Figure 5:Vaejovis lapidicola, male topotype, mating plug from left hemispermatophore: distal (5a)... more Figure 5:Vaejovis lapidicola, male topotype, mating plug from left hemispermatophore: distal (5a), dorsal (5b), proximal (5c), ventral (5d), internal (5e), and external (5f) views. Scale bar: 200 µm.
Figures 1–2. Vaejovis lapidicola, male topotype on red sandstone from the type locality (1) and f... more Figures 1–2. Vaejovis lapidicola, male topotype on red sandstone from the type locality (1) and female topotype in vivo habitus (2).
USENIX conference on Hot topics in parallelism, Jun 7, 2012
For communication-intensive applications on distributed memory systems, performance is bounded by... more For communication-intensive applications on distributed memory systems, performance is bounded by remote memory accesses. Task migration is a potential candidate for reducing network traffic in such applications, thereby improving performance. We seek to answer the question: can a runtime profitably predict when it is better to move the task to the data than move the data to the task? Using a simple model where local work is free and data transferred over the network is costly, we show that a best case task migration schedule can achieve up to 3.5x less total data transferred than no migration for some benchmarks. Given this observation, we develop and evaluate two online task migration policies: Stream Predictor, which uses only immediate remote access history, and Hindsight Migrate, which tracks instruction addresses where task migration is predicted to be beneficial. These predictor policies are able to provide benefit over execution with no migration for small or moderate size tasks on our tested applications. Min. total_bytes
Preston Briggs is a Senior Engineer at Reservoir Labs and an Affiliate Professor in Computer Scie... more Preston Briggs is a Senior Engineer at Reservoir Labs and an Affiliate Professor in Computer Science and Engineering at the University of Washington. He received a PhD in computer science from Rice University in 1992. [email protected] The rising importance of data-intensive applications has fueled the growth of a plethora of distributed computing frameworks, including Hadoop, Spark, and GraphLab. We have developed a system called Grappa [1, 2] to aid programmers in developing new frameworks. Grappa provides a distributed shared memory abstraction to hide complexity from the programmer, and takes advantage of parallelism in the data to hide remote access latency and to trade latency for more performance. These techniques allow it to outperform existing frameworks by up to an order of magnitude.
Proceedings of the 50th ACM Technical Symposium on Computer Science Education, 2019
The computer organization course must help students acquire difficult conceptual knowledge and de... more The computer organization course must help students acquire difficult conceptual knowledge and design skills, and improve their teamwork skills for subsequent project courses. Prior research supports that cooperative learning, in which students work together to achieve common goals, may address these challenges. We studied whether increasing the amount of guided inquiry activities and the cooperative support for it (teams and reflection) in an intermediate computer science course would improve achievement and engagement. The intervention group had lower scores on one of two achievement measures, lower engagement, and lower task value of collaborative activities. Qualitative analysis showed that students valued hands-on learning yet resisted guided inquiry, suggesting that sharing the purpose of each type of activity is important. Furthermore, the results showed that students valued learning with peers but were frustrated by group dysfunction, suggesting that instructors must address teamwork comprehensively to realize the benefits of cooperative learning.
Proceedings of the 52nd ACM Technical Symposium on Computer Science Education, 2021
Collaborative assessments, in which students take an individual assessment and then retake it wit... more Collaborative assessments, in which students take an individual assessment and then retake it with peers, are valued by students and contribute to learning. We studied a collaborative assessment technique as part of an instructional strategy called Team-Based Learning (TBL). This environment differs from those of prior studies on collaborative assessments in computing. Specifically, TBL uses teams over informal groups and 'medium-stakes' formative assessment that provides accountability for learning. We asked whether collaborative quizzes contribute to immediate conceptual understanding in CS2 and whether students improve the use of and attitudes toward peer learning over time. In our within-subjects study, we found learning gains on average and with three of five quizzes. We used a self-reported measure of team learning to model improvements in score, but it was not predictive. Looking across quizzes, we found modest evidence of improved learning over time, and by using pre/post semester surveys, we found an improvement in two of three measures of attitude towards peer learning.
