| Wednesday, March 17 |
| 3:00 pm - 4:30 pm | |
| |
Managing Quality of Service in Distributed Computing Infrastructures
(90 mins)
Costas Kotsokalis, Ramin Yahyapour
View Participants
Quality of Service has been a research topic in the area of Distributed Computing Infrastructures for a long time, as Grids were evolving into the mature infrastructures they now are, and currently as Clouds become all the more popular. Achieving QoS based on business requirements, both on the customer and the provider side, depends on a number of things: A way for customers and providers to communicate effectively and reach an agreement, proper planning methods on both sides so that a contract is feasible, provisioning of the infrastructure as requested/necessary, and efficient and practical monitoring to enforce the agreed QoS. This session explores the topic from various viewpoints and provides insight on latest results in the area, from ongoing research projects and standardization groups, also explicating the ties between standardization and successful production infrastructures. More specifically, the session will discuss how standards are important in this area to achieve QoS management in multi-tier service-based systems, eScience and research in general, business and the Cloud. Additionally to the presentations, a panel discussion at the end will try to decipher and elaborate on latest developments in the area, and suggest the future directions of QoS management in DCIs, especially with regards to their relevance to standards.
Quality of Service has been a research topic in the area of Distributed Computing Infrastructures for a long time, as Grids were evolving into the mature infrastructures they now are, and currently as Clouds become all the more popular. Achieving QoS based on business requirements, both on the customer and the provider side, depends on a number of things: A way for customers and providers to communicate effectively and reach an agreement, proper planning methods on both sides so that a contract is feasible, provisioning of the infrastructure as requested/necessary, and efficient and practical monitoring to enforce the agreed QoS. This session explores the topic from various viewpoints and provides insight on latest results in the area, from ongoing research projects and standardization groups, also explicating the ties between standardization and successful production infrastructures. More specifically, the session will discuss how standards are important in this area to achieve QoS management in multi-tier service-based systems, eScience and research in general, business and the Cloud. Additionally to the presentations, a panel discussion at the end will try to decipher and elaborate on latest developments in the area, and suggest the future directions of QoS management in DCIs, especially with regards to their relevance to standards.
Agenda:
1. "The importance of standards in multi-tier SLA management" (Ramin Yahyapour, SLA@SOI)
2. "Introducing commercial use cases to an academic e-Science infrastructure" (Philipp Wieder, SLA4DGRID)
3. "Addressing business-related QoS Requirements focusing on standardization opportunities" (Bastian Koller, BREIN)
4. "WS-Agreement: Foundations for standardized QoS negotiations" (Wolfgang Ziegler, Fraunhofer SCAI, GRAAP-WG)
5. "Advance Reservation API - a DRMAA follow up" (Mariusz Mamonski, PSNC)
6. "Standardizing infrastructure provisioning requests: The OCCI" (Thijs Metsch, OCCI)
7. Panel discussion
Location: HGB-A 125
|
| |
| |
|
| |
| | Slides: Bastian Koller: Addressing business-related QoS requirements focusing on standardization opportunities |
| | Slides (PDF): Mariusz Mamonski: Advance Reservation API - a DRMAA follow up |
| | Slides: Ramin Yahyapour: The importance of standards in multi-tier SLA management |
| | Slides (PDF): Wolfgang Zielger: WS-Agreement |