Distributing the power of OLAP

Abstract

In this paper we present the Brown Dwarf, a distributed system designed to efficiently store, query and update multidimensional data over an unstructured Peer-to-Peer overlay, without the use of any proprietary tool. Brown Dwarf manages to distribute a highly effective centralized structure among peers on-the-fly. Both point and aggregate queries are then naturally answered on-line through cooperating nodes that hold parts of a fully or partially materialized data cube. Updates are also performed on-line, eliminating the usually costly over-night process. Our initial evaluation on an actual testbed proves that Brown Dwarf manages to distribute the structure across the overlay nodes incurring only a small storage overhead compared to the centralized algorithm. Moreover, it accelerates cube creation up to 5 times and querying up to several tens of times by exploiting the capabilities of the available network nodes working in parallel.

Pocket-cubes, bringing multidimensional data views to mobile platforms

“…Nowadays, the drawbacks of the computation and compression processes of a data cube are well known [13,14,19]: the computation time, the huge space of memory required to process a cube, the huge space needed to store the data (even compressed) of the information as well as to ensure the semantic of the OLAP operations, queries types and incremental maintenance of the information of the data cube. On another hand it is recognized the most frequent problems when we intent to use the information of a data cube in mobile environments, namely the ones related to frequent disconnections, power and screen size limitations of the mobile device, and the space limitation of the storage in the mobile platform, which difficult the storage of all the information of the data cube.…”

“…Michalarias and Omelchenko [15] presented another infrastructure for the dissemination of multidimensional information using the Dwarf method [10] to compute and compress the aggregations to be used in mobile environments. Doka et al [14] approached the Dwarf representation of a data cube in a mobile environment. They generate the Dwarf structure and store it in several servers, using an online approach to access to the information of mobile devices.…”

Efficient updates for a shared nothing analytics platform

“…Brown Dwarf (BD for short) [7] is a system that distributes Dwarf over a network of interconnected nodes onthe-fly, in a way that all queries that were originally answered through the centralized structure are now distributed over a P2P overlay. The general approach is that each vertex of the dwarf graph (henceforth termed as dwarf node) is designated with a unique ID (UID) and assigned to a network node.…”

“…In this work, we do not explicitly deal with loadbalancing nor fault tolerance, as we believe it is orthogonal. An initial approach to tackle such issues is discussed in [7].…”

Brown dwarf

“…We have created an always-on, distributed data warehousing system, the Brown Dwarf (BD) [3], where geographically spanned users, without the use of any proprietary tool, can share and query information. Our system employs a robust and efficient adaptive replication scheme, perceptive both to workload skew and node churn using only local load measurements and overlay knowledge.…”