UCAR Information Technology Council, Boulder
CO, September 1997 - September 1998.
The strategic direction recommended for UCAR is to create an information environment that ultimately can support all-electronic operations, effective communications (corporate-wide and external), departmental specialization, and archiving, all with appropriate levels of security. Intellectual artifacts (e.g. papers, software, equipment, ideas, etc.) are among the most important resources and outcomes associated with UCAR's mission and efforts, and this architecture recognizes that such artifacts, along with financial transactions and business documents, are becoming entirely digital at a strikingly rapid rate.
UCAR's strategy for achieving this vision is embodied in eight specific recommendations; primary among these is for the corporation to move toward distributed object-oriented methodologies, to adopt a three-tiered architecture with emphasis on standardized interfaces in the middle, or service, tier, and to adopt widely emerging open standards whenever possible. (The full recommendations are provided in the body of this document.)
The vision and strategy were developed through UCAR-wide workshops (see Appendix A), and by modeling groups who developed case studies of UCAR's varied information dissemination, communication and manipulation needs by functional area (research, education, management etc). These functional sets were derived from preparing a model of how UCAR shares and uses information - research, technical and business - both internally and externally. All of these discussions led, in addition to the specific recommendations, to the very important general conclusion that the handling of these multiple "kinds" of information has significant commonality - this is not to say the information is the same, but rather that the paths of use and transformation in fact are. The importance of this general conclusion is that it provides the rationale for recommending object-oriented methodologies and the three-tiered architecture.
With respect to implementation, UCAR's strategy will be to build incrementally on many efforts across the institution that are consonant with the IT recommendations. This fashion of implementation leverages resources throughout the institution and is likely to be the lowest cost approach. For example, networking as managed in NCAR's Scientific Computing Division, and the plans in UCAR's Finance and Administration group, to examine and meet business needs systematically and in a standard way will be encouraged, expanded where possible, and promoted as examples of better ways to do business.
In times of rapid IT advances, an additional important aspect of UCAR's implementation effort will be to identify and resolve the most pressing "high return" issues first. For example, there is widespread interest, expressed in the workshops, modeling sessions, and e-mail communications from staff, for solving some problems quickly; dominant issues seem to be: transfer of attachments via e-mail, "smart" forms creation, authentication systems, and embedding of business rules within computerized procedures. There is also widespread agreement that a robust, ongoing training program for staff is essential.
The first half of this document describes the rationale for developing
the plan, UCAR's Information Technology vision and its attributes, and
the ITC's eight specific recommendations. For those readers interested
in the underlying technologies, the second half of the plan elaborates
upon those strategies; describes the modeling activities, the system architecture
and implementation approach, and the strategy for immediately advancing
IT at UCAR, specific action items and goals for the next two years, and
some conclusions.
It was becoming increasingly obvious to many at UCAR that common solutions to shared needs should be explored and could contribute to increased ease, efficiency and effectiveness of our operations. This realization was echoed by a recommendation of one of the National Science Foundation (NSF) review panels in mid-1997, in which UCAR was enjoined to seek an organization-wide strategy for information technology management, and to do so in such a way as to be a leader for the community. Accordingly, the UCAR President appointed the Information Technology Council (ITC - see Appendix B for the original charter and membership), with the charge to develop a strategic information plan for UCAR and to promulgate it amongst staff, the NSF and interested community-based advisory bodies.
The goal of this institution-wide plan is to make recommendations on
using and advancing modern information technology to enhance the effectiveness
of all UCAR activities, and provide community-wide leadership.
The ITC chose a unique combination of institution-wide modeling, based on object-oriented methodologies, system architecture evaluations and strategies for implementing and advancing UCAR's IT.
Many UCAR staff participating in these efforts asked questions such
as: "what will this mean for me in my daily work life?", followed closely
by "how much will this cost me?", and "when will this be done?"
The ITC recommendations are:
These methods should be accompanied by object modeling on an enterprise-wide scale. Anticipated benefits include: more rapid and robust software development; software systems which better match UCAR's needs and which are easily adapted as those needs change; greater reuse of existing software components; and better utilization of the distributed computing capabilities provided by our networks.