Euscorpius is the first research publication completely devoted to scorpions (Arachnida: Scorpion... more Euscorpius is the first research publication completely devoted to scorpions (Arachnida: Scorpiones). Euscorpius takes advantage of the rapidly evolving medium of quick online publication, at the same time maintaining high research standards for the burgeoning field of scorpion science (scorpiology). Euscorpius is an expedient and viable medium for the publication of serious papers in scorpiology, including (but not limited to): systematics, evolution, ecology, biogeography, and general biology of scorpions. Review papers, descriptions of new taxa, faunistic surveys, lists of museum collections, and book reviews are welcome. Derivatio Nominis The name Euscorpius Thorell, 1876 refers to the most common genus of scorpions in the Mediterranean region and southern Europe (family Euscorpiidae).
Herein we describe a new species of the Buthidae scorpion genus Centruroides Marx, 1890. It occur... more Herein we describe a new species of the Buthidae scorpion genus Centruroides Marx, 1890. It occurs at least in two small offshore islands of the southwestern Caribbean: Cozumel in Mexico and Guanaja in Honduras, based upon type specimens from the former and photographic evidence from the latter. It belongs in the "gracilis" species-group and is most closely related to both Centruroides gracilis (Latreille, 1805) and Centruroides nigrescens (Pocock, 1898).
Resource Managers like YARN and Mesos have emerged as a critical layer in the cloud computing sys... more Resource Managers like YARN and Mesos have emerged as a critical layer in the cloud computing system stack, but the developer abstractions for leasing cluster resources and instantiating application logic are very low level. This flexibility comes at a high cost in terms of developer effort, as each application must repeatedly tackle the same challenges (e.g., fault tolerance, task scheduling and coordination) and reimplement common mechanisms (e.g., caching, bulk-data transfers). This article presents REEF, a development framework that provides a control plane for scheduling and coordinating task-level (data-plane) work on cluster resources obtained from a Resource Manager. REEF provides mechanisms that facilitate resource reuse for data caching and state management abstractions that greatly ease the development of elastic data processing pipelines on cloud platforms that support a Resource Manager service. We illustrate the power of REEF by showing applications built atop: a distr...
We present Radish, a query compiler that generates distributed programs. Recent efforts have show... more We present Radish, a query compiler that generates distributed programs. Recent efforts have shown that compiling queries to machine code for a single-core can remove iterator and control overhead for significant performance gains. So far, systems that generate distributed programs only compile plans for single processors and stitch them together with messaging. In this paper, we describe an approach for translating query plans into distributed programs by targeting the partitioned global address space (PGAS) parallel programming model as an intermediate representation. This approach affords a natural adaptation of pipelining techniques used in singlecore query compilers and an overall simpler design. We adapt pipelined algorithms to PGAS languages, describe efficient data structures for PGAS query execution, and implement techniques for mitigating the overhead resulting from handling a multitude of fine-grained tasks. We evaluate Radish on graph benchmark and application workloads and find that it is 4× to 100× faster than Shark, a recent distributed query engine optimized for in-memory execution. Our work makes important first steps towards ensuring that query processing systems can benefit from future advances in parallel programming and co-mingle with state-of-the-art parallel programs. Prior systems have demonstrated orders of magnitude performance improvements over iterator based query processing by compiling plans for single processors [10, 23, 27, 34]. Frameworks that generate distributed programs only compile plans for individual processors and stitch them together with communication calls, retaining the iterator model [15, 33]. These systems have a common shortcoming: they depend directly on a single-node compiler (e.g. LLVM, JVM) to perform machine-level optimizations, but these compilers cannot reason about a distributed program. The alternative, which we explore in this paper, is to generate programs for a partitioned global address space (PGAS) language (e.g., Chapel [11] or X10 [13]), then compile and execute these distributed programs to evaluate the query. A key feature of these languages is a partition, which is a region of memory local to a particular processor and far from other processors. Consider this query with a join and multiplication:
The implementation of scalable synchronized data structures is notoriously difficult. Recent work... more The implementation of scalable synchronized data structures is notoriously difficult. Recent work in shared-memory multicores introduced a new synchronization paradigm called flat combining that allows many concurrent accessors to cooperate efficiently to reduce contention on shared locks. In this work we introduce this paradigm to a domain where reducing communication is paramount: distributed memory systems. We implement a flat combining framework for Grappa, a latency-tolerant PGAS runtime, and show how it can be used to implement synchronized global data structures. Even using simple locking schemes, we find that these flat-combining data structures scale out to 64 nodes with 2x-100x improvement in throughput. We also demonstrate that this translates to application performance via two simple graph analysis kernels. The higher communication cost and structured concurrency of Grappa lead to a new form of distributed flat combining that drastically reduces the amount of communication necessary for maintaining global sequential consistency.
