in external data stores, which significantly increases the latencies until events become visible to analytical queries. Proprietary solutions have been created to meet data freshness constraints. These solutions are expensive, error-prone, and difficult to maintain. Main-memory database systems, such as HyPer, achieve extremely low query response times while maintaining high update rates, which makes them well-suited for analytical streaming workloads. In this article, we explore extensions to database systems to match the performance and usability of streaming systems.
{"title":"Scalable Analytics on Fast Data","authors":"Andreas Kipf, Varun Pandey, Jan Böttcher, Lucas Braun, Thomas Neumann, A. Kemper","doi":"10.1145/3283811","DOIUrl":"https://doi.org/10.1145/3283811","url":null,"abstract":"Today’s streaming applications demand increasingly high event throughput rates and are often subject to strict latency constraints. To allow for more complex workloads, such as window-based aggregations, streaming systems need to support stateful event processing. This introduces new challenges for streaming engines as the state needs to be maintained in a consistent and durable manner and simultaneously accessed by complex queries for real-time analytics. Modern streaming systems, such as Apache Flink, do not allow for efficiently exposing the state to analytical queries. Thus, data engineers are forced to keep the state in external data stores, which significantly increases the latencies until events become visible to analytical queries. Proprietary solutions have been created to meet data freshness constraints. These solutions are expensive, error-prone, and difficult to maintain. Main-memory database systems, such as HyPer, achieve extremely low query response times while maintaining high update rates, which makes them well-suited for analytical streaming workloads. In this article, we explore extensions to database systems to match the performance and usability of streaming systems.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"6 1","pages":"1 - 35"},"PeriodicalIF":0.0,"publicationDate":"2019-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90035050","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}
W. Fan, Jingbo Xu, Yinghui Wu, Wenyuan Yu, Jiaxin Jiang, Zeyu Zheng, Bohan Zhang, Yang Cao, Chao Tian
This article presents GRAPE, a parallel GRAPh Engine for graph computations. GRAPE differs from prior systems in its ability to parallelize existing sequential graph algorithms as a whole, without the need for recasting the entire algorithm into a new model. Underlying GRAPE are a simple programming model and a principled approach based on fixpoint computation that starts with partial evaluation and uses an incremental function as the intermediate consequence operator. We show that users can devise existing sequential graph algorithms with minor additions, and GRAPE parallelizes the computation. Under a monotonic condition, the GRAPE parallelization guarantees to converge at correct answers as long as the sequential algorithms are correct. Moreover, we show that algorithms in MapReduce, BSP, and PRAM can be optimally simulated on GRAPE. In addition to the ease of programming, we experimentally verify that GRAPE achieves comparable performance to the state-of-the-art graph systems using real-life and synthetic graphs.
{"title":"Parallelizing Sequential Graph Computations","authors":"W. Fan, Jingbo Xu, Yinghui Wu, Wenyuan Yu, Jiaxin Jiang, Zeyu Zheng, Bohan Zhang, Yang Cao, Chao Tian","doi":"10.1145/3282488","DOIUrl":"https://doi.org/10.1145/3282488","url":null,"abstract":"This article presents GRAPE, a parallel GRAPh Engine for graph computations. GRAPE differs from prior systems in its ability to parallelize existing sequential graph algorithms as a whole, without the need for recasting the entire algorithm into a new model. Underlying GRAPE are a simple programming model and a principled approach based on fixpoint computation that starts with partial evaluation and uses an incremental function as the intermediate consequence operator. We show that users can devise existing sequential graph algorithms with minor additions, and GRAPE parallelizes the computation. Under a monotonic condition, the GRAPE parallelization guarantees to converge at correct answers as long as the sequential algorithms are correct. Moreover, we show that algorithms in MapReduce, BSP, and PRAM can be optimally simulated on GRAPE. In addition to the ease of programming, we experimentally verify that GRAPE achieves comparable performance to the state-of-the-art graph systems using real-life and synthetic graphs.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"21 1","pages":"1 - 39"},"PeriodicalIF":0.0,"publicationDate":"2018-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84991244","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}
We study the enumeration complexity of Unions of Conjunctive Queries (UCQs). We aim to identify the UCQs that are tractable in the sense that the answer tuples can be enumerated with a linear preprocessing phase and a constant delay between every successive tuples. It has been established that, in the absence of self-joins and under conventional complexity assumptions, the CQs that admit such an evaluation are precisely the free-connex ones. A union of tractable CQs is always tractable. We generalize the notion of free-connexity from CQs to UCQs, thus showing that some unions containing intractable CQs are, in fact, tractable. Interestingly, some unions consisting of only intractable CQs are tractable too. We show how to use the techniques presented in this article also in settings where the database contains cardinality dependencies (including functional dependencies and key constraints) or when the UCQs contain disequalities. The question of finding a full characterization of the tractability of UCQs remains open. Nevertheless, we prove that, for several classes of queries, free-connexity fully captures the tractable UCQs.