This architecture allows the separation of both user-interface and data-storage components from the service-provision layer. Hence legacy databases can be encapsulated and made available through the net, and users can more easily perform a wide variety of daily tasks with their favorite applications. A service tier also will facilitate implementing UCAR's business rules and common services (policies, overhead rates, work flows, authentication, etc.) in an appropriately centralized fashion.
A top-quality network is required infrastructure for IT advances and is increasingly essential for UCAR to fully utilize its intellectual resources. This net must support multi-media applications, such as video conferencing, as well as more traditional data flows. Furthermore, the net is becoming UCAR's primary interface to the external world, ranging from basic Web pages and data access, to complex scientific collaborations and distributed computations.
Open standards provide numerous advantages over proprietary (vendor-specific) standards: implementation choices are increased, multi-vendor hardware and software interoperability is enhanced, and the risk of dependency on a single vendor is reduced. Open standards (including TCP/IP, the World Wide Web, etc.) historically have thrived much longer than proprietary ones.
Technological progress at UCAR often has been driven by individual programs striving to fulfill their missions. While common standards must be offered and their use encouraged, individual groups will remain free to solve problems in ways they judge to be best. A suitable implementation framework must be put in place to enable groups to take advantage of institution-wide advances as well as to leverage efforts at the division and program level, where warranted.
UCAR can exert leadership in the IT arena only if staff are conversant with current technologies and have appropriate tools. The rate of change in these areas requires continual learning. Without adequate and ongoing training, adoption of new tools and techniques is unlikely to be successful.
This group should be based on the existing IT Council and include liaisons with the other IT-related committees (DSAC, CSAC, NCAB, etc.), with careful selection from UCAR's research, technical, and managerial staff. This group would: further refine the IT strategy; oversee the development of service layer (and other) standards; and coordinate UCAR's various IT efforts. The committee should report directly to the UCAR President.
Key services--such as authentication--which must be developed and administered centrally, would be handled by such groups. Additionally, responsiblity must be assumed for: refining the enterprise-wide object model; developing, disseminating, and maintaining a library of reusable classes and other software to support the model; and addressing other IT-related issues that concern the entire institution, such as e-mail with attachments or tools for network conferencing and collaboration.
UCAR also should consider appropriate administrative structures for the new engineering groups. While a single administrative entity may eventually be desirable for these groups, the present recommendation is to allow a distributed management structure as indicated in a later section. This compromise represents a practical approach to allow existing engineering groups to continue operating without any significant disruption and at the same time, to develop new groups supported by divisions or programs where needs or opportunities arise.
Object modeling seeks to create a description ("model") of UCAR and its operations in terms of "objects." Objects are simply "things" that we encounter as we carry out our business. They can be physical items (i.e. "desk", "radar", or "office"), people and the roles they play ("employee", "software engineer"), or logical concepts ("encumberance"). Objects can also represent the processes carried out by UCAR ("operate an observing system," or "buy a new computer").
Describing UCAR in terms of its component objects has a number of advantages. Objects hide within themselves all the complexity of the thing they represent. Just as one can submit a purchase requisition without really knowing what will happen to implement it, one can ask an object to carry out a desired task. Objects can be defined in a hierarchical manner which makes it easy to specify what two objects have in common, and how they differ. Thus, the potential exists to describe financial and scientific data streams in the same manner, and to use many of the same tools to work with them. Finally, since objects form a model of UCAR, any software which implements these objects also models UCAR. That means that the structure of our software matches the structure of our organization. The result is software which is better suited to UCAR's needs, and which is more adaptable as those needs change in the future.
Object models are often developed using the "Use Case" method, whereby particular day-to-day tasks are selected and studied in detail to determine all organization, resource and process components as well as identifying roles that people or computers play, and their key responsibilities.