We present Grappa, a modern take on software distributed shared memory (DSM) for in-memory data-i... more We present Grappa, a modern take on software distributed shared memory (DSM) for in-memory data-intensive applications. Grappa enables users to program a cluster as if it were a single, large, non-uniform memory access (NUMA) machine. Performance scales up even for applications that have poor locality and input-dependent load distribution. Grappa addresses deficiencies of previous DSM systems by exploiting application parallelism, trading off latency for throughput. We evaluate Grappa with an in-memory MapReduce framework (10⇥ faster than Spark [74]); a vertex-centric framework inspired by GraphLab (1.33⇥ faster than native GraphLab [48]); and a relational query execution engine (12.5⇥ faster than Shark [31]). All these frameworks required only 60-690 lines of Grappa code.
Grappa is a runtime system for commodity clusters of multicore computers that presents a massivel... more Grappa is a runtime system for commodity clusters of multicore computers that presents a massively parallel, single address space abstraction to applications. Grappa's purpose is to enable scalable performance of irregular parallel applications, such as branch and bound optimization, SPICE circuit simulation, and graph processing. Poor data locality, imbalanced parallel work and complex communication patterns make scaling these applications difficult. Grappa serves both as a C++ user library and as a foundation for higher level languages. Grappa tolerates delays to remote memory by multiplexing thousands of lightweight workers to each processor core, balances load via fine-grained distributed work-stealing, increases communication throughput by aggregating smaller data requests into large ones, and provides efficient synchronization and remote operations. We present a description of the Grappa system and an evaluation of its performance on microbenchmarks.
Dissatisfaction with relational databases for large-scale graph processing has motivated a new cl... more Dissatisfaction with relational databases for large-scale graph processing has motivated a new class of graph databases that offer fast graph processing but sacrifice the ability to express basic relational idioms. However, we hypothesize that the performance benefits amount to implementation details, not a fundamental limitation of the relational model. To evaluate this hypothesis, we are exploring code-generation to produce fast in-memory algorithms and data structures for graph patterns that are inaccessible to conventional relational optimizers. In this paper, we present preliminary results for this approach on path-counting queries, which includes triangle counting as a special case. We compile Datalog queries into main-memory pipelined hash-join plans in C++, and show that the resulting programs easily outperform PostgreSQL on real graphs with different degrees of skew. We then produce analogous parallel programs for Grappa, a runtime system for distributed memory architectures. Grappa is a good target for building a parallel query system as its shared memory programming model and communication mechanisms provide productivity and performance when building communication-intensive applications. Our experiments suggest that Grappa programs using hash joins have competitive performance with queries executed on Greenplum, a commercial parallel database. We find preliminary evidence that a code generation approach simplifies the design of a query engine for graph analysis and improves performance over conventional relational databases.
In this demo proposal, we describe REEF, a framework that makes it easy to implement scalable, fa... more In this demo proposal, we describe REEF, a framework that makes it easy to implement scalable, fault-tolerant runtime environments for a range of computational models. We will demonstrate diverse workloads, including extract-transform-load MapReduce jobs, iterative machine learning algorithms, and ad-hoc declarative query processing. At its core, REEF builds atop YARN (Apache Hadoop 2's resource manager) to provide retainable hardware resources with lifetimes that are decoupled from those of computational tasks. This allows us to build persistent (cross-job) caches and cluster-wide services, but, more importantly, supports high-performance iterative graph processing and machine learning algorithms. Unlike existing systems, REEF aims for composability of jobs across computational models, providing significant performance and usability gains, even with legacy code. REEF includes a library of interoperable data management primitives optimized for communication and data movement (wh...
This study sought to determine what aspect of divorce was most likely associated with an increase... more This study sought to determine what aspect of divorce was most likely associated with an increase in criminally delinquent behavior among adolescent children under the age of eighteen years. The researcher utilized the National Longitudinal Survey of Youth 1997, NLSY97, as the primary research tool which was a ten year survey of nearly 9,000 youths and household members. The researcher
Figure 5:Vaejovis lapidicola, male topotype, mating plug from left hemispermatophore: distal (5a)... more Figure 5:Vaejovis lapidicola, male topotype, mating plug from left hemispermatophore: distal (5a), dorsal (5b), proximal (5c), ventral (5d), internal (5e), and external (5f) views. Scale bar: 200 µm.
