REPORT ON INFORMATION HANDLING FACILITIES FROM WORKING GROUP I - PHASE I IHS STRATEGIC PLANNING EFFORT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP86B00689R000300040010-3
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 23
• Document Creation Date: December 20, 2016
• Document Release Date: November 14, 2007
• Sequence Number: 10
• Publication Date: January 19, 1982
• Content Type: REPORT

Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
INFORMATION HANDLING FACILITIES
FROM
WORKING GROUP 1 - PHASE I
January 19, 1982
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
Information Handling Facilities Report
Sect ion
Introduction
Conclusions
2 Terminal Facilities
Introduction
A Terminal For Most Agency Employees by 1990
Terminal Estimates
Office Automation
One Terminal-Many Systems
Planning for the Evolution of Terminals
TEMPEST - A Deterrent to Low-Cost?
Decentralization via Intelligent Terminals
and Office Automation
Conclusions
Introduction
Ruffing Center
CAMS
SAFE/ADSTAR
Do Information Systems
Data Base Growth
Conclusions
Dissemination
Network Facilities (MERCURY)
Services
Conclusions
5 Workshop Participants
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
SECRET
SECTION 1
SUMMARY
INTRODUCTION
The plan for developing the Strategic Plan for the Agency's
information handling system is comprised of four phases. In the first
phase a series of working. groups, with a user orientation, were to
meet to address and define strategic. planning goals; these goals are
to provide the basis for implementation strategies to be formulated in
later phases. This report is the product of the Phase I Working Group
concerned with Information Handling Facilities.
Information Handling Facilities were considered in three parts:
o Terminal Facilities (including Office Automation Equipment)
This report includes a section on each of the three types of
facilities, describing the highlights of the Working Group proceedings
and presenting conclusions.
The conclusions in the sections include planning goals which e.ze
extracted and restated in the following paragraphs as recommended
goals.
CONCLUSIONS
Recommended planning goals are as follows:
Goal 1-1 Provide an electronic workstation (terminal) for most Agency
personnel by 1990. The goal is to have installed by 1990
approximatel computer terminals (Delta Data class
device) an Office Systems (word processors and
distributed EDP systems).
The Working Group did a rough survey to provide the
estimates for the numbers. The DO and NFAC estimates
reflect the concept of one terminal for every individual,
whereas the other directorates were more conservative,
estimating on the order of 1 terminal for every 4 individuals.
Goal 1-2 (a) Standardize on terminal characteristics to put the Agency
in a favorable position for,procuring from a competitive
marketplace.
(b) Standardize on system and terminal interconnect protocols
to provide for maximum flexibility and interoperability.
25X1
.25X1
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
the Agency must be sensitive to aligning its requirements
with a large sector of the marketplace; buying mass produced
devices in a competitive situation can mean the difference
of millions of dollars and whether or not Goal 1-1 is
affordable.
The level of interconnect and interoperability that
will be required between terminal-to-terminal and
terminal-to-host can be attained only through
standardization of interconnect protocols.
Requirements for one terminal connecting to multiple
hosts were expressed repeatedly by Working Group members.
Goal 2-1 Provide Data Base Management Systems (DBMS's) which are
at the forefront of technology, and which have the highest
potential for remaining at the forefront over the long term.
The creation, maintenance, and processing of data bases are
among the most important of Agency information system
capabilities. The DBMS is the cornerstone of that capability,
providing the primitives and interfaces upon which many
applications are dependent; Query Language, Data Dictionaries,
Program Generators, Report Generators, and other tools evolve
from and are tied to the DBMS. User friendly characteristics
are mandatory. It is essential that the Agency be at the
forefront of this technolgy.
Goal 2-2 Provide administrative procedures and system information
facilities to better manage one of the Agency's most valuable
resources -- information.
In years to come the Agency's information resources will expand
rapidly, more of the information-will, be in electronic form,
there will be a need to process the same information in
various systems, and information will, where appropriate,
be shared and utilized more widely. Administrative procedures
and systemacilities must be designed for this environment;
dictionaries describing information resources must be automated;
and standards must be promulgated for the interchange of
information.
