Concurrency Control in Distributed Database Systems
ACM Computing Surveys (CSUR)
JSSPP '02 Revised Papers from the 8th International Workshop on Job Scheduling Strategies for Parallel Processing
Livelock Avoidance for Meta-Schedulers
HPDC '01 Proceedings of the 10th IEEE International Symposium on High Performance Distributed Computing
ICPADS '05 Proceedings of the 11th International Conference on Parallel and Distributed Systems - Volume 01
Advance Reservation and Co-Allocation Protocol for Grid Computing
E-SCIENCE '05 Proceedings of the First International Conference on e-Science and Grid Computing
Consensus on transaction commit
ACM Transactions on Database Systems (TODS)
IPDPS '09 Proceedings of the 2009 IEEE International Symposium on Parallel&Distributed Processing
JSSPP'07 Proceedings of the 13th international conference on Job scheduling strategies for parallel processing
A deadlock and livelock free protocol for decentralized Internet resource coallocation
IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans
| Hi-index | 0.00 |
In a Cloud environment, users face the challenge of selecting and composing resources and services from a single or multiple providers. As several negotiations can occur concurrently, information on service and resource availability may be out-of-date, thus requiring several iterations between users and providers until an agreement is achieved. To address this problem, we introduce CANPRO, a Conflict-Aware Negotiation Protocol for allocating Cloud resource and services aimed at reducing cancellation messages during negotiation. CANPRO allows users (or entities on their behalf) to know the amount of resources being concurrently negotiated by other users and the number of users interested in such an amount, while still keeping users' information private. By knowing this information, users can, for instance, confirm allocation requests with lower chances of having collisions with other users. In addition, for the same reason, users can increase their time deciding which (combination of) resources they want to allocate. The paper presents comparative results of CANPRO against the popular two-phase commit protocol (2PC) and a state-of-the-art protocol named SNAP-3PC. We used think time, network overhead, number of concurrent negotiations and providers as main metrics. The results are promising and the protocol can be used in scenarios other than Cloud Computing; for instance, bookings of health services, cars, tickets for venues, schedule of appointments, among others.