Scheduling heterogeneous multimedia servers: different QoS for hard, soft and non real-time clients
Euromicro-RTS'00 Proceedings of the 12th Euromicro conference on Real-time systems
| Hi-index | 0.00 |
ATM has offered QoS guarantees for nearly a decade, but the push is now for IP-based solutions. IP is ubiquitous in today's congested networks. Applications are more complex, users are more demanding, standards bodies are more receptive, and technology is more sophisticated. All this has focused attention on ways to add QoS to IP networks without exorbitant cost. But with progress has come some confusion. Telecommunications carriers and service providers are already enticing customers by offering two or three distinct classes of service over their IP networks. Vendors are beginning to ship QoS-capable hardware and software. ATM is firmly established in the Internet's core. A lot has happened in a short time, which means that users and providers must be aware of where things are going and what the various QoS technologies can actually do. Some QoS mechanisms deliver strict, absolute performance guarantees. Others merely offer assurances that one service class will take priority over another when resources are scarce