Center for Experimental Research in Computer Systems

CERCS group is involved in developing middleware and kernel level benchmarks and tools. A comprehensive list is as below :

Web Services Interoperability Benchmarks

Description:

Web Services Interoperability(WS-I) Organization provide resources for web services developers to create interoperable web services and verify that their results are compliant with WS-I guidelines.It includes Profiles, Sample Applications and Testing Tools. Sample Applications demonstrate Web services applications that are compliant with WS-I guidelines.These implementations are developed using multiple platforms, languages and programming tools, demonstrating interoperability in action, and providing readily usable resources for the Web services developer. Sample applications serve as working examples for developers looking to follow the WS-I guidelines in their programming environment of choice. More information on the sample application's provided by WS-I can be found at: WSI Organization Page

IBM provided our group a version of sample webservice retail application for WebSphere 6.0. The sample webservice retail application is ported to ServiceMix ESB.
The source code is available at : WSI sample application port to ServiceMix ESB

MSO Workloads

Description:

Example robotics application for use with Mobile Service Overlay (MSO). This package builds into a set of shared objects that can be used as a "dll" in EVpath-based overlays.

The components for each module are:

Localize: localize/localize.so, expects rawmap in the same directory. This replaces an extra dependence on mapserver to obtain the map. Expects laser data as input and produces positions as output.
Navigator: navigator/navigator.so, expects rawmap in the same directory. Obtains the laser data, positions from localize, and odometry, to produce the path plan.
Convert: navigator/convert.so. Used to convert odometry data to a format readable by navigator. Should soon be deprecated.
Edge detection:edge/edge.so. Edge detection on ppm images
Blob finder: blob/blobfinder.so, expects colors.txt in the same directory. Output is the same format as input
Bayesian classifier: bayes/bayes.so, expects the file under FILENAME in main.c in the same directory, number of iterations can be changed by changing #define ITER in main.c. For now does some dummy bayesian inference, ignoring the inputs. Ideally, it should use the inputs to extract some features, use the .xdsl file as a lookup table to perform the inference, and produce the output as a probability (that the area is a target/not).

NOTE: This module needs expat library.

StreamPerf

Description:

StreamGen is a workload generator which is targeted at distributed information flow applications. These include the event streaming services used in wide-area publish/subscribe systems or in operational information systems, the data streaming services used in remote visualization or collaboration, and the continuous data streams occurring in download services. Running across heterogeneous distributed platforms, these services are implemented by computational component that capture, manipulate, and produce information streams and are linked via overlay topologies. StreamGen can be used to produce the distributed computational and communication loads imposed by these applications. Dynamic application behaviors can be created with mathematical specifications or with behavior traces collected from application-level traces. An interesting set of traces presented in this paper is derived from long-term observations of the FTP download patterns observed at the Linux mirror site being run by the CERCS research center at the Georgia Institute of Technology. Two different flow-based applications are created and evaluated with StreamGen. The first emulates the data streaming behavior in a distributed scientific collaboration, where a scientific simulation (i.e., a molecular dynamics code) produces simulation data sent to and displayed for multiple, interactive remote users. The second emulates portions of the event-streaming behavior of an operational information system used by a large U.S. corporation. Parametric studies with StreamGen.s FTP traces applied to these applications are used to evaluate different load balancing strategies for the cluster machines manipulating these applications. data streams

More information can be found here.

Data Sets and Codes

Description:

The following is list of data sets and related codes available to Systems students:

Title Source Languages Description

Delta IXP, C, E-code Some FAA feeds plus other data that represents an Operational Information System

(OIS) Molecular Data IXP, C, E-code Output from the a molecular dynamics simulation

Hydrology Data Java, C, E-code

OpenGL/PPM IXP, Kernel, C, E-code Streaming images from visualizing the MD or Hydro data

Atmospheric data Weather simulation data

Linux mirror site traces Analysis of traffic at the CERCS Linux mirror site

Title	Source Languages	Description
Delta	IXP, C, E-code	Some FAA feeds plus other data that represents an Operational Information System
(OIS) Molecular Data	IXP, C, E-code	Output from the a molecular dynamics simulation
Hydrology Data	Java, C, E-code
OpenGL/PPM	IXP, Kernel, C, E-code	Streaming images from visualizing the MD or Hydro data
Atmospheric data		Weather simulation data
Linux mirror site traces		Analysis of traffic at the CERCS Linux mirror site

More information on this can be found here.

Dynamic Data Access to the GT/CERCS Linux Mirror Site

Description:

CERCS operates the largest FTP mirror site in the southeastern US. We analyzed the dynamic access patterns to this large data repository to better understand the dynamic data sharing behavior of certain classes of grid end users.

More information on this can be found here.

Nutch Modifications

Description:

We have created a web service application based upon the Nutch open source search engine. In particular, we modify the behavior of Nutch in order to resemble observed aspects of actual deployed web services. We drive Nutch with requests from different client classes that are in turn associated with varying SLA requirements, including the quality of results in responses. We incorporate both a differentiated service model and a response quality metric by defining requests to require a varying number of search responses in order to be completed, resulting in more (or less) work by the search engine.

This allows client class SLAs to be based upon both latency and average quality of information (QoI).

Service Oriented Application - Reference Implementation

Description:

In order to facilitate future research on service oriented architectures, we create a simulation system which can be used as an experimental platform and a benchmark. This simulator is based on a real world system: an airline fare pricing/ticketing website and the necessary services to support it. We describe the system our simulator is based on and the modifications we made to make the simulation more amenable to supporting research.

More information on this can be found here.

Hermes - Information Service

Description:

HERMES is an information service system with a high performance management and dissemination architecture. It leverages distributed architecture to provide relevant information tailored to user and capabilities to allow user to make timely decisions. It processes distributed heterogeneous structured data intelligently and build correlations amongst the data at runtime. This systems utilizes the existing middleware frameworks such as EVPath developed by MWARE group.

Some of the applications of this systems are campus network monitoring and decision support system for financial analysts.