Goal 3-1 Define the system architecture for information dissemination
network.
Dissemination is one of the Agency's more advanced information
handling applications requiring multiple systems to function
together in a network. Architecture tenets are needed to
guide system designers in treating network dependent design
variables.
Goal 3-2 Define standards for network information exchange.
The proliferation of cable formats and its impact on the
cost of information handling systems exemplifies the need
for standards. As new modes of information exchange are
introduced, the need for standards will be even greater.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
Goal 3-3 Define requirements for local area networking needs and
develop a plan for a uniform system of local networks.
There will be increasing needs to link office systems and
data processing systems in local areas. These local networks
will link into wider area networks; to avoid great difficulty
and cost in providing this connectivity a uniform approach
is needed.
Goal 3-4 Develop tools. and methods for specifying performance
objectives and validating the design of information
system networks against those objectives.
In a complex network it is extremely difficult to achieve
an adequate level of assurance that the performance objective
of the individual systems, and the network as a whole, will
be met. It is even difficult to specify the objectives.
But we must begin to apply methods which will give more
assurance that systems will support the operational needs
when deployed. Information services must be characterized
as to the performance requirements they will impose on
underlying facilities. Facilities must be analyzed and
modeled to validate that they will accommodate the
requirements.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
SECRET
Section 2
Terminal Facilities
INTRODUCTION
The Working Group considered terminals in two general classes:
o Class I - an intelligent interactive terminal of
the Delta Data 7260 class.
o Class II - an integrated office automation system of
the type currently being procured (standard
word processor in its most basic configuration).
The Working Group was challenged to provide: estimates on numbers
of terminals by class to be installed through 1989; characterize the
usage of terminals; develop planning. assumptions for distributed
viersus centralized functionality; and determine deficiencies in
current terminal programs.
Discussion points regarded as most significant are summarized in
following sections, ending with conclusions.
A, TERMINAL FOR MOST AGENCY EMPLOYEES BY 1990
Serving as a basis for the discussion, the IHSA Point Paper and
other planning documents, project a goal that most Agency employees
will have an electronic workstation (terminal) for his personal use by
the end of the decade. The IHSA Paper goes further to estimate that
terminals, at a cost of more th will be
required to meet this goal.
It was suggested that there are classes of employees that will not
require terminals. Examples given included FBIS newspaper readers,
laboratory technicians, and certain field personnel. Yet, there was
consensus that a large portion of the Agency's population would indeed
require a terminal by 1990; the actual estimates developed by the
Working Group and presented in the following section confirm this.
There was considerable discussion over whether the request for
estimates of terminals, by organization, through the eighties, was a
reasonable request; and if so, what the assumptions were to be in
making such estimates. It was pointed out that the experience has
been that budgetary constraints determine the number of terminals that
are procured, quite independent of stated requirements; ODP in
'immediate years has funds for less than 30 percent of stated
requirements. The counter argument was that a well conceived plan
will be necessary to lift budget constraints. It was eventually
agreed that Working Group members would survey their home
organizations; the assumptions used and the estimates are presented
below.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
TERMINAL ESTIMATES
A summary of the estimates provided by each Directorate is
provided below in Table 1. The more detailed data is included in
Appendix A.
The total number of terminals (excludes NPIC and other
requirements outside the CIA budget) for 1989 is estimated at
devices; this includes both Office Systems (Class II and conventional
computer terminals (Class There are currently devices
installed meaning thatew terminal installations will be
required over the next eight years to meet the estimates; in addition,
many of the currently installed devices will have to be replaced.
It is anticipated that the installation rate will gradually
accelerate, reaching a peak in the 1985-87 time frame when major
system developments become fully operational and when Office
Automation has realized its potential.
It is noted that DO and NFAC fully subscribe to the idea of one
terminal for every individual, whereas DS&T and DA are more
conservative in their estimates. The reasons can be attributed to the
charter of the organizations; these organizations are more involved in
the substance of intelligence, particularly analysis, and therefore
see greater individual involvement with information handling systems.
