Search for: All records

The physical data layout significantly impacts performance when database systems access cold data. In addition to the traditional row store and column store designs, recent research proposes to partition tables hierarchically, starting from either horizontal or vertical partitions and then determining the best partitioning strategy on the other dimension independently for each partition. All these partitioning strategies naturally produce rectangular partitions. Coarse-grained rectangular partitioning reads unnecessary data when a table cannot be partitioned along one dimension for all queries. Fine-grained rectangular partitioning produces many small partitions which negatively impacts I/O performance and possibly introduces a high tuple reconstruction overhead. This paper introduces Jigsaw, a system that employs a novel partitioning strategy that creates partitions with arbitrary shapes, which we refer to as irregular partitions. The traditional tuple-at-a-time or operator-at-a-time query processing models cannot fully leverage the advantages of irregular partitioning, because they may repeatedly read a partition due to its irregular shape. Jigsaw introduces a partition-at-a-time evaluation strategy to avoid repeated accesses to an irregular partition. We implement and evaluate Jigsaw on the HAP and TPC-H benchmarks and find that irregular partitioning is up to 4.2× faster than a columnar layout for moderately selective queries. Compared with the columnar layout, irregular partitioning only transfers 21% of the data to complete the same query. 

Many applications are increasingly becoming I/O-bound. To improve scalability, analytical models of parallel I/O performance are often consulted to determine possible I/O optimizations. However, I/O performance modeling has predominantly focused on applications that directly issue I/O requests to a parallel file system or a local storage device. These I/O models are not directly usable by applications that access data through standardized I/O libraries, such as HDF5, FITS, and NetCDF, because a single I/O request to an object can trigger a cascade of I/O operations to different storage blocks. The I/O performance characteristics of applications that rely on these libraries is a complex function of the underlying data storage model, user-configurable parameters and object-level access patterns. As a consequence, I/O optimization is predominantly an ad-hoc process that is performed by application developers, who are often domain scientists with limited desire to delve into nuances of the storage hierarchy of modern computers.This paper presents an analytical cost model to predict the end-to-end execution time of applications that perform I/O through established array management libraries. The paper focuses on the HDF5 and Zarr array libraries, as examples of I/O libraries with radically different storage models: HDF5 stores every object in one file, while Zarr creates multiple files to store different objects. We find that accessing array objects via these I/O libraries introduces new overheads and optimizations. Specifically, in addition to I/O time, it is crucial to model the cost of transforming data to a particular storage layout (memory copy cost), as well as model the benefit of accessing a software cache. We evaluate the model on real applications that process observations (neuroscience) and simulation results (plasma physics). The evaluation on three HPC clusters reveals that I/O accounts for as little as 10% of the execution time in some cases, and hence models that only focus on I/O performance cannot accurately capture the performance of applications that use standard array storage libraries. In parallel experiments, our model correctly predicts the fastest storage library between HDF5 and Zarr 94% of the time, in contrast with 70% of the time for a cutting-edge I/O model. 

Array management libraries, such as HDF5, Zarr, etc., depend on a complex software stack that consists of parallel I/O middleware (MPI-IO), POSIX-IO, and file systems. Components in the stack are interdependent, such that effort in tuning the parameters in these software libraries for optimal performance is non-trivial. On the other hand, it is challenging to choose an array management library based on the array configuration and access patterns. In this poster, we investigate the performance aspect of two array management libraries, i.e., HDF5 and Zarr, in the context of a neuroscience use case. We highlight the performance variability of HDF5 and Zarr in our preliminary results and discuss potential optimization strategies. 

Scientific data analysis pipelines face scalability bottlenecks when processing massive datasets that consist of millions of small files. Such datasets commonly arise in domains as diverse as detecting supernovae and post-processing computational fluid dynamics simulations. Furthermore, applications often use inference frameworks such as TensorFlow and PyTorch whose naive I/O methods exacerbate I/O bottlenecks. One solution is to use scientific file formats, such as HDF5 and FITS, to organize small arrays in one big file. However, storing everything in one file does not fully leverage the heterogeneous data storage capabilities of modern clusters. This paper presents Henosis, a system that intercepts data accesses inside the HDF5 library and transparently redirects I/O to the in-memory Redis object store or the disk-based TileDB array store. During this process, Henosis consolidates small arrays into bigger chunks and intelligently places them in data stores. A critical research aspect of Henosis is that it formulates object consolidation and data placement as a single optimization problem. Henosis carefully constructs a graph to capture the I/O activity of a workload and produces an initial solution to the optimization problem using graph partitioning. Henosis then refines the solution using a hill-climbing algorithm which migrates arrays between data stores to minimize I/O cost. The evaluation on two real scientific data analysis pipelines shows that consolidation with Henosis makes I/O 300× faster than directly reading small arrays from TileDB and 3.5× faster than workload-oblivious consolidation methods. Moreover, jointly optimizing consolidation and placement in Henosis makes I/O 1.7× faster than strategies that perform consolidation and placement independently. 

A ubiquitous problem in aggregating data across different experimental and observational data sources is a lack of software infrastructure that enables flexible and extensible standardization of data and metadata. To address this challenge, we developed HDMF, a hierarchical data modeling framework for modern science data standards. With HDMF, we separate the process of data standardization into three main components: (1) data modeling and specification, (2) data I/O and storage, and (3) data interaction and data APIs. To enable standards to support the complex requirements and varying use cases throughout the data life cycle, HDMF provides object mapping infrastructure to insulate and integrate these various components. This approach supports the flexible development of data standards and extensions, optimized storage backends, and data APIs, while allowing the other components of the data standards ecosystem to remain stable. To meet the demands of modern, large-scale science data, HDMF provides advanced data I/O functionality for iterative data write, lazy data load, and parallel I/O. It also supports optimization of data storage via support for chunking, compression, linking, and modular data storage. We demonstrate the application of HDMF in practice to design NWB 2.0, a modern data standard for collaborative science across the neurophysiology community.