Hiroshi Nakagoe

Flow Control for Higher Resiliency in Wide Area Distributed Cloud Computing

Wide area distributed cloud computing, where data are processed and stored at micro datacenters (DCs) located around terminals, improves response time and reduces the amount of traffic in wide area networks. However, it is difficult for the cloud to maintain service in the case of micro DC failure because the robustness of micro DCs is substantially lower than that of conventional DCs. We propose a flow control system that transfers the flow to other micro DCs by converting the L2, L3, and L4 headers of received packets via OpenFlow when micro DC failures occur. The proposed system enables flow to be transferred to other micro DCs even from M2M terminals such as sensors that are unable to change the destination IP address. We also developed a distributed control system to solve the critical problem of switching delay caused by the packet-in mechanism used when applying the OpenFlow to the distributed cloud. Emulation results showed that throughput in the distributed control became stable 12 times as quickly as that in the centralized control when flow was transferred.

Wide area tentative scaling (WATS) for quick response in distributed cloud computing

As cloud computing is increasingly adopted by enterprises, a stringent service level agreement such as response time is required by the cloud in order to run business applications. This can be problematic because activity from geographically distributed terminals owing to globalization often increases the average response time. Federating various clouds enables to utilize datacenters in various geographical regions regardless of their servicers. The response time can be reduced by replicating the applications and related data at datacenters near the terminals by considering the factors of delay (e.g., data synchronization, distribution of multi-tier applications, and influence of other applications). However, it is unrealistic to accurately analyze all of the factors without any errors. We propose wide area tentative scaling (WATS) to improve the response time in a phased manner by repetitively replicate a part of the application and related data at other datacenters and selecting a better organization. The WATS approach is to repetitively change the organization to reduce the response time even if the analysis is incorrect, unlike mathematical-formulae-based approaches that require precise analysis. The drawback of this approach is that it consumes more computing resources due to repeating the replication. We therefore, applied Bayesian inference to search for a better organization with fewer trials. Evaluation results showed that WATS successfully reduced the response time in a phased manner.