The highest level of the model (shown below) was constructed using the most distinct set of entities with which UCAR has relationships. Its purpose was to set the context for the modeling exercise and assist participants in understanding our broader audiences. These factors are important since a particularly useful part of the modeling exercise is the consideration of the extended enterprise and the extent to which relations, collaborations, and interactions with entities at the global level can be modeled effectively, then improvements in internal IT strategies can be logically extended outside of UCAR. The ITC was also motivated by the strong community orientation that UCAR has. At the completion of the modeling exercise, the global level model was revisited and checked for consistency. It was found that it still provides a suitable basis for future modeling efforts.
To accommodate the very broad range of activities at an organization the size and diversity of UCAR, as well as to eliminate the effects of organizational boundaries, the ITC decided to investigate eight high level groupings based on the actual functions being performed. This approach allowed the working groups to uncover potential functional commonalities in distributed parts of the organization, and assisted in the identification of hierarchies, which were later used in the object modeling exercise.
Thus, UCAR's world was divided into functional areas (shown below), which then led to the creation of working groups to deal with the Use Cases, from which commonalities were identified. UCAR's science, technical and business information "world" is populated with people and looks like the following:
All of these aims were achieved. The overall impression from the consultants was that UCAR would be in the top few percent of all organizations undertaking such a comprehensive planning effort. Additionally the working groups built a representative list of Use Cases and their findings provided a strong basis for the Architecture and Implementation planning efforts.
One surprise among the findings was the identification of a startling degree of commonality across the organization: e.g. data analyses for science and finance follow the same process and have the same bottlenecks and problem areas; aggregation of scientific information for general audiences is required in almost every part of the organization,; and strong access authentication for scientific and business information was widely needed.
Another significant aspect of the modeling exercise was the assembly of a list of ``roles'' that people (or computers) are asked to perform in the process of complete specific job tasks (or Use Cases). A complete list of these roles can be found in Appendix C. When combined with an object modeling hierarchy, roles that people play are a resource object for the organization.
To assist participants in the modeling exercise and for readers of this plan, it is useful to know that an object represents a real or an abstract thing, with a name, e.g. Plan, Regulation, Budget, Report, Model, Presentation, Purchase, Approval, Team, etc. Furthermore, objects have hierarchies, e.g. Plan.Strategic, Report.Research.Annual, Presentation.Oral and instances of the class contain the values, e.g. Plan.Strategic(``IT''), Report.Research.Annual(``NCAR''). Objects communicate via messages which are a basic tool in a distributed computing environment and this directly affects choices that were made for the architecture of the information system.
Our strategy is to engage in and encourage such change, both internally exploiting the power of desktop computers and departmental expertise to meet specialized needs and externally enhancing UCAR as a focus for efforts that build on the aggregated resources and expertise of the global geoscience community.
Guided by the model working group discussions and outcomes, several meetings were held on system architecture and implementation strategies.The overall aim for this section is to present key technologies and organizational changes that can create an information environment that ultimately can support all-electronic operations, effective corporate-wide and external communications, departmental specialization, and robust security and archiving.
Also known as the "presentation" tier, this category encompasses user interface software, ranging from general-purpose tools, such as browsers and spreadsheets, to more specialized systems, such as for requesting a check, analyzing a scientific dataset, or placing an order.
Also called "applications services," "interoperability layer," or "mediation layer," this tier will foster coherence among the decentralized and generally autonomous units that comprise UCAR's information system. Web servers are the most prominent components of this tier, and their primary purpose--implementing a standard protocol--is typical: Tier-2 components "encapsulate" a service with a known interface and published semantics, and the software to perform these functions often is shared and reusable. Encapsulation facilitates creating user interfaces because it hides variations among and complexities in the underlying (Tier-3) databases. Tier 2 components eventually will encapsulate "business rules," including indirect cost formulas, work-flow definitions, and authorization hierarchies, obviating the need for coding and updating these rules directly in every system where they apply.
Also known as "data management and distribution" or "foundation layer," this tier encompasses all of the (specialized) systems used to acquire, store and maintain various types of data. Tier-3 functions range in complexity from simple file retrieval to concurrent, read-write database access. Accessing Tier-3 systems via Tier-2 services, such as viewing a collection of files or a DBMS through the Web, is a means for creating a degree of uniformity across otherwise dissimilar data sources.
