developerWorks: Real-time Resource Management

"A real-time system must provide a response to a request within a specified period of time. Most often, requests are recurring, with requests made on a periodic basis for a common set of services. Real-time systems provide these periodic services, which produce output upon request to control systems. Examples might include a digital control for an antilock braking system, a digital autopilot to maintain steady, level flight for an aircraft, or video stream compression and decompression.

"For systems where requests are not periodic, the system can be analyzed to determine a worst-case interarrival rate for service requests. For example, in a fault-handling system, an upper bound on the fault rate that the system can handle can and should be established. In normal operation, the fault rate might be zero, but the assumption of a maximum rate ensures that computing resources are held in reserve..."

Complete Story

Related Stories:

• SearchOpenSource: Linux Virtual Address Randomization and Impacting Buffer Overflows(Nov 18, 2005)
• NewsForge: Linux Wireless Freedom with OpenWrt(Aug 19, 2005)
• Linux.com: CLI Magic: Use Extended Attributes for Better File Management(Jun 21, 2005)
• NewsForge: OSCommerce in Action(Jun 03, 2005)