6 Ways to Business Intelligence System Scalability
This Information Builders video explains how the WebFocus business intelligence platform builds in scalability so that hundreds of thousands of concurrent users can be served. Six ways of ensuring scalability are discussed. The concepts are:
- Persistence vs. Non-persistence (use non-persistent connections)
- Multi-threading (multi-threaded solutions allows multiple users to share the same resources concurrently)
- Native Data Adapters (ODBC not good enough and leads to an under-utilized DBMS server and overworked BI server. Use native data adapter to allow database server to take the load.)
- Superior Data Manipulation (Typical SQL from most business intelligence tools require multiple passes through the data. Use advanced data manipulation for more efficient reading, aggregation and manipulation of data.)
- Multi-tier Configuration (Many BI solutions have limited support for hardware platforms. Most data reside both on large mainframes as well as middle-tier machines. Beware of transfering more data than needed wasting valuable network resources. Try to process data on or close to the platform where the data resides in order to minimize network usage and optimize performance.)
- Clustering, Load Management & Autonomic Functions (Link servers in a chain and automatically and intelligently distribute traffic across servers. Use autonomic functionality to all the server to wake up additional machines as the traffic requires.