A hierarchial CPU scheduler for multimedia operating systems
OSDI '96 Proceedings of the second USENIX symposium on Operating systems design and implementation
Adaptive algorithms for managing a distributed data processing workload
IBM Systems Journal
Apache: the definitive guide
Resource containers: a new facility for resource management in server systems
OSDI '99 Proceedings of the third symposium on Operating systems design and implementation
Defending against denial of service attacks in Scout
OSDI '99 Proceedings of the third symposium on Operating systems design and implementation
Self-paging in the Nemesis operating system
OSDI '99 Proceedings of the third symposium on Operating systems design and implementation
Experience with processes and monitors in Mesa
Communications of the ACM
Supporting Service Level Agreements on IP Networks
Supporting Service Level Agreements on IP Networks
Proportional Share Scheduling of Operating System Services for Real-Time Applications
RTSS '98 Proceedings of the IEEE Real-Time Systems Symposium
The eclipse operating system: providing quality of service via reservation domains
ATEC '98 Proceedings of the annual conference on USENIX Annual Technical Conference
Retrofitting quality of service into a time-sharing operating system
ATEC '99 Proceedings of the annual conference on USENIX Annual Technical Conference
Managing energy and server resources in hosting centers
SOSP '01 Proceedings of the eighteenth ACM symposium on Operating systems principles
Isolated Dynamic Clusters for Web Hosting
Euro-Par '01 Proceedings of the 7th International Euro-Par Conference Manchester on Parallel Processing
Kernel Mechanisms for Service Differentiation in Overloaded Web Servers
Proceedings of the General Track: 2002 USENIX Annual Technical Conference
Kernel Support for Open QoS-Aware Computing
RTAS '03 Proceedings of the The 9th IEEE Real-Time and Embedded Technology and Applications Symposium
Scale and performance in the Denali isolation kernel
ACM SIGOPS Operating Systems Review - OSDI '02: Proceedings of the 5th symposium on Operating systems design and implementation
Sharc: Managing CPU and Network Bandwidth in Shared Clusters
IEEE Transactions on Parallel and Distributed Systems
Stateful distributed interposition
ACM Transactions on Computer Systems (TOCS)
Scale and performance in the Denali isolation kernel
OSDI '02 Proceedings of the 5th symposium on Operating systems design and implementationCopyright restrictions prevent ACM from being able to make the PDFs for this conference available for downloading
Using runtime paths for macroanalysis
HOTOS'03 Proceedings of the 9th conference on Hot Topics in Operating Systems - Volume 9
Path-based faliure and evolution management
NSDI'04 Proceedings of the 1st conference on Symposium on Networked Systems Design and Implementation - Volume 1
Model-based resource provisioning in a web service utility
USITS'03 Proceedings of the 4th conference on USENIX Symposium on Internet Technologies and Systems - Volume 4
Reliable framework for RFID devices
Proceedings of the 5th Middleware doctoral symposium
| Hi-index | 0.00 |
Modern server operating systems (OS's) do not address the issue of interference between competing applications. This deficiency is a major road-block for Internet and Application Service Providers who want to multiplex server resources among their business clients. To insulate applications from each other, we introduce Virtual Services (VSs). Besides providing per-service resource budgets, VSs drastically reduce cross-service interference in the presence of shared backend services, such as databases and name services. VSs provide dynamic per-service resource partitioning and management in a manner completely transparent to applications. To accomplish this goal, we introduce a kernel-based work classification mechanism called gates. Gates track work that propagates from one service to another and are configured by the system administrator via simple rules. They automate the binding of processes and sockets to VSs, and ensure that any work done on behalf of a VS, even if it is done by shared services, is charged to the resource budget of the VS that requested it. Using our experimental Linux 2.0.36-based implementation we applied them effectively to co-hosted Web servers. Thus, nearly eliminating performance interference between the co-hosted sites.