About ScimoreDB Grid Technology

Scimore engineers have, over the last 5 years, developed a data streaming grid technology.

The core idea behind the technology is that a user, using a simple distributed query language (DQL), can manage unit-of-process (UOP) and data streams in a distributed environment.

The data streams are streams of tuples/rows, which are used for interfacing between UOP. Each UOP exposes methods to control data flow between UOPs and UOP distribution in the Grid. Queries may be written using the DQL language, which defines UOP needed for the query, UOP query parameters and the relations between UOPs.

The focus for our grid technology is data-driven distributed solutions for large data volumes, with high transaction throughput. We see grid technology as the ideal platform for resource-intensive, data-driven applications.

The core features of the Scimore Grid technology are:

• Distributed Query management
  • Distributed Query optimization
  • Automatic Query distribution
• Distributed Transaction Management
  • Distributed Deadlock preventation
  • ACID transaction properties
• Concurrency control
• A simple Distributed Query Language (DQL)
• Health and activity monitoring
• Seamless embedding of external data driven products

ScimoreDB is implemented using Scimore Grid Techonology. We designed UOPs optimized for relational databases. At the core is a UOP for accessing tables. On top is the UOP's implementing relational algebra.

We then developed a translater and query optimizer, which translates SQL language to DQL.

It sounds simple, but it is not. It was all about getting the core DQL design correct, then everything did fall into place. In total we have 200,000 lines of c++ code for ScimoreDB - including the Scimore Grid Technology. Learn more ».

ScimoreDB requirements

• XP / Vista / Windows Server 2000 / Windows Server 2003 / Windows Server 2008
• 10 Mb RAM
• 16 MB disk space