UCAR's newest standing committee, the Desktop Systems Advisory Committee (DSAC: see Appendix D for a description of UCAR's current standing IT subcommittees) has gathered crucial information and is preparing recommendations on a strategy for improving corporate-wide e-mail capabilities, including specific software options and conventions for "attachments." The ITC will evaluate these recommendations and present them to the UCAR Management Committee or President's Council for action.
The network is increasingly essential for UCAR to fully utilize its capabilities. To do so, UCAR networks must support multi-media applications, such as video conferencing, as well as more traditional data flows. Furthermore, it must serve as UCAR's primary interface to the external world, with services ranging from basic Web pages and data access, to remote scientific collaboration and distributed computing.
Two of UCAR's standing IT committees, the Network Coordination Advisory Board (NCAB) and the Computer Security Advisory Committee (CSAC), play crucial roles in meeting the often overlapping and occasionally conflicting needs of the corporation for state-of-the-art networking with adequate security. It is also likely that the DSAC recommendations will have security and networking implications. The ITC will create a forum for debating matters of common concern to these committees, in the context of an overall IT strategy. It also will create an avenue for recommendations on these topics to reach the UCAR Management Committee and/or President's Council as needed.
Complementary to concurrent efforts in Finance and Administration (F&A), the ITC will launch and oversee efforts to define and implement electronic business rules and workflow mechanisms on an institution-wide scale. This effort likely will include: authentication; calendars and scheduling; automated workflows for travel, purchasing, time cards, personnel data, scientific reviews; associated rules, such as authorization hierarchies; and other business rules, such as authorizations to modify/access data, spend money, control an instrument, etc.
To these ends, the ITC will create or recommend the following:
The ITC will create a standing committee to advise F&A and the engineering group on the computerization of workflows and business rules at UCAR, including matters of priority and potential funding sources.
The ITC will recommend creating an engineering group, guided by the Business Rules Advisory Committee, to be responsible for service-layer capabilities needed to automate business oriented workflows, including authentications and authorizations. The BRAC, perhaps by creating a task force, will help develop specifications for these capabilities via models and use cases, per the IT Strategic Plan.
UCAR must position itself for applying distributed computing methods to major challenges facing the geosciences in the next decade. These challenges encompass needs for increased resolution, scale interactions, cross-disciplinary model interactions with feedback, and investigator collaboration utilizing advanced visualization and analysis tools. Solutions will require various forms of problem segmentation, messaging, load balancing, and interface specification, some of which may permit large calculations to be run as community endeavors. Observing systems, especially those that are remote or comprise many instruments at diverse locations, also may be enhanced by UCAR-led efforts in distributed computing. Strategies for distributed computing will be developed by a new task force that investigates object-oriented approaches, including Java.
Growing amounts of information published via the World Wide Web--coupled with the diverse nature of UCAR's organizational structure, programmatic activities, and technical expertise--have made organizing, coordinating, and finding that information increasingly difficult. Development of a consistent approach to site architecture, navigation, and search capabilities to facilitate access and retrieval is a high priority need identified by numerous visitors to our WWW sites as well as by participants of the Information Model working groups. At its June 1998 meeting, the President's Council approved an effort to investigate design, searching and architecture issues. A Web Advisory Committee will guide this effort and the ITC will coordinate it with related IT activities.
Appropriate and timely education and training are critical to the IT strategy and these must be provided both for IT professionals, to ensure their technical currency, and for users, so they can take maximum advantage of the tools available. Training must encompass unique applications developed at UCAR as well as commonly used commercial packages.
To this end, the Human Resources department of F&A is establishing a corporate IT training facility, and guidance will be provided by the ITC. This facility will be dedicated to IT training and equipped with appropriate hardware, software and network connectivity. UCAR has identified space for this facility and is in the process of identifying equipment needs.
