We propose a framework to model on-line resource management problems based
on an on-line version of positive linear programming. We consider both min
cost problems and max benefit problems and propose logarithmic competitive
algorithms that are optimal up to a constant factor.
The proposed framework provides a general methodology that applies to a wid
e class of on-line problems: shop scheduling, packet routing, and in genera
l a class of packing and assignment problems. Previously studied problems a
s on-line multiprocessor scheduling and on-line virtual circuit routing can
also be modeled within this framework.