Those involved in administrative and other support processes, such as
program management, see information handling systems as being less
integral to their working environment. Perhaps influencing this
difference in per..pective is the emphasis to date on systems in the
operational domain, rather than the support domain.
25X1
25X1
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
Office Automation was considered as those functions described in
the current RFP for Office Automation equipment. Included are:
o Word Processing
o Document Storage and Retrieval
o Electronic Mail
o Records Management
o Local EDP Capabilities
o Local Networking for Office Interconnection
Working Group members were supportive of Office Automation
c ncepts, believing that the technology has enormous potential for
improving information handling processes. As a group, there was clear
understanding of the application and need for word processing, and for
communicating word processors - this technology is widely used in the
Agency today.
There is less understanding of the other aforementioned Office
Automation functions and how those functions are to be exploited. The
Office Automation equipment which the Agency is about to procure
offers the potential for electronically connecting every desk in the
Agency, for major reduction in paper dissemination and paper files,
and for data processing in the office environment. But, before this
piptential can be fully realized there are questions to be answered,
such as: What are the applications of electronic storage and
retrieval of documents in the office environment--is it cost
effective--is it secure? What are the applications of electronic
mail--is electronic mail to become pervasive in the Agency's conduct
of its business? What EDP applications are appropriately performed at
the office level?
As with-any significant new systems technology, it will take years
to understand and fully exploit office automation. Answers to
questions will come as we deploy office systems for limited
applications, such as word processing, and progress to more
sophisticated applications as we learn.
It is clear that the impact of office systems will be pervasive:
In the systems domain EDP Systems and Communication Systems must
interface and provide interoperability; in the office domain
procedures for conduct of business will be altered. What this
strongly suggests is that there be increased emphasis on planning for
the application of Office Automation Systems.
Other discussion points were:
o Each of the major computer centers offer, or are developing,
functions which are equivalent to those provided by Office
Systems--electronic mail and word processing being the prime
examples. It is assumed that there will be requirements for
an end user at an office system to exchange mail, or a text
file, with an end user interfaced to SAFE or the Ruffing
Computer Center. NFAC, for example, will use SAFE, VM, and
local office systems for mail and word processing; it is
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
clear that information must pass between these facilities,
and further, that a degree of universality in the user
interface is required.
o The primary reasons for the attractiveness of office
systems as compared to central systems with the same
function, are availability,
responsiveness, and user friendliness. The balance
between distributed office systems and central systems
will depend heavily on improvements to central systems
in these areas.
o It was suggested that the justification for procurement of
terminal equipment should be in the form of a broad based
plan for automating the Agency, rather than justifying
the equipment in small lots, as is the current practice
for word processors.
o There was limited discussion of other concepts often
considered under office automation, such as, voice mail,
teleconferencing, and voice I/O.
ONE TERMINAL - MULTIPLE SYSTEMS
Throughout the discussion there were statements to the effect that
the end user must have access to multiple systems through a single
terminal. Examples include:
o NFAC users must have access to both SAFE and the RCC.
In addition, as a longer range goal, SAFE users will
be provided the capability to access community data
bases and systems.; DIA and NPIC were specifically
mentioned as having data bases that the NFAC analyst
would benefit by having access to.
o The DO has a one-way link to the RCC.
o NPIC has a requirement to provide gateways from NDS
to SAFE, and other community systems.
There was discussion of regional printers and whether they should
be accessible through a network by both SAFE and the RCC, or whether
they should be dedicated to one or the other; there was concern
expressed that the user would not be properly serviced in a networked
environment.
PLANNING FOR THE EVOLUTION OF TERMINALS
During later phases of the strategic planning effort, it is
expected that a terminal strategy will be formulated. Discussion in
the Working Group pertinent to this strategy is summarized in the
following paragraphs.
Given the "best of all worlds", the best terminal solution for the
Agency would be a family of terminals, configured from common and
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
modular building blocks, covering the least to the most complex
configuration. The distinction between an ADP terminal (DD 7260) and
a communicating Word Processor would be only in the software, with
both functions being available in a properly configured terminal.
