Quality of Service and Real-Time Stream Data Processing
Sang Hyuk Son
Dept. of Computer Science
University of Virginia
Abstract
There is a growing need for real-time data services
in distributed environments.
Many real-time applications and information services
are becoming very sophisticated in their data needs that span
the spectrum from low level status data, typically
acquired from sensors, to high level aggregated data.
Examples include aerospace and defense systems, sensor networks
and ubiquitous computing, traffic control, and web-based applications.
Providing quality-of-service guarantees for data services
in a distributed environment is a challenging task. The
presence of multiple sites in distributed environments raises
issues that are not present in centralized systems. The
transaction workloads in distributed real-time databases may not
be balanced and the transaction access patterns may be
time-varying and skewed. Data replication is an effective
method to help data service systems meet the stringent temporal
requirements of real-time applications.
Further, some of the applications need to
operate on continuous unbounded streams of data. These
applications have inherent real-time performance requirements
that have to be met under high-volume, time-varying incoming data
streams.
In this lecture, we will first discuss quality of service metrics
in distributed real-time data services and present several approaches
used to provide specified level of QoS in distributed real-time
applications. We will also discuss a real-time data stream query
model named PQuery, which provides periodic real-time queries on
data streams.
Pre-requisite:
Some knowledge in database system concepts, such as transaction
management, concurrency control, and replication control will be
necessary.
In addition, some knowledge in real-time systems, such as temporal
consistency and real-time scheduling will be helpful to follow
the material.
Note on lecture contents:
The contents of the first 2 lectures will be based on the material
presented in ARTES summer school in 2004. The lecture #3 will be on
stream data management, based on the lecture given at the University
of Skövde on June 14, 2006.
