This year, we have 5 graduate students leaving REINS with their Master's degrees. Among them, Yin, Zhida will join Optiver;Zhu, Yiming will join Zhejiang People’s Government; Yu, Fangxiaoqi will join Baidu;Cao, Yuting will join eBay;Shen, Jiamei will join PayPal. More details can be found at their personal pages and the alumni page. Congratulations! And best of luck in their future endeavors!

This year, we have 7 graduate students leaving REINS with their Master's degrees. Among them, Li, Ran will join MiHoYo;Wu, Ye will join WangYi; Liu, Yu will join Baidu;Zhang, Xin will join JingDong;Zhao, Zihao will join PayPal; Wu, Qiangqiang will join Alibaba and Gong, Lihang will join IQiYi. More details can be found at their personal pages and the alumni page. Congratulations! And best of luck in their future endeavors!

EDGE 2017 CLOUD 2017 SCC 2017

REINSians attended SCF 2017 with three presentations: EAERS: An Enhanced Version of Autonomic and Elastic Resource Scheduling Framework for Cloud Applications, PRMRAP: A Proactive Virtual Resource Management Framework in Cloud and Back Propagation Grouping: Load Balancing At Global Scale When Sources Are Skewed. They were accepted by CLOUD 2017, EDGE 2017 and SCC 2017 respectively.

In the paper of EAERS, we designed a framework called EAERS which is an enhanced version of our previous work (AERS). It eliminates modeling the specific cloud application and instead determines the relationship between workloads and the number of virtual machines (VMs) through self-learning so that the whole scheduling can be carried out in a more transparent manner. Dynamic consolidation is also designed, implemented and integrated into EAERS.

In the paper of PRMRAP, we designed a framework called PRMRAP which is a proactive framework based on the prediction of resource amount to cope with sudden traffic change. Compared with post-action methods and existing proactive methods, we have lower time latency and we consider not only horizontal resizing but also vertical resizing of scaling group which makes our model quicker and much more cost-effective.

In the paper of Back Propagation Grouping, we found out that the upstream skewed sources can also exacerbate the load imbalance in the downstream workers and this bottleneck cannot be handled well by existing schemes. Thus, we propose a novel stream partitioning solution called BACK PROPAGATION GROUPING (BPG), and its core components are key splitting, back propagation and calibration signal.

The above pictures were shot at the SCF 2017, which took place at Honolulu, USA, June 25 - June 30, 2017.