ExtenDB Parallel Server Standard Edition 1.1.1

ExtenDB Parallel Server project is a loosely-coupled clustered database system that provides the heart of your Business Intelligence infrastructure at a very low cost.
It includes intelligence to maximize parallelization over multiple servers, achieving near linear scalability for fast query response times.
At the same time, by utilizing reliable commodity hardware and proven open source technology, the ExtenDB solution costs a fraction of that of traditional vendors.
The ExtenDB Parallel Server is a high-performance, highly scalable, enterprise-class, clustered database solution for data warehousing and data marts. It leverages open source databases running on PC-based commodity hardware (nodes), and intelligently coordinates work between them.
All of the nodes are interconnected, typically via Gigabit Ethernet and logically, in the ExtenDB Parallel Server software, via the ExtenDB Network Messaging Layer.
All the Node Storage Subsystems connected to each node of an ExtenDB Cluster are referred to collectively as the Distributed Data Domain. The Distributed Data Domain is the repository for all the Business Intelligence data that makes up the ExtenDB database.
Each node has an assigned multi-threaded Node Agent that executes and manages a pool of connections to the underlying database, and executes the commands sent from the designated coordinator node.
The ExtenDB Parallel Server loads Business Intelligence data into the Distributed Data Domain using an intelligent partitioning engine. Simply put, a hash value is used to determine the destination of a particular row. The column used in computing the hash value can also be chosen by the Database Administrator (DBA) as part of the overall schema strategy.
In addition, other tables may be designated as replicated lookup tables, which results in an exact copy of the table being distributed to all nodes in the cluster. These are generally relatively small tables.
The ExtenDB Parallel Server’s cost-based Optimizer uses this partitioning and replication information in determining an optimal strategy for processing queries by maximizing parallelism. In choosing an optimal query plan, it also avoids shipping rows of data on the network if possible. Hence, it is the geography of partitioned customer data in the Distributed Data Domain determined by the intelligent parallelization engine that lies at the core of the blistering performance and enterprise scalability of the ExtenDB Parallel Server.
Version restrictions:
- Limited to 4 nodes