CHIMIT 2009 in Detail.

Saturday, November 7

8:00 - 9:00
Registration, full breakfast Atrium
9:00 - 10:30    Welcome

Invited Speaker: Johnvey Hwang, Splunk

"Why did we build Splunk? Product Evolution and Innovation"

Abstract: Ask any IT administrator to describe his or her job and you may get something along the lines of, "do more with less".  As nearly every aspect of business is being routed through the datacenter, the volume of data that needs to be monitored and analyzed is increasing at a phenomenal -- and unmanageable -- rate.  Instead of just coping with this overload, Splunk embraces complexity and tries to do more with its "index first, interpret later" philosophy towards IT management.  In this session, we will discuss the inspirations that drive Splunk's product evolution, how we use simple constructs to do the heavy lifting, and how we try to make IT operations a little more fun.

Bio: Johnvey Hwang heads the user interface development team at Splunk.  His current role focuses on creating a flexible user interface as well as maintaining developer-friendly APIs on top of the Splunk Server.  He often explores remixing various interfaces into Splunk, in search of efficient design patterns.  His previous work centered around emerging AJAX technologies, some presented at the AJAXWorld conference.  Johnvey has been with Splunk for over 4 years and holds a B.S. in Operations Research and Industrial Engineering from Cornell University.
10:30 - 11:00
Break Atrium
11:00 - 12:00
Papers - Field studies of system administration
  • System Administrators as Broker Technicians,

    Nicole F. Velasquez, The University of Arizona, Suzanne P. Weisband, The University of Arizona

    This research investigates the work practices of system administrators. Using semi-structured interviews and an analysis of existing system administrator literature, we theorize that system administrators act as technical brokers who bridge two communities, the end users they support and their own technical community. We also show that system administrators, like other technical workers, rely on contextual knowledge. This knowledge is largely acquired through practice and less through formal education and certification. Through a discussion of common reactive and proactive system administrator duties, we present system administrators as broker technicians who must mediate between the end users they support and their technical community. We end with a discussion of the changing role of sysadmins as their tools and users get more sophisticated.

  • Scripting Practices in Complex Systems Management
    Eser Kandogan, IBM Almaden Research Center,  Paul P. Maglio, IBM Almaden Research Center,  Eben M. Haber,  IBM Almaden Research Center,  John H. Bailey, IBM Almaden Research Center,  


    System administrators are end-users too. And as end-users, they develop tools, create web pages, write command-line scripts, use spreadsheets, and repurpose existing tools. In short, they engage in end-user programming activities in support of their systems management work. We examined system administrator practices in software tool development, operations, and maintenance based on ethnographic field studies at service delivery centers and data centers across the United States. Our findings suggest that software practices were mostly informal and collaborative and mixed within formal change processes; tool development and debugging were interleaved with tool use and maintenance as they interacted with live systems; and the complexity of large-scale systems and the risks involved in changing live and critical systems put increased demands on system administrators. We argue that system administrators might benefit from certain software engineering methodologies such as agile software development and software modeling.

12:00 - 12:30
12:30 - 1:30 
Invited Speaker: Carolyn Hennings, Windward IT Solutions

"Do I Know Who, What, When, Where, How, and WHY?"

Abstract: The quest to effectively and efficiently manage IT services involves a continual translation of data into information that provides knowledge to individuals needing to make wise decisions. ITIL® describes many sources of data and identifies the information that can be derived from that data. ITIL® also recommends that this information be made available at the right time, in the right place, in order to enable informed decisions.

All rather lofty statements … how do we bring these concepts to reality for people managing IT services and making decisions on a daily basis?

This presentation identifies the types of data, information, and knowledge needed by IT personnel on a regular basis. By profiling the types of decisions that need to be made, my goal is to help participants identify new and innovative ways of enabling those decisions and observing the impact of the decisions.

Bio: Carolyn M. Hennings is a consultant at Windward IT Solutions, which provides service management consulting services, including ITIL process design and implementation. At Windward, Carolyn provides ITIL consulting services to clients in the federal and commercial sectors. She has over 18 years of experience in helping IT organizations understand and advance their operational and project management processes. In utility, consulting, technology, and Internet companies, she has focused on how things work in IT. Through this experience and her research into both project management and IT best practices, Carolyn provides guidance to organizations interested in providing IT services in effective and efficient ways.

