LEXST Cluster is a simple database management system which provides both horizontal partitioning and vertical partitioning. It's designed to provide high performance when querying a huge table which exceeds 1 million rows. Lexst scaling is linear, the biggest cluster allows 20,000 nodes included, and each node can have its own standby node. Lexst Database cluster was published in 2008 in China.

LEXST is based on a Hash table

Hash table

In computer science, a hash table or hash map is a data structure that uses a hash function to map identifying values, known as keys , to their associated values . Thus, a hash table implements an associative array...

 searching mechanism, written in Java code and runs in a Java

Java Development Kit

The Java Development Kit is an Oracle Corporation product aimed at Java developers. Since the introduction of Java, it has been by far the most widely used Java SDK. On 17 November 2006, Sun announced that it would be released under the GNU General Public License , thus making it free software...

 environment.

Architecture

Lexst Database Cluster has a few new concepts and also disadvantages, which shall be noticed by users and learners.

Replication

Lexst Database Cluster uses asynchronous replication, when a standby node is introduced into the node, according to the role assigned by the Master Node, it would replicate the data from the main node automatically in "good time".

Horizontal Partitioning

Data within Lexst tables is automatically partitioned across all of the data nodes. This is done based on a hashing algorithm based on the PRIMARY KEY on the table, and is nearly transparent to the end application, users can also define their own partitioning schemes, usually by specifying the space limit on each node.

Fault Tolerance

If a node fails, a standby node would automatically switch in, and play the role. It guarantees the high availability

High availability

High availability is a system design approach and associated service implementation that ensures a prearranged level of operational performance will be met during a contractual measurement period....

.

Variable-width columns

For 'char' type columns, the length of the column can be specified as "0". In this case, it can store as much as the user wants. This saves the storage space and allows rich media files to be stored here.

Implementation

There are 6 kinds of Node in a cluster:

• Master Node: Master node is the nerve center, it controls and monitors all the nodes, administrator manages the users and setting here.

• Log Node: Records every activities in the cluster.

• User Node: Informations of all users are stored here.

• Data Nodes: Nodes store the data.

• Call Node: Call nodes execute the queries.

• Remote Node: Allow users or administrator to log into system remotely.

Usually, one node shall run on one physical server, however, it's even allowed that all 6 kinds of node to be installed on one server.

Limitations

Lexst Database Cluster is a simple database system, it currently doesn't support full-text search, users can only insert rows, delete rows and update rows. Transaction has not been added into the system, although there is a plan in future.

Currently, only Windows versions can be provided. Java Environment is also a must for running of Lexst Database cluster.

Hardware Requirements

Minimum system requirements are as follows (for each node) 3 Machines Minimum

OS: Windows 2000/xp/vista

CPU: Intel/AMD x86

Memory: 128MB RAM

HDD: 3GB

Network: 1+ nodes (Standard Ethernet - TCP/IP)

Reference manual

• Administrator's Guide for Lexst Database Cluster
• API Manual for LEXST
• Mysql cluster and Lexst_cluster comparison