DMCA
Compiled Plans for In-Memory Path-Counting Queries (2013)
| Citations: | 1 - 0 self |
BibTeX
@MISC{Myers13compiledplans,
author = {Brandon Myers and Jeremy Hyrkas and Daniel Halperin and Bill Howe},
title = {Compiled Plans for In-Memory Path-Counting Queries},
year = {2013}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
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.
Keyphrases
in-memory path-counting query competitive performance good target query engine preliminary evidence graph processing new class implementation detail shared memory conventional relational database different degree large-scale graph process-ing special case code generation approach basic relational idiom graph database relational model analogous parallel program relational database fundamental limitation distributed memory architecture data structure real graph parallel query system hash join graph pattern communication-intensive application conventional relational optimizers fast in-memory algorithm commercial parallel database preliminary result path-counting query grappa program hash-join plan runtime system datalog query communication mechanism graph analysis triangle counting performance benefit amount