Figures 1–2. Vaejovis lapidicola, male topotype on red sandstone from the type locality (1) and f... more Figures 1–2. Vaejovis lapidicola, male topotype on red sandstone from the type locality (1) and female topotype in vivo habitus (2).
USENIX conference on Hot topics in parallelism, Jun 7, 2012
For communication-intensive applications on distributed memory systems, performance is bounded by... more For communication-intensive applications on distributed memory systems, performance is bounded by remote memory accesses. Task migration is a potential candidate for reducing network traffic in such applications, thereby improving performance. We seek to answer the question: can a runtime profitably predict when it is better to move the task to the data than move the data to the task? Using a simple model where local work is free and data transferred over the network is costly, we show that a best case task migration schedule can achieve up to 3.5x less total data transferred than no migration for some benchmarks. Given this observation, we develop and evaluate two online task migration policies: Stream Predictor, which uses only immediate remote access history, and Hindsight Migrate, which tracks instruction addresses where task migration is predicted to be beneficial. These predictor policies are able to provide benefit over execution with no migration for small or moderate size tasks on our tested applications. Min. total_bytes
Preston Briggs is a Senior Engineer at Reservoir Labs and an Affiliate Professor in Computer Scie... more Preston Briggs is a Senior Engineer at Reservoir Labs and an Affiliate Professor in Computer Science and Engineering at the University of Washington. He received a PhD in computer science from Rice University in 1992. [email protected] The rising importance of data-intensive applications has fueled the growth of a plethora of distributed computing frameworks, including Hadoop, Spark, and GraphLab. We have developed a system called Grappa [1, 2] to aid programmers in developing new frameworks. Grappa provides a distributed shared memory abstraction to hide complexity from the programmer, and takes advantage of parallelism in the data to hide remote access latency and to trade latency for more performance. These techniques allow it to outperform existing frameworks by up to an order of magnitude.
Proceedings of the 50th ACM Technical Symposium on Computer Science Education, 2019
The computer organization course must help students acquire difficult conceptual knowledge and de... more The computer organization course must help students acquire difficult conceptual knowledge and design skills, and improve their teamwork skills for subsequent project courses. Prior research supports that cooperative learning, in which students work together to achieve common goals, may address these challenges. We studied whether increasing the amount of guided inquiry activities and the cooperative support for it (teams and reflection) in an intermediate computer science course would improve achievement and engagement. The intervention group had lower scores on one of two achievement measures, lower engagement, and lower task value of collaborative activities. Qualitative analysis showed that students valued hands-on learning yet resisted guided inquiry, suggesting that sharing the purpose of each type of activity is important. Furthermore, the results showed that students valued learning with peers but were frustrated by group dysfunction, suggesting that instructors must address teamwork comprehensively to realize the benefits of cooperative learning.
Proceedings of the 52nd ACM Technical Symposium on Computer Science Education, 2021
Collaborative assessments, in which students take an individual assessment and then retake it wit... more Collaborative assessments, in which students take an individual assessment and then retake it with peers, are valued by students and contribute to learning. We studied a collaborative assessment technique as part of an instructional strategy called Team-Based Learning (TBL). This environment differs from those of prior studies on collaborative assessments in computing. Specifically, TBL uses teams over informal groups and 'medium-stakes' formative assessment that provides accountability for learning. We asked whether collaborative quizzes contribute to immediate conceptual understanding in CS2 and whether students improve the use of and attitudes toward peer learning over time. In our within-subjects study, we found learning gains on average and with three of five quizzes. We used a self-reported measure of team learning to model improvements in score, but it was not predictive. Looking across quizzes, we found modest evidence of improved learning over time, and by using pre/post semester surveys, we found an improvement in two of three measures of attitude towards peer learning.