{"title":"On the Enumeration Complexity of Unions of Conjunctive Queries","authors":"Nofar Carmeli, Markus Kröll","doi":"10.1145/3450263","DOIUrl":"https://doi.org/10.1145/3450263","url":null,"abstract":"We study the enumeration complexity of Unions of Conjunctive Queries (UCQs). We aim to identify the UCQs that are tractable in the sense that the answer tuples can be enumerated with a linear preprocessing phase and a constant delay between every successive tuples. It has been established that, in the absence of self-joins and under conventional complexity assumptions, the CQs that admit such an evaluation are precisely the free-connex ones. A union of tractable CQs is always tractable. We generalize the notion of free-connexity from CQs to UCQs, thus showing that some unions containing intractable CQs are, in fact, tractable. Interestingly, some unions consisting of only intractable CQs are tractable too. We show how to use the techniques presented in this article also in settings where the database contains cardinality dependencies (including functional dependencies and key constraints) or when the UCQs contain disequalities. The question of finding a full characterization of the tractability of UCQs remains open. Nevertheless, we prove that, for several classes of queries, free-connexity fully captures the tractable UCQs.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"37 1","pages":"1 - 41"},"PeriodicalIF":0.0,"publicationDate":"2018-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77062561","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}
Niccolò Meneghetti, Oliver Kennedy, Wolfgang Gatterbauer
Tuple-independent and disjoint-independent probabilistic databases (TI- and DI-PDBs) represent uncertain data in a factorized form as a product of independent random variables that represent either tuples (TI-PDBs) or sets of tuples (DI-PDBs). When the user submits a query, the database derives the marginal probabilities of each output-tuple, exploiting the underlying assumptions of statistical independence. While query processing in TI- and DI-PDBs has been studied extensively, limited research has been dedicated to the problems of updating or deriving the parameters from observations of query results. Addressing this problem is the main focus of this article. We first introduce Beta Probabilistic Databases (B-PDBs), a generalization of TI-PDBs designed to support both (i) belief updating and (ii) parameter learning in a principled and scalable way. The key idea of B-PDBs is to treat each parameter as a latent, Beta-distributed random variable. We show how this simple expedient enables both belief updating and parameter learning in a principled way, without imposing any burden on regular query processing. Building on B-PDBs, we then introduce Dirichlet Probabilistic Databases (D-PDBs), a generalization of DI-PDBs with similar properties. We provide the following key contributions for both B- and D-PDBs: (i) We study the complexity of performing Bayesian belief updates and devise efficient algorithms for certain tractable classes of queries; (ii) we propose a soft-EM algorithm for computing maximum-likelihood estimates of the parameters; (iii) we present an algorithm for efficiently computing conditional probabilities, allowing us to efficiently implement B- and D-PDBs via a standard relational engine; and (iv) we support our conclusions with extensive experimental results.
{"title":"Learning From Query-Answers","authors":"Niccolò Meneghetti, Oliver Kennedy, Wolfgang Gatterbauer","doi":"10.1145/3277503","DOIUrl":"https://doi.org/10.1145/3277503","url":null,"abstract":"Tuple-independent and disjoint-independent probabilistic databases (TI- and DI-PDBs) represent uncertain data in a factorized form as a product of independent random variables that represent either tuples (TI-PDBs) or sets of tuples (DI-PDBs). When the user submits a query, the database derives the marginal probabilities of each output-tuple, exploiting the underlying assumptions of statistical independence. While query processing in TI- and DI-PDBs has been studied extensively, limited research has been dedicated to the problems of updating or deriving the parameters from observations of query results. Addressing this problem is the main focus of this article. We first introduce Beta Probabilistic Databases (B-PDBs), a generalization of TI-PDBs designed to support both (i) belief updating and (ii) parameter learning in a principled and scalable way. The key idea of B-PDBs is to treat each parameter as a latent, Beta-distributed random variable. We show how this simple expedient enables both belief updating and parameter learning in a principled way, without imposing any burden on regular query processing. Building on B-PDBs, we then introduce Dirichlet Probabilistic Databases (D-PDBs), a generalization of DI-PDBs with similar properties. We provide the following key contributions for both B- and D-PDBs: (i) We study the complexity of performing Bayesian belief updates and devise efficient algorithms for certain tractable classes of queries; (ii) we propose a soft-EM algorithm for computing maximum-likelihood estimates of the parameters; (iii) we present an algorithm for efficiently computing conditional probabilities, allowing us to efficiently implement B- and D-PDBs via a standard relational engine; and (iv) we support our conclusions with extensive experimental results.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"50 1","pages":"1 - 41"},"PeriodicalIF":0.0,"publicationDate":"2018-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81284892","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}
In this article, we show that key-value stores backed by a log-structured merge-tree (LSM-tree) exhibit an intrinsic tradeoff between lookup cost, update cost, and main memory footprint, yet all existing designs expose a suboptimal and difficult to tune tradeoff among these metrics. We pinpoint the problem to the fact that modern key-value stores suboptimally co-tune the merge policy, the buffer size, and the Bloom filters’ false-positive rates across the LSM-tree’s different levels. We present Monkey, an LSM-tree based key-value store that strikes the optimal balance between the costs of updates and lookups with any given main memory budget. The core insight is that worst-case lookup cost is proportional to the sum of the false-positive rates of the Bloom filters across all levels of the LSM-tree. Contrary to state-of-the-art key-value stores that assign a fixed number of bits-per-element to all Bloom filters, Monkey allocates memory to filters across different levels so as to minimize the sum of their false-positive rates. We show analytically that Monkey reduces the asymptotic complexity of the worst-case lookup I/O cost, and we verify empirically using an implementation on top of RocksDB that Monkey reduces lookup latency by an increasing margin as the data volume grows (50--80% for the data sizes we experimented with). Furthermore, we map the design space onto a closed-form model that enables adapting the merging frequency and memory allocation to strike the best tradeoff among lookup cost, update cost and main memory, depending on the workload (proportion of lookups and updates), the dataset (number and size of entries), and the underlying hardware (main memory available, disk vs. flash). We show how to use this model to answer what-if design questions about how changes in environmental parameters impact performance and how to adapt the design of the key-value store for optimal performance.