It was anticipated that Delta Data would be replaced beginning in
the mid-80's; the new terminal would conform to the family concept,
and offer more processing, storage, and software support. There was
discussion about what impact that might have on current systems where
the software tends to be coupled rather tightly to the terminal; the
upshot was that system designs should be moving in the direction of
implementing standard protocols which would ease the way to
interfacing new terminals.
A representative from ODP indicated that there are initiatives
underway to enhance the Delta Data as a local processor; this may
eventually result in the micro operating system, CP/M, being available
along with some of the packaged software that runs in that
environment. In addition, there is an initiative to enhance host
software to make the graphics capability of the 7260 useful.
TEMPEST - A DETERRENT TO LOW-COST?
With the large quantities of terminals projected for the eighties,
cost was a recurring issue. The Agency currently pays approximately
$6000 for a minimum configured Delta Data terminal; some thought it
should be much less, as little as $3000. When considering thousands
of terminals the differential is attention getting.
There is clearly enormous cost leverage in procuring terminals
which are mass produced for a large marketplace. The experience,
however, has been that the TEMPEST requirement has been instrumental
in forcing the Agency into a special and relatively small marketplace,
and therefore a costly marketplace. The result is that-the TEMPEST
requirement can as much as double the cost of the equipment. It
should be said that TEMPEST has not been the only special requirement,
but it is the most intractable and one which will be with us in the
Yong term. Perhaps a coordinated effort to focus and stimulate the
marketplace would result in lower costs.
DECENTRALIZATION VIA INTELLIGENT TERMINALS AND OFFICE AUTOMATION
As terminals expand in storage and processing capability, and as
Office Automation Systems are introduced, the Agency will surely do
more of its automated information handling functions in the office
environment. There will be an overlap of capability between central
and office systems, and choices involving complex tradeoffs will have
to be made. Some of the tradeoffs are likely to be similar to those
confronted in the past over the deployment of minicomputers.
It was suggested that a hybrid configuration offering both central
and local facilities to the end user is an opportunity to take
advantage of the best of both. The lines of demarcation are not
spelled out, and perhaps it is too soon to do so, except in the most
general way.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
Most members of the Working Group foresee a highly connected
network of central computer systems and office equipment. There was
concern over-the security of such a network; a view was expressed that
certain organizations would not allow their office information
handling facilities to be connected to a network because of security
considerations.
What is clear, is that a network of computer systems and office
automation requires a high degree of interoperability where files and
other information flows between systems and can be processed
interchangeably, and where functions such as mail transcend system
boundaries. Standard protocols, addressing the lowest communication
protocols to the highest application levels, are essential to achieve
such interoperability.
CONCLUSIONS
Conclusions are derived from the above sections stated as planning
goals. These goals are to be reduced to more specific objectives as
planning progresses through subsequent phases.
Goal 1 - Provide an electronic workstation (terminal)
for most Agency personnel by 199 Tie
estimates call for approximately computer
terminals (Class I devices) and office
systems (Class II devices) by 1990.
Goal 2 - (a) Standardize on terminal characteristics so
that the Agency procures from a competitive
marketplace.
(b) Standardize on system and terminal
interconnect protocols to provide for
maximum flexibility and interoperability.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
INTRODUCTION
It is anticipated that the eighties will bring a significant
increase in demand on data base facilities. There will be
requirements for more and larger data bases, for greater accessibility
to data bases, and for tools to allow effective utilization of large
quantities of data.
The Working Group was challenged to develop planning goals and
objectives which are responsive to this anticipated demand. The Group
was tasked to provide: estimates for quantities and sizes of data
bases; define data base related tools; provide estimates for community
data base access; and review plans for current large data base
facilities.
There was particular emphasis on reviewing large programs such as
SAFE since these programs represent a very large commitment to the
future.
This section is organized to first summarize presentations and
discussion regarding major programs; followed by data base growth,
ending with conclusions.
RUFFING COMPUTER CENTER
The Working Group was briefed on status and plans for the Ruffing
Computer Center, through the 1987 period.