Euscorpius is the first research publication completely devoted to scorpions (Arachnida: Scorpion... more Euscorpius is the first research publication completely devoted to scorpions (Arachnida: Scorpiones). Euscorpius takes advantage of the rapidly evolving medium of quick online publication, at the same time maintaining high research standards for the burgeoning field of scorpion science (scorpiology). Euscorpius is an expedient and viable medium for the publication of serious papers in scorpiology, including (but not limited to): systematics, evolution, ecology, biogeography, and general biology of scorpions. Review papers, descriptions of new taxa, faunistic surveys, lists of museum collections, and book reviews are welcome. Derivatio Nominis The name Euscorpius Thorell, 1876 refers to the most common genus of scorpions in the Mediterranean region and southern Europe (family Euscorpiidae).
Herein we describe a new species of the Buthidae scorpion genus Centruroides Marx, 1890. It occur... more Herein we describe a new species of the Buthidae scorpion genus Centruroides Marx, 1890. It occurs at least in two small offshore islands of the southwestern Caribbean: Cozumel in Mexico and Guanaja in Honduras, based upon type specimens from the former and photographic evidence from the latter. It belongs in the "gracilis" species-group and is most closely related to both Centruroides gracilis (Latreille, 1805) and Centruroides nigrescens (Pocock, 1898).
Resource Managers like YARN and Mesos have emerged as a critical layer in the cloud computing sys... more Resource Managers like YARN and Mesos have emerged as a critical layer in the cloud computing system stack, but the developer abstractions for leasing cluster resources and instantiating application logic are very low level. This flexibility comes at a high cost in terms of developer effort, as each application must repeatedly tackle the same challenges (e.g., fault tolerance, task scheduling and coordination) and reimplement common mechanisms (e.g., caching, bulk-data transfers). This article presents REEF, a development framework that provides a control plane for scheduling and coordinating task-level (data-plane) work on cluster resources obtained from a Resource Manager. REEF provides mechanisms that facilitate resource reuse for data caching and state management abstractions that greatly ease the development of elastic data processing pipelines on cloud platforms that support a Resource Manager service. We illustrate the power of REEF by showing applications built atop: a distr...
We present Radish, a query compiler that generates distributed programs. Recent efforts have show... more We present Radish, a query compiler that generates distributed programs. Recent efforts have shown that compiling queries to machine code for a single-core can remove iterator and control overhead for significant performance gains. So far, systems that generate distributed programs only compile plans for single processors and stitch them together with messaging. In this paper, we describe an approach for translating query plans into distributed programs by targeting the partitioned global address space (PGAS) parallel programming model as an intermediate representation. This approach affords a natural adaptation of pipelining techniques used in singlecore query compilers and an overall simpler design. We adapt pipelined algorithms to PGAS languages, describe efficient data structures for PGAS query execution, and implement techniques for mitigating the overhead resulting from handling a multitude of fine-grained tasks. We evaluate Radish on graph benchmark and application workloads and find that it is 4× to 100× faster than Shark, a recent distributed query engine optimized for in-memory execution. Our work makes important first steps towards ensuring that query processing systems can benefit from future advances in parallel programming and co-mingle with state-of-the-art parallel programs. Prior systems have demonstrated orders of magnitude performance improvements over iterator based query processing by compiling plans for single processors [10, 23, 27, 34]. Frameworks that generate distributed programs only compile plans for individual processors and stitch them together with communication calls, retaining the iterator model [15, 33]. These systems have a common shortcoming: they depend directly on a single-node compiler (e.g. LLVM, JVM) to perform machine-level optimizations, but these compilers cannot reason about a distributed program. The alternative, which we explore in this paper, is to generate programs for a partitioned global address space (PGAS) language (e.g., Chapel [11] or X10 [13]), then compile and execute these distributed programs to evaluate the query. A key feature of these languages is a partition, which is a region of memory local to a particular processor and far from other processors. Consider this query with a join and multiplication:
The implementation of scalable synchronized data structures is notoriously difficult. Recent work... more The implementation of scalable synchronized data structures is notoriously difficult. Recent work in shared-memory multicores introduced a new synchronization paradigm called flat combining that allows many concurrent accessors to cooperate efficiently to reduce contention on shared locks. In this work we introduce this paradigm to a domain where reducing communication is paramount: distributed memory systems. We implement a flat combining framework for Grappa, a latency-tolerant PGAS runtime, and show how it can be used to implement synchronized global data structures. Even using simple locking schemes, we find that these flat-combining data structures scale out to 64 nodes with 2x-100x improvement in throughput. We also demonstrate that this translates to application performance via two simple graph analysis kernels. The higher communication cost and structured concurrency of Grappa lead to a new form of distributed flat combining that drastically reduces the amount of communication necessary for maintaining global sequential consistency.
