With the temporal audio-visual and sensory information in various distributed multimedia applications, the provision of end-to-end quality of service (QoS) guarantees is a major acceptance factor for these applications.

We designed and implemented a set of end-point services and protocols for provision of end-to-end QoS guarantees which rely on a new resource model and a robust time-variant QoS management, accessible to any application. The resource model incorporates, in addition to a resource scheduler, a new component, called the resource broker, which provides a reservation capability to use a shared resource such as CPU, network or memory. The resource brokers are intermediary resource management agents which provide a better, more predictable and fine granularity control of resources to the application. The QoS management consists of a set of distributed QoS management entities, called `QoS brokers', which translate and negotiate/renegotiate end-to-end QoS parameters, and coordinate resource reservation according to end-to-end time-variant QoS requirements such as jitter, synchronization skew, end-to-end delays to establish a QoS contract.

This talk presents the time-variant QoS management architecture, design, implementation and results of QoS-aware CPU, memory and communication servers comprising our QoS management platform, and various lessons we learned during the design and implementation of the QoS brokerage and enforcement services and protocols.