Optimization of checkpointing/recovery strategy in cloud computing with adaptive storage management

摘要

Cloud Computing is a type of distributed system that is usually based on the services offered tothe user based on SLA contract. In this case, the implementation of a fault-tolerant system thatensures the reliability and the services continuity becomes amajor requirement. In this paper, wepropose a fault tolerance strategy based on checkpointing and replication. Our approach uses asmart checkpoint infrastructure for cloud computing tasks. The checkpoints are stored in alternativealready paid VMs. This allows resuming a task execution faster and cheaper after a nodecrash. Since checkpoints are distributed and replicated, our approach increases also the systemreliability. The experimental results show the effectiveness of the proposed strategy in term ofenergy consumption, SLA (System Level Aggregation) violation, and reliability.