An adaptive peer-to-peer network for distributed caching of OLAP results

Abstract

A b s t r a c tPeer-to-Peer (P2P) systems are becoming increasingly popular as they enable users to exchange digital information by participating in complex networks. Such systems are inexpensive, easy to use, highly scalable and do not require central administration. Despite their advantages, however, limited work has been done on employing database systems on top of P2P networks.Here we propose the PeerOLAP architecture for supporting On-Line Analytical Processing queries. A large number of low-end clients, each containing a cache with the most useful results, are connected through an arbitrary P2P network. If a query cannot be answered locally (i.e. by using the cache contents of the computer where it is issued), it is propagated through the network until a peer that has cached the answer is found. An answer may also be constructed by partial results from many peers. Thus PeerOLAP acts as a large distributed cache, which amplifies the benefits of traditional client-side caching. The system is fully distributed and can reconfigure itself on-the-fiy in order to decrease the query cost for the observed workload. This paper describes the core components of PeerOLAP and presents our results both from simulation and a prototype installation running on geographically remote peers.

Brown Dwarf: A fully-distributed, fault-tolerant data warehousing system

“…In [18], the authors consider a number of DWs and peers, forming an unstructured P2P overlay for caching OLAP views. Views are divided in chunks and peers retrieve cached chunks from the network and the DW if needed.…”

“…Furthermore, the need to keep large volumes of historical data online and ensure their availability and fast access even under heavy workload dictates a continuous investment in hardware, electrical power and software maintenance. Some works in the field propose distributed warehousing systems [18,4,2], but the warehouse and its aggregation, update and querying functionality remain centralized. Recently, effort has been made to distribute the data warehouse itself by applying techniques from the field of Peer-to-Peer (P2P) computing [12], but with no a priori consideration for group-by queries.…”

The Piazza Peer Data Management System

“…In [35], the authors study a variant of the data placement problem and focus on intelligently caching and reusing queries in an OLAP environment. Recently, [36] described local relational models as a formalism for mediating between different peers in a PDMS and a sound and complete algorithm for answering queries using the formalism, but do not describe the expressive power of the formalism compared to previous ones in the data integration literature.…”

Distributing the power of OLAP

“…Centralized warehousing systems offer indexing schemes for storing and efficiently querying data cubes (e.g., [5,6]), but only work in controlled environments, failing to scale. Some Table 1: A sample fact table with three dimensions and one measure DIM1 DIM2 DIM3 Measure S1 C2 P2 $70 S1 C3 P1 $40 S2 C1 P1 $90 S2 C1 P2 $50 works in the field propose distributed warehousing systems (e.g., [1,4]), but the warehouse and its aggregation, update and querying functionality remain centralized. Recently, effort has been made to distribute the data warehouse itself by applying techniques from the field of P2P computing [3], but with no a priori consideration for group-by queries.…”