{"title":"Optimal Bloom Filters and Adaptive Merging for LSM-Trees","authors":"Niv Dayan, Manos Athanassoulis, Stratos Idreos","doi":"10.1145/3276980","DOIUrl":"https://doi.org/10.1145/3276980","url":null,"abstract":"In this article, we show that key-value stores backed by a log-structured merge-tree (LSM-tree) exhibit an intrinsic tradeoff between lookup cost, update cost, and main memory footprint, yet all existing designs expose a suboptimal and difficult to tune tradeoff among these metrics. We pinpoint the problem to the fact that modern key-value stores suboptimally co-tune the merge policy, the buffer size, and the Bloom filters’ false-positive rates across the LSM-tree’s different levels. We present Monkey, an LSM-tree based key-value store that strikes the optimal balance between the costs of updates and lookups with any given main memory budget. The core insight is that worst-case lookup cost is proportional to the sum of the false-positive rates of the Bloom filters across all levels of the LSM-tree. Contrary to state-of-the-art key-value stores that assign a fixed number of bits-per-element to all Bloom filters, Monkey allocates memory to filters across different levels so as to minimize the sum of their false-positive rates. We show analytically that Monkey reduces the asymptotic complexity of the worst-case lookup I/O cost, and we verify empirically using an implementation on top of RocksDB that Monkey reduces lookup latency by an increasing margin as the data volume grows (50--80% for the data sizes we experimented with). Furthermore, we map the design space onto a closed-form model that enables adapting the merging frequency and memory allocation to strike the best tradeoff among lookup cost, update cost and main memory, depending on the workload (proportion of lookups and updates), the dataset (number and size of entries), and the underlying hardware (main memory available, disk vs. flash). We show how to use this model to answer what-if design questions about how changes in environmental parameters impact performance and how to adapt the design of the key-value store for optimal performance.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"1 1","pages":"1 - 48"},"PeriodicalIF":0.0,"publicationDate":"2018-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86446463","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}
Firas Abuzaid, Peter D. Bailis, Jialin Ding, Edward Gan, S. Madden, D. Narayanan, Kexin Rong, S. Suri
As data volumes continue to rise, manual inspection is becoming increasingly untenable. In response, we present MacroBase, a data analytics engine that prioritizes end-user attention in high-volume fast data streams. MacroBase enables efficient, accurate, and modular analyses that highlight and aggregate important and unusual behavior, acting as a search engine for fast data. MacroBase is able to deliver order-of-magnitude speedups over alternatives by optimizing the combination of explanation (i.e., feature selection) and classification tasks and by leveraging a new reservoir sampler and heavy-hitters sketch specialized for fast data streams. As a result, MacroBase delivers accurate results at speeds of up to 2M events per second per query on a single core. The system has delivered meaningful results in production, including at a telematics company monitoring hundreds of thousands of vehicles.
{"title":"MacroBase","authors":"Firas Abuzaid, Peter D. Bailis, Jialin Ding, Edward Gan, S. Madden, D. Narayanan, Kexin Rong, S. Suri","doi":"10.1145/3276463","DOIUrl":"https://doi.org/10.1145/3276463","url":null,"abstract":"As data volumes continue to rise, manual inspection is becoming increasingly untenable. In response, we present MacroBase, a data analytics engine that prioritizes end-user attention in high-volume fast data streams. MacroBase enables efficient, accurate, and modular analyses that highlight and aggregate important and unusual behavior, acting as a search engine for fast data. MacroBase is able to deliver order-of-magnitude speedups over alternatives by optimizing the combination of explanation (i.e., feature selection) and classification tasks and by leveraging a new reservoir sampler and heavy-hitters sketch specialized for fast data streams. As a result, MacroBase delivers accurate results at speeds of up to 2M events per second per query on a single core. The system has delivered meaningful results in production, including at a telematics company monitoring hundreds of thousands of vehicles.","PeriodicalId":6983,"journal":{"name":"ACM Transactions on Database Systems (TODS)","volume":"17 1","pages":"1 - 45"},"PeriodicalIF":0.0,"publicationDate":"2018-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73922817","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：