The plan for upgrading the Center is designed to be responsive to
growth rates of 20-25 percent per year in general computing. The plan
shows that by gradual replacement of installed computers with newer
and higher performance computers, the Center will be able to service
the growth rate and still fit in the same physical space. The plan
appears to be a.comfortable one with little risk.
It was stated that by 1984 there is the expectation of being able
to service 600-1000 simultaneous VM users.
The Center will continue to offer GIMS and RAMIS in the
foreseeable future for DMBS services, GIMS for the large production
applications and RAMIS for the smaller ad hoc applications. There are
plans to upgrade GIMS to run under VM, and within the CAMS II Project,
plans to improve GIMS performance with data base hardware. In
addition, ODP has an initiative to investigate backend data base
machines.
There was some concern expressed by Working Group members over the
plan: One person indicated that there are not enough people to
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86BOO689ROO0300040010-3
respond'to demands for building GIMS data bases. Another expressed
concern over GIMS as the standard DBMS offering, to the effect that
GIMS as an industry "orphan" would not be viable over the longer term.
The Working Group was briefed on CAMS II plans. As a community
sponsored system, it is of less concern in the Agency planning
process; however, a few comments are appropriate.
CAMS II will become operational in the 1984-86 timeframe. It is
to support improved collection systems, new users with new demands,
and is to further integrate collection and exploitation processes.
Perhaps the most significant aspect of CAMS are requirements which
are very similar to other Agency systems. For example: there are
requirements for interfacing electronically to other community
systems; for interactive graphics; for analyst support system
including electronic mail;. and for sophisticated DBMS facilities.
Later planning phases may consider how to best provide functions which
are generally common across several systems.
SAFE/ADSTAR
SAFE and ADSTAR plans were presented to the Working Group. These
systems are of particular interest since they represent new
initiatives in information systems which will be exploited and further
developed through the eighties. ADSTAR is scheduled to become
operational within the next few months an; the initial block of SAFE
in early 1983.
The current development contract for SAFE extends through 1984 at
which time the system is to have 600 terminals. It is planned that
the system be extended to support 1200 terminals in the 1986
timeframe, and up to 2100 terminals by the end of the decade. Other
possible initiatives for enhancing SAFE during the last half of the
decade include:
o Improved networking support: SAFE is to be coupled closely
with the RCC so that the analyst can conveniently use that
facility for analysis to augment SAFE capabilities. In
addition, requirements for interfacing to community
systems are anticipated.
o Enhancements to support translation from paper to electronic
form (OCR) for retrieval purposes.
o Automation of biographic files.
o Inclusion of wire services in dissemination process.
o Interface to ADSTAR.
A major initiative is planned for ADSTAR to support up to'150
remote softcopy/hardcopy terminals, to become operational in the 86-87
Approved For Release 2007/11/14: CIA-RDP86BOO689ROO0300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
ti.meframe. A new communications capability for the Headquarters
building and outbuildings in the Washington area for high speed image
distribution will be required to support the initiative.
The'response of the Working Group confirmed that the capabilities
to be provided by SAFE are well conceived. It is clear that many of
the information handling capabilities (compose, indexing, retrieval,
presentation) are universally applicable across the Agency. As noted
it other parts of this report, SAFE capabilities overlap with RCC and
will overlap with Office Automation systems; there is therefore a need
to fit these systems into an overall information system architecture
bringing to the end user a consistent set of tools to manage and
process information flowing across system boundaries.
DO INFORMATION SYSTEMS
The Working Group received an overview briefing on DO Information
Systems status and plans.
The universality of those functions to be provided by SAFE was
noted in the discussion: the major information systems in the Agency
have common requirements for information storage and retrieval, and
for the composition and coordination of textual information - the
challenge is to limit the numbers of times these tools are
implemented. Thq~ Study will answer some of the key questions 25X1
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
concerning how best to implement SAFE functions in the DO environment,
and how to best exploit the investment in SAFE.
