Guo Weichao

TL;DR: An availability and cost aware bidding framework is proposed that can reduce the costs of a distributed lock service and a distributed storage service by 81.23% and 85.32% respectively while still keeping availability level the same as it is by using on-demand instances.

TL;DR: A prototype system, MARS, to enable page swapping and cache more applications, which can speed up the application relaunching and restore the application state and can outperform the conventional Linux page swapping up to four times.

TL;DR: In this article, an anomaly detection method for calculating intensive parallel tasks is presented, which comprises the following steps that program machine codes are analyzed to obtain function overriding information of the program machine code; task schedule trace functions are selected according to the function overriding Information, and instrumentation is conducted on inlets of the task scheduling trace functions to obtain instrumented program machines codes; a plurality of tasks of work are executed by a number of calculation nodes through the instrumented machine code.

TL;DR: This work proposes a new framework designed for compute-intensive tasks by using instrumentation and automatic instrument point selector, and employs a clustering method to identify outliers at runtime and perform speculative execution/aborting, speeding up task execution by up to 25%.

TL;DR: In this article, the authors proposed a method for deploying distributed services through bidding nodes in a cloud computing platform, which comprises steps as follows: obtaining historical price data of the bidding nodes from independent areas from the cloud computing platforms, establishing a bidding node failure probability forecast model according to the historical prices, estimating the failure probability of the bidder nodes in the independent areas in a preset bidding cycle according to bidding node prediction model, and obtaining the optimal bidding strategy of the bid nodes in independent areas with an online bidding algorithm.