1:30 - 2:30
Papers - Novel visualization support
  • Visual Support for Analyzing Network Traffic and Intrusion Detection Alerts using TreeMap and Graph Representation

    Florian Mansmann, University of Konstanz, Fabian Fischer, University of Konstanz,  Daniel A. Keim, University of Konstanz,  Stephen C. North, AT&T Research

    Network security depends heavily on automated Intrusion Detection Systems (IDS) to sense malicious activity. Unfortunately, IDS often deliver both too much raw information, and an incomplete local picture, impeding accurate assessment of emerging threats. We propose a system to support analysis of IDS logs, that visually pivots large sets of Net-Flows. In particular, two visual representations of the flow data are compared: a TreeMap visualization of local network hosts, which are linked through hierarchical edge bundles with the external hosts, and a graph representation using a force-directed layout to visualize the structure of the host communication patterns. Three case studies demonstrate the capabilities of our tool to 1) analyze service usage in a managed network, 2) detect a distributed attack, and 3) investigate hosts in our network that communicate with suspect external IPs.


  • Network Stack Diagnosis and Visualization Tool

    Krist Wongsuphasawat, University of Maryland, Pornpat Artornsombudh, University of Maryland, Bao Nguyen, University of Maryland, Justin McCann, University of Maryland,

    End users are often frustrated by unexpected problems while using networked software, leading to frustrated calls to the help desk seeking solutions. However, trying to locate the cause of these unexpected behaviors is not a simple task. The key to many network monitoring and diagnosis approaches is using cross-layer information, but the complex interaction between network layers and usually large amount of collected data prevent IT support personnel from determining the root of errors and bottlenecks. There is a need for the tools that reduce the amount of data to be processed, offer a systematic exploration of the data, and assist whole- stack performance analysis.


    In this paper, we present Visty, a network stack visualization tool that allows IT support personnel to systematically explore network activities at end hosts. Visty can provide an overview picture of the network stack at any specified time, showing how errors in one layer affect the performance of others. Visty was designed as a prototype for more advanced diagnosis tools, and also may be used to assist novice users in understanding the network stack and relationships between each layer.

2:30 - 2:35
Stretch Break

2:35 - 3:15
Short Papers - Innovative support tools
  • Soramame: WYSWYG Access Control Tool for Mashups

    Nachi Ueno, NTT Information Sharing Platform Laboratories, Ryota Hashimoto, NTT Information Sharing Platform Laboratories, Michio Shimomura, NTT Information Sharing Platform Laboratories, Kenji Takahashi, NTT Information Sharing Platform Laboratories,

    We discuss the design of a novel tool for usable and secure access control named Soramame. We focused on the user problem of creating a mental model of access control. To help better create mental models, we propose using Soramame. Soramame is a WYSWYC (What You See is What You Control) access control user interface. It extracts and visualizes the data-flows of access control policies and uses animation to help users better understand the policies.

  • An Intelligent Contextual Support System for Intrusion Detection Tasks

    Balaji Rajendran, Centre for Development of Advanced Computing, Dnyanesh Pawar, Centre for Development of Advanced Computing

    The cognitively challenging task of Intrusion Detection, undertaken by Network Security Engineers (NSE) even through a sophisticated Intrusion Detection System (IDS), demands considerable effort and time in figuring out the real intrusion attempts. We believe that a support system providing contextual assistance in the task of Intrusion detection would be of immense help to the Network Security Engineers. In this paper, we discuss about the features and mechanisms of such an intelligent contextual support system, along with the implementation and quantitative evaluation of it by using certain observable parameters of performance. The results indicated improvement in the productivity of the NSE and also the usability of the IDS.

