AUTOMATED META-CONTROL FOR ADAPTABLE REAL-TIME SOFTWARE

The software meta-controller is an online agent responsible for dynami cally adapting an application's software configuration, e.g. altering operational modes and migrating tasks, to best accommodate varying run time circumstances. In distributed real-time applications such adaptat ions must be carried out in a manner which maintains the schedulabilit y of all critical tasks while maximizing some notion of system value f or all other tasks. For large-scale real-time applications, considerin g all possible adaptations at the task-level is computationally intrac table. This paper presents an automated aggregate approach to software meta-control, appropriate for large-scale distributed real-time syste ms. The aggregate automated meta-control problem is !;till NP-hard, bu t it has very practical approximate solutions. Introduced, here, are t wo very-effective approximation algorithms, QDP and G(2), with very re asonable polynomial time complexity. Both algorithms also provide us w ith upper bounds for optimum system values, useful for deriving absolu te, albeit somewhat pessimistic, measures of actual performance. Exten sive Monte Carlo analysis is used to illustrate that expected performa nce for both algorithms is generally suboptimal by no more than a few percent. Our flexible software meta-control model is also shown to be readily applied to a wide range of time-sensitive applications.