Network Weather Service User's Guide
Online guide to running NWS at NMI
The North Carolina Internet2 Technology Evaluation Center was chartered by Internet2 and the University Corporation for Advanced Internet Development (UCAID) to serve as a national network testing laboratory for the Internet2 community. The two major aspects of the NC-ITEC mission are:
1: to provide testing and evaluation services in support of the existing national backbone network, Abilene. The focus of this effort is on optimizing technologies currently deployed in Abilene.
2: to support research into new technologies that have the potential to progress the state of the art in networking infrastructure and advanced services.
ITEC-Ohio (Internet2 Technology Evaluation Center) is a consortium of Ohio universities, government agencies, and corporate partners whose purpose is to examine critical new applications, infrastructure problems and emerging technologies in order to meet future academic, engineering and industrial needs in research, education, government and commerce. ITEC-Ohio was created in December 1999 by UCAID (University Corporation for Advanced Internet Development) and is chartered by Internet2 to serve as a national network-testing laboratory for the Internet2 community. ITEC-Ohio is one of only two ITEC testing laboratories in the United States. The other ITEC facility is located in North Carolina.
Here are some links with description from Steve Senger at Unv Wisconsin - LaCrosse.
http://astra.csfac.uwlax.edu/InformationChannels/
InformationChannels is the mechanism we use to advertise services offered by our applications. There are three redundant registration servers running (2 @ La Crosse, 1 @ Stanford). Apps register their available channels using multicast packets. The page contains a list of service categories which link to a cgi script which interrogates the registry and builds a list of channel links. The pages are fairly crude at this point. There is some explanatory material but not enough. I really need to find time to write up more details but it is always a choice between making new things work and writing up old things. We are working on more polished front door pages for use by NLM in their show&tell center.
http://astra.csfac.uwlax.edu/INet2
These are graphs we have been keeping since last April when WiscNet finally got I2 connectivity configured for me. I've been tracking roundtrip latency, pkt loss, tcp (not particularly optimized, and a max udp rate with small loss. We are moving comparable measurements over to the WeatherStations.
http://astra.csfac.uwlax.edu/WthrStns
These are the new graphs from the WeatherStations. There are 3 running: 2 @ La Crosse (saber & melonhead), 1 @ Stanford(elegon). You will notice that there is something very peculiar happening on this campus. Since the WeatherStations advertise their services through the InfomrationChannels, other apps will be able to directly obtain information from them. right now there is a logging app which periodically looks for registered stations and records their data into RRD tables and graphs.
From Steve Senger at the Unv Wisconsin - La Crosse Information Channels:
"The term "information channel" refers to a data or visualization service provided by a network application. The network client and server applications developed for the NGI project are able to announce the availability of their services through a registration server application. This page provides links to query the registration server for currently active channels of different types."
DataGrid:
Grid Network Monitoring Document Accompanying Deliverable D7.2 Demonstration of Enhanced Monitoring Tools
http://edms.cern.ch/document/334883
Introduction and Objectives of the document:
1. INTRODUCTION
This document is provided as an accompaniment to the WP7 PM12 deliverable (D7.2), a prototype demonstrating enhanced network monitoring tools. It provides a framework for network monitoring against which WP7 has operated in preparing this prototype. The prototype is designed to demonstrate how basic network monitoring metrics can be made available to the Grid middleware in addition to the visualisation of the metrics via the Web. The architecture upon which this is based is simple and easily extensible. It is in these respects that WP7 views this deliverable to be "enhanced" with respect to the operation of the Grid.
WP7 expects the prototype monitoring tool to be developed further. Specifically WP7 is focusing its work in the area of predicting and forecasting, network metrics based upon current measurements. WP7 believes that if Grid middleware is to make best possible use of metric information what is required is a current measurement and a prediction over the "next" time interval. Based upon this work WP7 expects to review its monitoring architecture with a view to accommodating forecasting techniques within the existing scheme, or if necessary to replace what currently exists with a more suitable environment.
WP7 will also work with the other Work Packages to ensure that the network metrics from monitoring are of value to the EDG middleware (the resource brokers). In addition, WP7 will review the output from its network monitoring deployed on the Testbed1 and will also accommodate the monitor ing data collected by the Grid applications, that is to say, the Grid traffic.
1.1. OBJECTIVES OF THIS DOCUMENT
This document provides a review of the requirements for monitoring with respect to Grid operations, a review of the classic network monitoring metric s and a review of the well known network monitoring tools. It describes the tools that have been selected for this prototype release and how they provide metric information to both the Grid middleware and, via visualisation, to a human observer. Finally this document outlines future areas of work that WP7 believes are required to complete its task.
The Pioneer page with links to Pioneer Overview and reports from Abilene & Envoy.
Draft document for Starbox Appliance Development.
Yet another George notebook for capturing ideas, links, notes, and comments about the E2Epi network measurment appliance.
Have a look at Ruggidized Luggable Lunchbox & Crushproof Suitcase PC's/Servers from TeamMicro.
The Grid High-Performance Networking Research Group (GHPNRG) focuses on the relationship between network research and Grid application and infrastructure development. The objective of GHPNRG is to bridge the gap between the networking and grid research communities. It accomplishes its goal by serving as a forum for information exchange on advances and requirements in both fields, as well as by providing a focal point for liaison activities between the GGF and the various networking standards bodies.
The interface between Grid applications and network services Deployment of new technologies on the Internet and overlay networks The GHPNRG provides a forum for such topics until sufficient maturity and interest to both communities is reached that naturally results in the formation of a separate WG or RG to pursue them further. The discussions of GHPN are carried out both during the meetings and on the GHPN mailing list.
There's a tool list at WareOnEarth
They "specialize in high performance Internet engineering, network measurement and analysis, and information systems security. Our current projects include working with the Defense Research Engineering Network (DREN), the Space and Naval Warfare Systems Command (SPAWAR), and the National Laboratory for Applied Network Research (NLANR)."
SONAR is a service which, when presented with a list of Internet Protocol addresses, attempts to order that list according to the proximity from the SONAR server to each address. SONAR is designed to help networked applications make a reasonable choice between several alternative locations for a network-accessible service.
The LBNL Data Intensive Distributed Computing Group maintains a Tuning Guide for Wide Area Distributed Applications
Dunigan's TCP auto-tuning zoo is a "brief summary of auto-tuning work to date, starting with network-aware applications and then kernel modifications to tune TCP flows."
Here is Tom Dunigan's Network Peformance Links. It is a long list that he keeps pretty well up to date.