Shortcomings seen in today's Information System operation and
practices that must be dealt with in the future were identified as:
o Challenge in coping with growing volumes of traffic
in electrical form.
o Lack of budget success in funding terminals
o System availability
o Reluctance of people to use computers
o Shortage of qualified people
o Lack of Agency level planning for ADP and
communications support.
DATA BASE GROWTH
The Working Group in responding to the task of estimating data
base growth over the next decade, for the most part, chose to
extrapolate from past experience rather than survey t:'eir
organizations. SAFE was the only area where there has been
considerable attention given to quantifying the size and
characteristics of data bases based on a survey of the user population
and analysis of the input-. more on this below.
Regarding online data storage capacity, the experience of the ODP
Computer Centers has been that capacity has increased approximately 20
percent per year over the last four years. Whether this growth is
fully responsive to the demand. or is impacted by some of the realities
of the supply side is not clear, nonetheless, the 20 percent growth
rate is accepted as a reasonable planning objective. With the gradual
replacement of-disk devices with new technology, which is currently
available, this growth rate can be sustained through 1987; beyond that
time either more space for disks or a technology advancement (a likely
possibility) will be required.
A representative from NFAC was able to provide estimates for data
bases in the SAFE environment. Summary estimates are: 0 private
analyst data bases for indexing, fast retrieval, and other support
data; general document access index data bases of 16.million records;
and a cable data base covering 10 years of traffic. In the aggregate
there may be as much as 100 billion bytes of information. NFAC also
has requirements for verb' large data bases residing on Ruffing
Computer Center facilities; for example, there are 20 gigabytes of
international trade data currently on magnetic tape.
The DDA estimated five to seven new large centralized data bases
over the next ten years; in addition, it is anticipated that there may
be a need for 35 to 40 smaller data bases on local processors.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
It is clear that the problems of managing and extracting useful
information from data bases overshadows the capacity issue. Data Base
Administration came up in the discussion repeatedly in terms of how to
manage our data bases so that information can be shared by larger
communities; so that duplication is avoided; so that the integrity of
the information is maintained; and so that records can be properly
maintained. Requirements to integrate administrative data bases,
concerning different aspects of personnel management, were described.
Needs for data fusion, analyst tools, geographic data bases, and data
base design tools were noted.
There were examples cited where organizations are becoming more
aware of data base administration requirements and responding by
setting up special organizations to fulfill the need; the scope of
these organizations tends to be narrow however, usually related to a
specific system. There is a need for database administration
functions which are more global in scope, for the purpose of managing
one of the Agency's more valuable resources -- information.
CONCLUSIONS
A good deal of time was spent reviewing major system programs,
since they represent a significant commitment to the future. The
reviews were informative and members of the Working Group seemed to be
in general accord with the direction b-ing taken. Perhaps the most
important point which showed in the review was that the programs are
working many of the same requirements with different solutions:
examples included analyst support systems, data base support, and
office autc.nation; this is an issue to be considered in later phases
of the planning effort.
The following long range planning goals are proposed:
Goal 1) Provide Data Base Management Systems (DBMS's) which are
at the forefront of technology and which have the
highest potential for remaining at the forefront over
the long term.
The creation, maintenance, and processing of data
bases are among the most important of our information
system capabilities. The DBMS is the cornerstone of
that capability, providing the primitives and interfaces
upon which many applications are dependent. Query
languages, data dictionaries, program generators,
report generators, and other tools evolve from and
are tied to the DBMS. Therefore, it is essential
that the Agency be at the forefront of this technology.
Goal 2) Provide administrative procedures and information
system facilities to better manage one of the Agency's
most valuable resources--information.
In years to come the Agency's information resources
will expand rapidly, more of the information will be
in electronic form, there will be a need to process
the same information in various systems, and information
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
will be shared and utilized more widely. Administrative
procedures and system facilities must be designed
for this environment. Much of the information is
a corporate resource--it must be managed as such.