The restructured ITC will have representation from the original members involved in the planning stages, and from existing IT committees that have focused on implementation issues for key topics that are relevant to the entire organization. Indeed, the most important IT-related resource at UCAR is the aggregated expertise of the staff, and much of this resource already has been organized and focused on specific IT topics through various UCAR standing committees (NCAB, CSAC, DSAC, LDHwg). The ITC will be able to quickly consider new and changing requirements within the organization, and make decisions to allow UCAR to move forward rapidly. The ITC will consult with and inform the UMC and President's Council on IT related matters. As a result of this action item, UCAR will have a high level technical focal point for all IT related issues and concerns.
Membership of the restructured ITC initially will be: Jack Fellows (chair), Harriet Barker, Lawrence Buja (LDHwg), Jon Corbet, Mike Daniels/Tim Fredrick (DSAC), Peter Fox, Dave Fulker, Karon Kelly, Rebecca Oliva, Pete Peterson, Katy Schmoll, Pat Waukau (NCAB), and Greg Woods (CSAC). The ITC will consist of members from the UCAR staff (appointed by the chair for terms of two years, which may be renewed) and the chairs of the standing IT comittees (whose terms are determined by the committee's rules). Additionally, the ITC will retain at least one member who is also a member of the UCAR President's Council.
The IT Council will report to the UCAR President and be charged:
In this diagram: The top layer denotes the high level advisory and policy making Councils. At the level below the ITC are the IT advisory committes; DSAC, CSAC and NCAB are existing committees (blue), BRAC is the proposed Business Rules Advisory Committee (green), IMAG is the Information Model Advisory Group (for refinement of the UCAR-wide model efforts and specific science and data applications; red), WAC is the Web Advisory Committee (green). The engineering groups are listed at the next level, including existing groups for security, networking and training, as well as proposed groups for business rules, desktop applications, modeling and web development. In the lower layer, the administrative or supervisory entities associated with each engineering group are indicated. Not all of the entities shown have been defined or agreed upon.
To enhance effectiveness in the context of such distributed management, especially in cases where concerns and responsibilities overlap, the ITC will promote feedback and interaction among the engineering groups and advisory committees. The ITC also will seek to ensure adequate participation and representation from all parts of UCAR to capture the full range of needs that must be meet, including those of the external community.
NASA is funding a collaborative effort involving HAO and Unidata plus the University of Rhode Island, MIT, and many other external players to employ the Distributed Ocean Data System (DODS) at UCAR and to support its use by a large community of scientific data users. The DODS technology which is not specific to ocean data makes exemplary use of the three-tiered architecture, permitting a variety of clients (including widely used packages, such as MatLab and IDL) to access heterogeneous, remote databases without knowing their formats.
Future development of DODS is planned to incorporate OO methods into the process of discovering and using relevant data from the immense but often unwieldy world of Internet-accessible resources.
COMET, PAGE and Unidata are developing a common strategy to support the sharing and use of information about people, organizations, and resources within their communities. This project is investigating an object-oriented approach to community-wide information sharing across a very broad range of content types. The outcome, a set of object-oriented community-information tools (OOCIT), potentially will provide integrated support for names and addresses, committee membership lists, special-interest mailing lists, e-mail archives, and pointers to on-line educational materials and other community resources. An initial set of requirements and design documents, developed using object models, already has been produced.
The objectives for near-term application development are:
Consistent with the Strategic Plan, the ITC will consider following the OOCIT pattern to establish a science-oriented modeling effort for all of UCAR. (This would continue the modeling work begun last year by the ITC.)
The ITC will monitor other object oriented development projects underway in ATD, JOSS, SCD, HAO, and elsewhere as they emerge.
Though funding has yet to be gained from these sources, the NSF and other organizations are agressively funding innovative advances in IT. The ITC will encourage the writing of proposals to tap these sources for UCAR's benefit.
