Tutorial 1 - Network Performance Monitoring and Measurement Techniques
Supratik Bhattacharyya (supratik@sprintlabs.com) and Sue B. Moon (sbmoon@sprintlabs.com), Sprint ATL
As the Internet continues to grow rapidly in size and complexity, it
has become increasingly clear that its evolution is closely tied to a
detailed understanding of network traffic. Consequently, the
development of tools and techniques to capture Internet traffic and
its properties has gained widespread attention. This tutorial covers
various approaches for monitoring and measuring Internet traffic and
techniques for analyzing and interpreting its properties.
In the first part of this tutorial, we will provide a comprehensive
survey of existing approaches and techniques, including discussions on
the types of data to be collected, points of data collection, common
metrics and the tradeoffs of different approaches. We will describe
standard tools and information sources such as such as ping,
traceroute, SNMP, etc., and survey pioneering projects in the area of
network monitoring.
In the second part of the tutorial, we will discuss how measurement
can help in the design, engineering and operation of IP backbones by
providing valuable input for resource provisioning, traffic
engineering, routing protocol operations, etc. We will illustrate this
section with a variety of tools, traces and observations from
operational networks.
Finally we will cover some unsolved aspects of network measurement
with focus on how completeness of data can impact the exactness of the
observation, and how to infer the characteristics of a network from
local or partial observations. We will conclude with a discussion on
future of Internet monitoring and measurement in anticipation of
higher-speed links and future router generations.
- Presenters' Biographies
Supratik Bhattacharyya is a member of the IP/Internetworking Group at
the Sprint Advanced Technology Labs, Burlingame, CA. He received a
Ph.D in Computer Science from the University of Massachusetts Amherst
in 1999. Prior to that, he earned an M.S. in Computer Science from the
University of Massachusetts in 1995, and a Bachelors degree in
Computer Science and Engineering from Jadavpur University, Calcutta,
India in 1992. His research interests include Internet measurement
and monitoring, traffic engineering, routing protocols and IP multicast.
Sue B. Moon received the B.S. and M.S. from Seoul National University,
Seoul, Korea, in 1988 and 1990, respectively, all in computer
engineering. From 1990 to 1991, she worked for IMIGE Systems, Inc. in
Seoul, Korea. In 1999 she received her Ph.D. from the Dept. of
Computer Science at the University of Massachusetts at Amherst in
1999, and joined Sprint ATL in Burlingame, California. She is a
member of the IP-Interworking group and works in the Sprint IP
monitoring project. Her current research topics include network delay
and scaling behavior analysis, network anomalies, and monitoring and
analysis system design.
Tutorial 2 - Service Network Control in Multi-Technology Environments
Aleksandar (Sasha) Ratkovic, CTO, CPLANE (sasha@cplane.com)
The network is a purpose built utility designed to generate service revenue for the provider and allow the growth of business. In addition to providing network service, Service Providers (SPs) also want to make optimal use of its network resources in order to derive maximum profit from the business. The basic mechanisms that are used to provide network services and optimize network resource utilization have been around for a while. All these mechanisms are parametrized by a set of well-defined parameters. What was subject to constant change were implementations of these mechanisms and protocols used for setting the mechanism's parameters. This implies that it is possible to implement a unified model (multi vendor, multi technology) of the network state, which contains the state of all modules implementing fundamental mechanisms.
In this talk we first define the problem facing SPs in their effort to offer network services in multi technology environments. We then define the set of basic mechanisms that are used to provide quantifiable SLAs in a diverse environment:
- Traffic conditioning
- Admission control
- Bandwidth management
- Forwarding Control
- Traffic Engineering
We then map these mechanisms into existing technologies and vendor implementations and show that coherent control of network infrastructure allows for end-to-end QoS. Technologies/mechanisms that will be addressed are MPLS, traffic engineering, Diff-serv.
We also discuss the importance of the common Information Model, which is defined as abstract representation of managed elements in networking environment that describes their properties, operations and relationships and is independent of the implementation platform. We then discuss the existing models and extensions required to describe the fundamental mechanisms described above. The Common Information Model (CIM) is seen by major service providers as the most important enabler for the next generation OSS.
- Presenter's Biography
Sasha Ratkovic is currently Chief Technology Officer at CPLANE, where he is in charge of the strategy and architecture for the family of service activation and network optimization products in multi-vendor multi-technology environments. His experience and research have helped him to build a world class technology team with broad expertise in IP, MPLS, ATM and optical networking technologies. Prior to joining CPLANE, he was with AT&T Labs, where he was a key technical lead in the architecture, development and delivery of an advanced services platform for next-generation IP services in AT&T's Internet Platform Organization. Sasha holds Ph.D in Electrical Engineering from the University of California, Los Angeles.
