 |
|
 |
 |
 |
 |
 |
 |
 |
 |
TruCQ Continuous Analytics Engine
You might have heard of zero-latency stream processing technology. You definitely know about relational databases. Truviso harnesses all the time-tested methodologies of relational database systems (and the standards that make them work) and has combined them with real-time data stream processing techniques. The result is a Stream-Relational system that processes and queries data before it’s stored, providing immediately updated analytics, alerts or triggers as soon as data hits the system. Continuous Analytics eliminates the traditional 12-24 hour processing lag times associated with batch loading data warehouses and data marts.
As data is pulled or pushed into the TruCQ Continuous Analytics engine from relational tables, files or streaming sources, it is correlated and queried. There’s no need to store data, then select data, then read and query the data later. Queries run continuously in parallel as data comes in, resulting in always-up-to-date answers.
Adaptive Query Processor
Core to Truviso's architectural design is the Adaptive Query Processor which intelligently, and on-the-fly, "folds" the processing steps of multiple different queries into a shared global query that effectively executes as a single query in the system. The result is massive query scalability that enables thousands of concurrent queries to be run continuously against incoming streams of data. The system has been benchmarked processing data streams in excess of 375,000 records/second on a single quad processor server with only 3 disks.
Each time queries are run the Optimizer produces an execution plan. If the query
is a snapshot in time, the TruCQ engine evaluates the execution plan individually,
as would be the case in a traditional RDBMS. If the query produces a continuous
dataset, as any streaming data source would produce, the associated execution
plan is evaluated using a control queue that queries based on standard time increments
or numbers of records.
|
 |
 |
 |
 |
 |
 |
 |
 |
 |
|
 |
|