As an example, one recent UCAR proposal to the NSF's KDI initiative titled A Geoscience Knowledge Community and submitted to the Knowledge Networking (KN) component of KDI aims to substantially accelerate the accumulation of knowledge across a diverse, worldwide network of researchers, educators, and students in geoscience and related fields. The proposal focuses on digital representations of scientific observations, simulations, and graphical depictions of earth- and sun-related phenomena and systems. At present these embody an unwieldy and highly dispersed collection of datasets, computer programs, and images on varied media. Their use and integration across discipline boundaries, even within the geosciences, is painfully tedious. The goal is to increase the coherency and hence the utility of these digital resources, without limiting their dispersion. The proposed method will be to provide a growing, networked "geoscience knowledge community" with new software and services that facilitate the creation, use, and sharing of digital scientific artifacts. In the hands of the diverse scientists, educators, and students who are connected to UCAR, this will yield benefits well beyond the geosciences.
The objectives are organized around a 3-tiered architecture to enable greater variation in user interfaces (user tier) and modes of data collection, processing, and storage (data tier). The implicit many-to-many connections between users and data in this scenario will be achieved economically by developing (service tier) software tools and servers that implement well defined interfaces and perform common data handling and processing functions. This design will facilitate creating "thin client" applications that are easy to use even though they provide powerful visualizations, perform complex data analysis, and access and integrate diverse types of data from multiple locations throughout our geoscience knowledge community. Clearly this effort, if funded, will advance the objectives articulated in UCAR's IT Strategic Plan.
The ITC, with extensive staff participation, has developed a Strategic
Plan that promises important and beneficial IT advances for all of UCAR.
This is a non-trivial management challenge, and the goal of the ITC is
to oversee, inform, advise, and guide the corporation on this important
path. In doing so, the ITC also identified issues needing immediate attention
and organizational approaches that appear effective for resolving them.
The ITC encourages feedback and exchange of information with respect to
this plan with its community, oversight committees, and sponsoring agencies.
The ITC recommends that this strategic plan be re-evaluated in three years.
The workshop was structured in four parts; (1) two keynote addresses to discuss, in conceptual terms, the challenges of implementing business and science IT strategies, (2) an open discussion to identify relevant information needs, successes, and challenges at UCAR, (3) presentations on new IT methods characterized as "middleware," and (4) four parallel breakout sessions to discuss what was learned during the morning sessions and to consider the draft IT principles. The workshop also included a presentation by Cliff Jacobs of the National Science Foundation (NSF) on the importance of an IT strategy for UCAR. Cliff stressed a community leadership role for UCAR regarding IT, and he pointed out the potential relationship between such a role and the NSF's new Knowledge and Distributed Intelligence (KDI) initiative, which seeks "to develop networks across disciplines via heterogeneous databases."
The workshop was effective, with participants gaining an overview of the importance of an IT strategy, hearing about design and implementation opportunities and pitfalls, learning about one another's information issues, and gaining a rough sense of how these issues might be addressed with new technologies. The organizing group got excellent feedback in the breakout sessions on the draft IT principles, as described in the next section.
The full Workshop
Summary is available online.
FROM: Rick Anthes
SUBJECT: Information Technology Council (ITC)
Attached you will find the charter for the ITC. I would like to thank all of you who participated in the September 9, 1997 Information Technology Workshop. The feedback provided at the workshop will be extremely important input to the ITC's activities. I would also like to thank the members of the ITC in advance for agreeing to participate in this important process.
It was challenging to select the membership of the ITC. There were many people who could serve effectively on this council. I felt it was extremely important to keep the number of members small, yet ensure that its membership reflected the broad range of UCAR/NCAR/UOP views. I know that the ITC members will be drawing heavily on UCAR-wide staff during its activities. My hope is that the ITC will be able to develop a draft strategy for our collective review in the early December timeframe.
Information Technology Council (ITC)
ITC Charter. At the September, 1997 UCAR Information Technology Workshop (ITW), a cross section of UCAR staff identified important principles to guide the creation of a UCAR-wide information technology (IT) strategy. The function of the ITC is to develop a roadmap for gaining a robust and flexible IT systems that employs these principles and serves UCAR (i.e., UCAR, NCAR, and UOP) and its associated communities effectively, now and for years to come. Components of the strategy will include:
Information Model (IM). The ITC will create an information model--reflecting UCAR's organizational structure, its processes, and resources consumed and generated--as the basis for enhancing UCAR's information systems, emphasizing institution-wide IT services and practices that will help UCAR and its members meet their goals (i.e., science data, administrative functions, libraries, etc.). The IM will employ a broad, multi-media definition of information.
