Multitenant Applications in the Cloud: Continuous Monitoring and Live Database Migration for Autonomic Load Management

Speaker:        Professor Divy Agrawal
                Department of Computer Science
                University of California at Santa Barbara

Title:          "Multitenant Applications in the Cloud:
                Continuous Monitoring and Live Database Migration for
                Autonomic Load Management"

Date:           Wednesday, 28 March 2012

Time:           3:00pm - 4:00pm

Venue:          Lecture Theatre F (near lifts 25/26), HKUST

Abstract:

Cloud platforms handle large numbers of applications with small data
footprints, different types of workloads, and unpredictable load
characteristics. A multitenant database management system (DBMS) storing
and serving the data for these applications (tenants) is therefore an
important component of the platforms' software stack. Continuous
monitoring of tenant and node performance is critical for resource
provisioning, reacting to changes in workload patterns, elastic load
balancing, and orchestrating database migration to ensure effective
resource utilization and minimize operational cost. A self-managing system
controller faces multiple algorithmic and system level challenges. For
instance, how to characterize a tenant given the variety of workloads a
tenant might issue; how to reduce the impact of co-locating multiple
tenants, i.e., which tenants to co-locate and how many tenants per server;
how to adapt to changes in behavior, i.e., how to detect changes in tenant
behavior, how to determine the root cause of the change, and how to
mitigate the impact of changed behavior; and how to associate a service
level objective to a tenant in a consolidated setting? In this
presentation, I will discuss some of the challenges we have faced in the
trenches of designing a self-managing coordinating component that monitors
system performance, controls tenant placement, and manages elasticity. Our
approach is to "learn" tenant behavior through observation and analyze the
tenants to characterize their behavior. We use a collection of system
level and DBMS agnostic database level performance measures to represent
tenant and node behavior, a combination of machine learning techniques to
characterize tenant and node behavior, and a reactive mechanism to detect
and mitigate performance crises arising from unpredictable workloads and
behavior. Our approach continuously learns and models behavior to maintain
a history as well as an up-to-date understanding of behavior. I will also
discuss the recent approaches we have developed for live database
migration in the cloud computing infrastructures.


*************************
Biography:

Dr. Divyakant Agrawal is a Professor of Computer Science at the University
of California at Santa Barbara. His research expertise is in the areas of
database systems, distributed computing, data warehousing, and large-scale
information systems. From January 2006 through December 2007, Dr. Agrawal
served as VP of Data Solutions and Advertising Systems at the Internet
Search Company ASK.com. Dr. Agrawal has also served as a Visiting Senior
Research Scientist at the NEC Laboratories of America in Cupertino, CA
from 1997 to 2009. During his professional career, Dr. Agrawal has served
on numerous Program Committees of International Conferences, Symposia, and
Workshops and served as an editor of the journal of Distributed and
Parallel Databases (1993-2008), the VLDB journal (2003-2008) and currently
serves on the editorial board of the ACM Transactions on Database Systems
and serves as the Editor-in-Chief of Distributed and Parallel Databases,
an International Journal. From 2012 through 2017, he will serve as a
member of the Board of Trustees of the VLDB Endowment. Also, he served as
the Program Chair of the 2010 ACM International Conference on Management
of Data and served as the General Chair of the 2009, 2010, and 2011 ACM
SIGSPATIAL Conferences on Advances in Geographical Information Systems.
Dr. Agrawal's research philosophy is to develop data management solutions
that are theoretically sound and are relevant in practice. He has
published more than 300 research manuscripts in prestigious forums
(journals, conferences, symposia, and workshops) on wide range of topics
related to data management and distributed systems and has advised more
than 30 Doctoral students during his academic career. The Academic Senate
at UC Santa Barbara awarded Dr. Agrawal 2010-11 Outstanding Graduate
Mentor Award. Dr. Agrawal is an ACM Distinguished Scientist, is a Fellow
of the ACM and is a Fellow of IEEE. His current interests are in the area
of scalable data management and data analysis in Cloud Computing
environments, security and privacy of data in the cloud, and scalable
analytics over social networks data and social media.