平成24年度 博士課程教育リーディングプログラム:東京大学 ソーシャルICT グローバル・クリエイティブリーダー育成プログラム

ホーム > GCL TechTalk シリーズ > 2013/11/15 Global Design Seminar:Dr. K. K. Ramakrishnan講演会

GCL TechTalk シリーズ

2013/11/15 Global Design Seminar:Dr. K. K. Ramakrishnan講演会

日時: 2013年11月15日金曜日 10:00-12:00

場所:工学部3号館1F電気系セミナー室2&3  HowToGetThere

講演題目:Networking Cloud Data Centers: Providing Dependable Storage and Computation(ICNP2013基調講演)

Cloud computing has changed how enterprises run and manage their IT systems. Cloud computing platforms provide customers with flexible on demand resources. Data centers exploit economies of scale and provide a low cost structure to users as a result of multiplexing gains achieved by exploiting virtualization. Applications running on these large data centers have also adapted to be able to sustain component failures. However, for wide-spread migration of enterprise applications to the Cloud, it is desirable to provide dependable compute and storage capabilities while exploiting the elasticity and flexibility offered by Cloud computing. We have proposed our CloudNet architecture to logically unify cloud and enterprise data center resources, providing opportunities for cross data center load balancing and dynamic application placement. We build on CloudNet to support Disaster Recovery (DR) solutions, addressing the needs of storage in the face of data center failures, and provide server availability in the face of transient server failures.

Our DR solution looks at the difficult problem of providing storage replication across Cloud sites that are geographically separated. Cloud based DR promises economies of scale and on-demand provisioning that is a perfect fit for the infrequent yet urgent needs of disaster recovery. However, the WAN latency between a cloud and enterprise site can be a major performance bottleneck when synchronously replicating an application’s data into the cloud. I will describe our proposal called ‘Pipelined synchrony’ to address this problem and show how PipeCloud can address even the stringent needs of financial applications that require DR without data loss. I will then describe our approach to provide high availability through replication of live memory snapshots for virtual machines. The use of a ‘transient server’ abstraction and utilizing a shared backup server allows us to deal with termination of VMs with little application impact. With Cloud services exploiting ‘spot instances’ to improve multiplexing and data centers moving to renewable energy sources, the ability to provide high availability in such environments is increasingly important. All of this work was done in collaboration with Prof. Prashant Shenoy of UMass and his students.

 

Profile

Dr. K. K. Ramakrishnan is a Distinguished Member of Technical Staff at AT&T Labs-Research. He joined AT&T Bell Labs in 1994 and has been with AT&T Labs-Research since its inception in 1996. Prior to 1994, he was a Technical Director and Consulting Engineer in Networking at Digital Equipment Corporation. Between 2000 and 2002, he was at TeraOptic Networks, Inc., as Founder and Vice President. Dr. Ramakrishnan is an AT&T Fellow, recognized for his fundamental contributions on communication networks and lasting impact on AT&T and the industry, including his work on congestion control, traffic management and VPN services. He is an IEEE Fellow, and has received other awards. His work on the “DECbit” congestion avoidance protocol was recognized in the 1995 retrospective issue of ACM Sigcomm Computer Communication Review as one of the 16 most important papers published over the previous 25 years in ACM Sigcomm publications. The work once again received the ACM Sigcomm Test of Time Paper Award in 2006. He has published nearly 200 papers and has more than 108 patents issued in his name. K.K. has been on the editorial board of several journals and has served as the TPC Chair and General Chair for several networking conferences and has been a member of the National Research Council Panel on Information Technology for NIST. K. K. received his MS from the Indian Institute of Science (1978), MS (1981) and Ph.D. (1983) in Computer Science from the University of Maryland, College Park, USA. His web page can be found at: (http://www.research.att.com/people/Ramakrishnan_Kadangode_K/)

 


« GCL TechTalk シリーズのトップに戻る