Issasquah, WA — (press release excerpt) — McObject's new white paper, “Main Memory vs. RAM-Disk Databases: A Linux-based Comparison” addresses performance and availability issues of different approaches to database management in embedded systems running on Linux.

It stands to reason that accessing data from memory will be faster than from physical media. A new type of database management system, the main memory database (MMDB), claims breakthrough performance and availability via memory-only processing. But doesn't database caching achieve the same result? And if complete elimination of disk access is the goal, why not deploy a traditional database on a RAM-disk, which creates a file system in memory?

McObject's benchmark tests the company's MMDB against a widely used embedded database, which is used in both traditional (disk-based) and RAM-disk modes. Deployment on RAM-disk boosts the traditional database's performance by as much as 74 percent, but still lags the memory-only database in this test, performed on Red Hat Linux version 6.2.

Fundamental architectural differences explain the disparity. Performance overhead in disk-based databases includes data transfer and duplication, unneeded recovery logic and, ironically, caching functions intended to avoid disk access. None of these can be “turned off” when the database is deployed in memory. In contrast, the MMDB is streamlined to provide only the functions required for memory-based data management in real-time embedded systems.

 
This article was originally published on LinuxDevices.com and has been donated to the open source community by QuinStreet Inc. Please visit LinuxToday.com for up-to-date news and articles about Linux and open source.