We present Grappa, a modern take on software distributed shared memory (DSM) for in-memory data-i... more We present Grappa, a modern take on software distributed shared memory (DSM) for in-memory data-intensive applications. Grappa enables users to program a cluster as if it were a single, large, non-uniform memory access (NUMA) machine. Performance scales up even for applications that have poor locality and input-dependent load distribution. Grappa addresses deficiencies of previous DSM systems by exploiting application parallelism, trading off latency for throughput. We evaluate Grappa with an in-memory MapReduce framework (10⇥ faster than Spark [74]); a vertex-centric framework inspired by GraphLab (1.33⇥ faster than native GraphLab [48]); and a relational query execution engine (12.5⇥ faster than Shark [31]). All these frameworks required only 60-690 lines of Grappa code.
Grappa is a runtime system for commodity clusters of multicore computers that presents a massivel... more Grappa is a runtime system for commodity clusters of multicore computers that presents a massively parallel, single address space abstraction to applications. Grappa's purpose is to enable scalable performance of irregular parallel applications, such as branch and bound optimization, SPICE circuit simulation, and graph processing. Poor data locality, imbalanced parallel work and complex communication patterns make scaling these applications difficult. Grappa serves both as a C++ user library and as a foundation for higher level languages. Grappa tolerates delays to remote memory by multiplexing thousands of lightweight workers to each processor core, balances load via fine-grained distributed work-stealing, increases communication throughput by aggregating smaller data requests into large ones, and provides efficient synchronization and remote operations. We present a description of the Grappa system and an evaluation of its performance on microbenchmarks.
Dissatisfaction with relational databases for large-scale graph processing has motivated a new cl... more Dissatisfaction with relational databases for large-scale graph processing has motivated a new class of graph databases that offer fast graph processing but sacrifice the ability to express basic relational idioms. However, we hypothesize that the performance benefits amount to implementation details, not a fundamental limitation of the relational model. To evaluate this hypothesis, we are exploring code-generation to produce fast in-memory algorithms and data structures for graph patterns that are inaccessible to conventional relational optimizers. In this paper, we present preliminary results for this approach on path-counting queries, which includes triangle counting as a special case. We compile Datalog queries into main-memory pipelined hash-join plans in C++, and show that the resulting programs easily outperform PostgreSQL on real graphs with different degrees of skew. We then produce analogous parallel programs for Grappa, a runtime system for distributed memory architectures. Grappa is a good target for building a parallel query system as its shared memory programming model and communication mechanisms provide productivity and performance when building communication-intensive applications. Our experiments suggest that Grappa programs using hash joins have competitive performance with queries executed on Greenplum, a commercial parallel database. We find preliminary evidence that a code generation approach simplifies the design of a query engine for graph analysis and improves performance over conventional relational databases.
In this demo proposal, we describe REEF, a framework that makes it easy to implement scalable, fa... more In this demo proposal, we describe REEF, a framework that makes it easy to implement scalable, fault-tolerant runtime environments for a range of computational models. We will demonstrate diverse workloads, including extract-transform-load MapReduce jobs, iterative machine learning algorithms, and ad-hoc declarative query processing. At its core, REEF builds atop YARN (Apache Hadoop 2's resource manager) to provide retainable hardware resources with lifetimes that are decoupled from those of computational tasks. This allows us to build persistent (cross-job) caches and cluster-wide services, but, more importantly, supports high-performance iterative graph processing and machine learning algorithms. Unlike existing systems, REEF aims for composability of jobs across computational models, providing significant performance and usability gains, even with legacy code. REEF includes a library of interoperable data management primitives optimized for communication and data movement (wh...
This study sought to determine what aspect of divorce was most likely associated with an increase... more This study sought to determine what aspect of divorce was most likely associated with an increase in criminally delinquent behavior among adolescent children under the age of eighteen years. The researcher utilized the National Longitudinal Survey of Youth 1997, NLSY97, as the primary research tool which was a ten year survey of nearly 9,000 youths and household members. The researcher
Uploads
Papers by Brandon Myers