IBM solidDB: In-Memory Database Optimized for Extreme Speed and Availability.

Abstract

A relational in-memory database, IBM solidDB is used worldwide for its ability to deliver extreme speed and availability. As the name implies, an in-memory database resides entirely in main memory rather than on disk, making data access an order of several magnitudes faster than with conventional, disk-based databases. Part of that leap is due to the fact that RAM simply provides faster data access than hard disk drives. But solidDB also has data structures and access methods specifically designed for storing, searching, and processing data in main memory. As a result, it outperforms ordinary diskbased databases even when the latter have data fully cached in memory. Some databases deliver low latency but cannot handle large numbers of transactions or concurrent sessions. IBM solidDB provides throughput measured in the range of hundreds of thousands to million of transactions per second while consistently achieving response times (or latency) measured in microseconds. This article explores the structural differences between in-memory and disk-based databases, and how solidDB works to deliver extreme speed. 1