3:15 - 4:15
Invited Speaker: Susan Winter, National Science Foundation

"Innovation Ecologies: Toward an Infrastructure for Knowledge Creation"
Abstract: Advances in computer and information technologies have enabled radical transformations in the organization of work allowing distributed, collaborative partnerships and global operations. These same technologies are beginning to enable transformations in the organization of scientific and engineering work.  Research is becoming global and, addressing more complex and integrative problems such as climate change.   Adressing such questions requires shared resources created, maintained and accessed by distributed, interdisciplinary, collaborative teams.  How do we build and support the shared infrastructure underlying such collaborations?  What resources need to be provided?  Who will build them? How will they be funded?  Our future innovative capabilities depend upon our ability to address these crucial questions.
Bio: Dr. Susan Winter is currently a Program Director in the Office of Cyberinfrastructure at the National Science Foundation.  Her particular areas of focus include the study of virtual organizations as socio-technical systems and cyberinfrastructure learning and workforce development.  Dr. Winter is also a Visiting Associate Professor at Portland State University and served on the faculty at the University of North Carolina, Charlotte for 7 years.  She has worked as a project manager at Duke Energy and has over 20 years of international consulting experience.   She has a Ph.D from the University of Arizona and her research on IT and the transformation of work has been presented at national conferences and published in top journals and has won numerous research awards.  She sits on the editorial board of three major journals and is currently a member of AIS and AOM
4:15 - 4:45
4:45 - 5:45
Panel - From the trenches: Are we heading in the right direction?
  • The world of system administration has changed in the last 15 years. We now have SANs, Clouds, and massive clusters. Why are so many of the problems still the same?  What is the next thing and the issues it will bring? 
    • Daniel Boyd, Google
    • AEleen Frish, Exponential Consulting
    • Joseph Kern, Deleware Dept. of Education
    • Daniel Klein, Lone Wolf Systems
    • Tom Limoncelli, Google (Organizer)
6:00 - 7:00
Reception, posters, demos

Affective Pre-Operative Health Care Communication Robot

R. Khosla, La Trobe University, Mei-Tai Chu, La Trobe University, K. G. Yamada, NEC Corporation, T. Yamaguchi, NEC Corporation


Evaluation of Collaborative GIS Usage

Jeffrey D. Campbell, University of Maryland Baltimore County, Michael McGuire, University of Maryland Baltimore County

Learn IT: A tool for search and visualization of IT projects
Maher Rahmouni, Andrew Byde, HP Labs Bristol, Claudio Bartolini, HP Labs Palo Alto

Lessons for IT Architecture from Physical Architecture

William T. Schiano, Bentley University, Leslie J. Waguespack, Bentley University, David J. Yates, Bentley University


Towards Developing Usability Heuristics for Evaluation of IT Security Management (ITSM) Tools

Pooya Jaferian, University of British Columbia, David Botta, University of British Columbia, Kirstie Hawkey, University of British Columbia, Konstantin Beznosov, University of British Columbia


Transforming the IT Workplace by Realizing Personal Online Networks

David Gurzick, University of Maryland Baltimore County

Use of Eyetracking in Usability Testing

Anjali Phukan, University of Maryland Baltimore County


Visual Analysis of NetFlows with VIAssist

John R. Goodall, Secure Decisions


Workflow and Resource Management for Energy-Efficient Supercomputing

*Richard Centauro, International Image System, Kurt L. Keville, MIT, Gaurav Khanna, University of Massachusetts, David J. Yates, Bentley University

Atrium 2nd Floor
7:00 -

Sunday, November 8

8:00 - 9:00
Full Breakfast
9:00 - 10:00
Invited Speaker - Raymond L. Paden, IBM

"Challenges in managing petabyte-scale storage systems"

