Exploiting Tuple Spaces to Provide Fault-Tolerant Scheduling on Computational Grids

Failure-aware resource management for high-availability computing clusters with distributed virtual machines

“…Tuple spaces along with checkpointing and replication mechanisms have been applied to grid scheduling in [12]. The performance of distributed checkpointing protocols has been evaluated by Agbaria and Friedman [4].…”

A Hybrid Scheduling Algorithm with Load Balancing for Computational Grid

“…After executing a task, the resources send the result to the broker and the broker assembles all results and returns them to the user. In [7], all communication between brokers/masters and resources/workers is done exclusively through the tuple space.…”

SLA-based Asynchronous Coordination in Metacomputing Environments

“…For example, Chao et al [2] propose a centralized method using BPEL whereas Eymann et al [3] and Cao et al [1] concentrate on a decentralized agent-based approach. Another notable example is the work of Favarim et al [4] who propose to use the tuple-space model for coordination. The last approach seems most promising since its space and time independence allow for late-binding and disconnected operationqualities that fit well to the decentralized service-2-service nature of the SORCER environment.…”

“…Those providers that fulfill functional requirements (offer the requested service type) read the corresponding envelops (3) and start the process of issuing an SLA offer. At first they call the SlaPrioritizer service and request a permission to execute and the assigned priorities (4). If a permission is given (5) the provider matches its current QoS parameter values against QoS requirements retrieved from the envelop (6) and based on the outcome creates an SLA offer (if all QoS requirements are met) or proposes an updated SLA contract (otherwise).…”