Legacy Real-Time Applications in a Reservation-Based System

A remarkable research activity has been carried out in the past few years to support real-time applications with appropriate scheduling solutions. Unfortunately, most of such techniques can be used only if real-time applications use a specialized API, and if some important information (such as the worst-case execution-time) are known a priori. In this paper, we present a novel technique, the legacy feedback scheduler (LFS), for a class of legacy applications that need the support of a real-time scheduler but are not written using a specialized API and have unknown or varying execution requirements. The approach is based on the combination of a resource reservation scheduler and a feedback-based adaptation mechanism for identifying the correct scheduling parameters.