Abstract
As computational power and storage capacity increase, processing and analyzing large volumes of data play an increasingly important part in many domains of scientific research. Typical examples of large scientific datasets include long running simulations of time-dependent phenomena that periodically generate snapshots of their state (e.g. hydrodynamics and chemical transport simulation for estimating pollution impact on water bodies [4, 6, 20], magnetohydrodynamics simulation of planetary magnetospheres [32], simulation of a flame sweeping through a volume [28], airplane wake simulations [21]), archives of raw and processed remote sensing data (e.g. AVHRR [25], Thematic Mapper [17], MODIS [22]), and archives of medical images (e.g. confocal light microscopy, CT imaging, MRI, sonography).
These datasets are usually multi-dimensional. The data dimensions can be spatial coordinates, time, or experimental conditions such as temperature, velocity or magnetic field. The importance of such datasets has been recognized by several database research groups and vendors, and several systems have been developed for managing and/or visualizing them [2, 7, 14, 19, 26, 27, 29, 31].
These systems, however, focus on lineage management, retrieval and visualization of multi-dimensional datasets. They provide little or no support for analyzing or processing these datasets -- the assumption is that this is too application-specific to warrant common support. As a result, applications that process these datasets are usually decoupled from data storage and management, resulting in inefficiency due to copying and loss of locality. Furthermore, every application developer has to implement complex support for managing and scheduling the processing.
Over the past three years, we have been working with several scientific research groups to understand the processing requirements for such applications [1, 5, 6, 10, 18, 23, 24, 28]. Our study of a large set of applications indicates that the processing for such datasets is often highly stylized and shares several important characteristics. Usually, both the input dataset as well as the result being computed have underlying multi-dimensional grids, and queries into the dataset are in the form of ranges within each dimension of the grid. The basic processing step usually consists of transforming individual input items, mapping the transformed items to the output grid and computing output items by aggregating, in some way, all the transformed input items mapped to the corresponding grid point. For example, remote-sensing earth images are often generated by performing atmospheric correction on several days worth of raw telemetry data, mapping all the data to a latitude-longitude grid and selecting those measurements that provide the clearest view.
In this paper, we present T2, a customizable parallel database that integrates storage, retrieval and processing of multi-dimensional datasets. T2 provides support for many operations including index generation, data retrieval, memory management, scheduling of processing across a parallel machine and user interaction. It achieves its primary advantage from the ability to seamlessly integrate data retrieval and processing for a wide variety of applications and from the ability to maintain and process multiple datasets with different underlying grids. Most other systems for multi-dimensional data have focused on uniformly distributed datasets, such as images, maps, and dense multi-dimensional arrays. Many real datasets, however, are non-uniform or unstructured. For example, satellite data is a two dimensional strip that is embedded in a three dimensional space; water contamination studies use unstructured meshes to selectively simulate regions and so on. T2 can handle both uniform and non-uniform datasets.
T2 has been developed as a set of modular services. Since its structure mirrors that of a wide variety of applications, T2 is easy to customize for different types of processing. To build a version of T2 customized for a particular application, a user has to provide functions to pre-process the input data, map input data to elements in the output data, and aggregate multiple input data items that map to the same output element.
T2 presents a uniform interface to the end users (the clients of the database system). Users specify the dataset(s) of interest, a region of interest within the dataset(s), and the desired format and resolution of the output. In addition, they select the mapping and aggregation functions to be used. T2 analyzes the user request, builds a suitable plan to retrieve and process the datasets, executes the plan and presents the results in the desired format.
In Section 2 we first present several motivating applications and illustrate their common structure. Section 3 then presents an overview of T2, including its distinguishing features and a running example. Section 4 describes each database service in some detail. An example of how to customize several of the database services for a particular application is given in Section 5. T2 is a system in evolution. We conclude in Section 6 with a description of the current status of both the T2 design and the implementation of various applications with T2.