System Architecture (SA). The ITC will identify candidate IT architectures, components, and tools with which to design, construct, and test enhancements to UCAR's information system, according to the IM and the ITW principles. This effort will incorporate lessons learned by organizations that have implemented institution-wide IT systems on scales comparable to UCAR's.
Implementation Process (IP). The ITC will recommend how to implement the IT strategy--including responsibilities, schedules, and incentives, as discussed in the ITW--within reasonable resource constraints.
ITC Structure and Schedule. The IM will be defined first at the macro level and then successively refined to embody greater levels of detail across the entirety of UCAR. After an initial model, which identifies the most crucial needs at present, successive refinements will be realized through design, construction, and testing activities. In this way, change will occur incrementally, address the most immediate needs, and inform the model-refinement effort. Thus the IM is the linchpin for the IT strategy; at each stage the IM will clearly and simply describe the current understanding of UCAR's information needs, and it will form the basis for subsequent SA and IP activities. As the model becomes more detailed, it will show increased specialization, such as distinctions between systems for science data and business data.
The ITC will be structured--with co-chairs for IM, SA, and IP--to initiate this cyclical process, and most of its early effort will be focused on developing the initial IM. The ITC will report to the UCAR President and abide by the following schedule:
Jack Fellows (chair), Peter Fox (IM co-chair), Rebecca Oliva (IP co-chair),
Dave Fulker (SA co-chair), Harriet Barker, Katy Schmoll, Bill Buzbee, Karon
Kelly, Jon Corbet, Jim Hurrel, Steve Williams
The NCAR/UCAR Network Coordination and Advisory Board (NCAB), reports to the Director of SCD and is chartered to assist SCD with network planning and management. The scope of the committee will encompass the performance and function of the network cabling plant and all associated network hardware. In particular, the NCAB will assist SCD in:
The NCAB will have regular meetings (to be determined by the group) for briefings on new networking initiatives, updates of ongoing networking projects, and discussion of networking issues and problems. SCD will keep minutes of each meeting that includes an ongoing, prioritized, and categorized list of subjects for committee discussion. The NCAB meetings will alternate between the Mesa and Foothills Labs.
The NCAR/UCAR Computer Security Advisory Committee (CSAC) reports to the Director of the Scientific Computing Division (SCD) and is chartered to assess the state of computer and network security at NCAR/UCAR and to make recommendations to assist NCAR and UCAR management in setting policies related to the security of computers and other devices attached to the NCAR/UCAR network. This committee works to build consensus on a security policy which, in turn, impacts the entire NCAR/UCAR organization and requires compliance from both SCD and other NCAR/UCAR entities to implement.
In particular, the CSAC:
The CSAC holds regular meetings for briefings on new security threats or new problems caused by security restrictions, updates of ongoing security work, and discussion of security issues and problems. SCD keeps minutes of each meeting that includes an ongoing, prioritized, and categorized list of subjects for committee discussion. The CSAC meetings alternate between the Mesa and Foothills Labs.
DSAC will explore, discuss, and recommend ways that UCAR organizationalunits can improve the quality, successful utilization of, and cost-effectiveness of the desktop computing environment.
With respect to Desktop computers and the Desktop computing environment, DSAC will consider:
NCAR's goal for handling large, scientific datasets is to enable access with techniques that are easy, efficient, and practical for both data providers and users. To achieve this goal, NCAR will work closely with data providers and data users to evolve toward a data handling environment that includes:
Some specific steps are identified for the next two years. Some of the shorter term of these have been realized (reconstitute the ITC, for example); others are yet to be tackled (carry out survey of existing IT systems, including of non-UCAR groups, develop specific cost-benefit analysis). The original steps were: