Abstract:
Tail latency reduction improves service experience. User-facing latency-sensitive cloud applications typically have multiple interactive tiers (e.g., Web, App, Database) running in different VMs with complex interaction patterns. Previous VM consolidation methods neglected such VM-tier interactions, resulting in poor application performance.
This article examines multi-tier interactive workload consolidation from a new perspective of user-perceived tail latency. A novel profiling-based consolidation method meets tail latency requirements while reducing physical machines. We first perform large-scale profiling experiments under various consolidation settings in a KVM virtualized private cluster to establish empirical performance values.
Interference with co-located VMs and tier interaction affect multi-tier workload tail latency. We model multi-tier workload consolidation as an optimization problem with different objectives and constraints and derive the consolidation schedule.
We test and compare the proposed models with other methods (without profiling or interaction influence). Extensive experimental results show that the proposed method can reduce tail latency by 5X compared to the method without profiling and 1.3X compared to the method without considering tier interaction.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
