Psydex - Technology - MPP (Massive Parallel Processing)


	TECHNOLOGY • AG (Analytics Grid) • AIM Indexes • Massive Parallel Processing • Universal Modeling • Predictive Analytics • PEP (Event Processor) • Partner Integration • Sources • Query Language • AG APIs BIG DATA - BIG THINKING The World will produce a ZetaByte of new data this year with 80% in the form of UnStructured emails, documents, text messages and social media conversations. Traditional approaches to analyzing and processing data do not work for the Real Time World. White Papers

Psydex is designed around a shared-nothing, MPP architecture and easily scales using commodity hardware. Each blue square in the diagram below represents a 4+ Gig AIM Index Instance (Time Slice) of a particular source. A source might have thousands of instances that function seamlessly as a single index. Federators communicate with Index Instances to intelligently route requests and aggregate, collate and fuse results.

Technology MPP

Communications between processes is based on a proprietary, compressed serialization protocol for optimized performance

Federation
Psydex AG employs a very specialized and advanced Federation/Routing model to satisfy the unique demands of seamlessly processing queries against large volumes of rapidly moving and historical data in real time. Federators facilitate routing of requests and the aggregation, collation and fusion of results from distributed nodes and processes. Roll up and aggregation of timeslots occurs at Query Time based on the smallest possible time slice.

Grid Management
Grid services are managed by Grid Managers which track and maintain overall health of the Grid. Grid Managers orchestrate the loading of Indexes and monitor the health and status of Nodes and Index Instances. Grid Services and Index Instances are instantiated as indexe instances surpass thresholds or exceptions are detected.