Abstract:
Datacenter applications need low-latency data service. Proper data placement in modern distributed storage systems reduces data movement delay, which can greatly reduce service latency. Existing data placement solutions often assumed data request distribution or discovered it through trace analysis. Data placement is difficult due to dynamic network conditions and time-varying user request patterns. Static model-based solutions struggle with dynamic systems.
DataBot+, a reinforcement learning-based framework, automatically learns optimal placement policies. DataBot+ uses Q-learning neural networks to estimate near-future latency from real-time data flow measurements.
DataBot+ is decoupled into two asynchronous production and training components to avoid data flow overheads and instantaneous decision making. Real-world traces prove our design works.
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
