Research Article
Prediction Based Proactive Thermal Virtual Machine Scheduling in Green Clouds
Algorithm 1
Proactive thermal VM scheduling.
IF (Event(Application_Arrival) ==TRUE) | Put new request in waiting queue | FOREACH physical machine in the Unified list | If (current temperature < threshold temperature) | Estimate for each node and put in the predict queue | END IF | END FOREACH | Sort predict Queue in the decreasing order of | FOREACH VM request in waiting queue | Check availability of VM in the first node in predict queue | IF Available then allocate VM to this VMRequest | Update Unified and Local lists. | ELSE | Check next node in the Predict Queue till last node is reached | END ELSEIF | IF no node is available | Event(Activate_SM) and allocate VMRequest to a new VM on this SM. | ENDIF | END FOREACH | ENDIF | IF (Event(Application_Complete) ==TRUE) | Event(Free_VM) | Update Unified and local lists | IF(Active_VM_Count == 0) | Event(Hibernate(SM)). | Update Unified and Local lists. | ENDIF | ENDIF |
|