Abstract: At one time PB sized disk storage systems seemed unthinkable, but today they are increasingly common, even in small clusters.  Vendors are responding by making larger, less expensive disks and controllers that manage over a PB.  Yet center managers and application developers do not appreciate the complexity of these large storage systems.  Too often they approach them as something peripheral rather than integral to the overall processing system.   Part of the problem is that they lack the tools necessary to properly assess their storage requirements (including how different applications interact with each other in their storage subsystem) and to view how their storage profile would behave under different storage products with different characteristics.  So they give detailed attention to other things, but simply ask vendors for a storage system like they had before, only bigger.  This often leads to an overly costly storage system that is often a performance bottleneck.
Bio: Dr. Ray Paden is currently an HPC Technical Architect with world wide scope in IBM's Deep Computing organization, a position he has held since June, 2000.  His particular areas of focus include HPC storage systems, performance optimization and cluster design.  Before joining IBM, Dr. Paden worked as software engineer doing systems programming and performance optimization for 6 years in the oil industry.  He also served in the Computer Science Department at Andrews University for 13 years, including 4 years as department chair.  He has a Ph.D. from the Illinois Institute of Technology in Computer Science.  He has done research and published papers in the areas of parallel algorithms and combinatorial optimization, performance tuning, file systems, and computer education.  He has served in various capacities on the planning committee for the Supercomputing conference since 2000.  He is currently a member of ACM, IEEE and Sigma Xi.  As a professor, he has won awards for excellence in both teaching and research.

10:00 - 10:30
10:30 - 12:00
Papers & Short Papers - Notes from the Field
  • A Case Study of Enterprise Identity Management System Adoption in an Insurance Organization

    Pooya Jaferian, University of British Columbia, David Botta, University of British Columbia, Kirstie Hawkey, University of British Columbia, Konstantin Beznosov, University of British Columbia

    This case study describes the adoption of an enterprise identity management (IdM) system in an insurance organization. We describe the state of the organization before deploying the IdM system, and point out the challenges in its IdM practices. We describe the organization's requirements for an IdM system, why a particular solution was chosen, issues in the deployment and configuration of the solution, the expected benefits, and the new challenges that arose from using the solution. Throughout, we identify practical problems that can be the focus of future research and development efforts. Our results confirm and elaborate upon the findings of previous research, contributing to an as-yet immature body of cases about IdM. Furthermore, our findings serve as a validation of our previously identified guidelines for IT security tools in general.

  • Storage Administration: Field Findings and Software Design Guidelines  (Short)

    Lance Bloom, Hewlett Packard, Nancy Clark, Hewlett Packard

    This paper reports field research of Enterprise Storage Administration. We observed that Storage Administration was complex and not always optimally supported by software tools. The findings highlight Storage Administrator work practices and challenges, and inform guidelines for designing software tools that support Storage Administrators’ work.

  • Self Service Technologies: Transforming a Traditional Call Center into a Virtual Customer Support Center (Short)

    Jakita O. Thomas, IBM Research Almaden, Yolanda A. Rankin, IBM Research Almaden, Neil Boyette, IBM Research Almaden

    Traditional call centers have become one of the primary means for providing customer support. Customer Service Representatives (CSRs) offer personalized attention to each customer via the telephone as the CSR interfaces with technology and customers to resolve customers’ requests. Because call centers are expensive, labor intensive, and produce latency during periods of high call volume, businesses utilize web-based Self Service Technologies (SSTs) to minimize costs, provide readily available resources as needed, and reduce latency while simultaneously serving multiple customers. However, web-based SSTs have their own share of challenges which can impede service quality. To better understand the painful front stage experiences attributed to traditional call centers, we conduct content analysis of interactions between CSRs and customers. Results provide a design framework for web-based customer support SSTs that eliminate the pain points attributed to traditional call centers and address the shortcoming of web-based SSTs.

  • Wiki Anxiety: Impediments to implementing wikis for IT support groups (Short)

    Kevin F. White, University of Maryland, David Gurzick, University of Maryland, Wayne G. Lutters, University of Maryland

    As web technologies have flourished, the workplace has become inundated with new, often-overlapping applications meant to assist busy employees with information management and collaboration. IT departments seeking to implement these systems encounter difficulties in determining which to use. This paper reports on the impediments that arose as a result of the installation of a knowledge sharing wiki in tandem with other knowledge sharing tools within six school technology departments. Analysis of the use and perception of the wiki revealed two prevalent issues: concern over achieving a critical mass of content and anxiety over potential unintended/unexpected content changes.

12:00 - 1:30
Closing, business meeting
Lunch (starts 12:30)