Dictionaries or catalogs describing information
resources must be automated so the end user is
informed on what information exists and how to
access it. Standards must be promulgated for the
interchange of information--so that files, messages,
and documents can flow between systems.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
Data Transmission, Dissemination and Distribution
INTRODUCTION
During the eighties the Agency will be challenged with providing
the necessary communications and dissemination facilities to carry and
disseminate increasing volumes of information to the right people in a
timely manner. Trends, which will have an impact, include:
o New overseas facilities and procedures, such as CRAFT
o Dispersion of more terminals in the Washington area
and the need for connectivity (terminal to terminal,
terminal to multiple computer systems)
o New Office Automation functions such as electronic mail,
teleconferencing, and shared electronic files place demands
on local communications capabilities
The Working Group was challenged to:
o Characterize and provide estimates for new operating
procedures and new services which drive requirements
for distribution and dissemination facilities.
o Express strategic, or qualitative arguments which should
be factored into the planning process.
o Review current and planned programs to determine deficiencies
in satisfying Agency needs through the eighties.
This section is organized to first address network facilities,
followed by dissemination facilities and a discussion of services,
ending with conclusions.
NETWORK FACILITIES
The Working Group was briefed on the MERCURY Program to provide an
awareness of current plans to improve the Staff Communications
Network. Regarding other aspects of those plans: the SKYLINK Program
was not discussed; there was a small amount of discussion regarding
the Washington area network.
The MERCURY Program is designed to incrementally modernize the
message and data communications computer systems through the eighties.
Requirements are premised on growing demand for narrative message and
data communication services, and increased use of ADP Systems and
office automation. Information was provided to the Working Group on
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
network size and traffic handling design objectives. as an attempt to
provide insight into the responsiveness of MERCURY to the demands of
the eighties: whereas the growth potential of MERCURY, both in the
number of connections and in the amount of traffic, seem impressive,
it was beyond the scope of the Working Group to provide other than a
superficial assessment of the program and its ultimate impact on the
IHS network.
i
It was pointed out that MERCURY does not support voice. It was,
noted that the DO does have a need for secure voice--the priority of
such a need was not discussed. It was mentioned that the STU-II
Program may be responsive to voice requirements.
It was noted that MERCURY will limit any single virtual channel to
I.
a maximum rate of 64Kb--this was not taken as a cause for concern.
MERCURY will follow the ISO Open System Connection Model in
iImplementing the communication and inter-computer protocols; the
transport level of protocol will be designed to the NBS Standard.
DISSEMINATION FACILITIES
The Working Group was briefed on the Cable Dissemination System
(CDS), and on plans for the CDS replacement system (MHF). In
addition, near term activities to alleviate CDS shortcomings were
described.
The discussion centered on the fact that there are several
systems, either in operation or planned, which are to service some
part of the Agency's cable dissemination needs, and which are not well
coordinated. Included are CDS/MHF, SAFE, and COMET. Present planning
requires these systems to coexist in a dissemination network, where
there are dependencies between them. System design requirements such
as sizing, availability, data integrity, and operational control, are
examples of dependent requirements which must be understood in the
context of the network for the most cost effective and operationally
satisfactory solutions. This line of discussion was ended with the
understanding that the issues raised would be dealt with in later
planning activities which are to be architecture oriented.
It was pointed out that increased reliance on automated
dissemination brings with it some issues which will require attention.
Included are:
o Dissemination of non-cable traffic - there will be
increasing emphasis on data traffic (e.g., files).
o Standardization of formats - lack of standards impose
a heavy burden on automated systems.
o Security - replacement of human review processes; electronic
analogs to paper privacy channels; dissemination levels.
o Electronic origin - as more electronic traffic is
originated in Headquarters Facilities, with outside
addressees, different control mechanisms may be required.
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
o System Availability and Integrity - shift to electronic
domain from paper places added importance to these
parameters.
SERVICES
In responding to the challenge to identify services which will be
of some importance during the eighties, and which will have an impact
On data transmission, dissemination, and distribution facilities, the
following comments are made:
o MERCURY appears to provide the framework for satisfying
the more important services to exist between Headquarters
and the field. The actual transmission facilities were
not considered.
o There are needs for voice and teleconferencing services
between Headquarters and the field which are not currently
planned.
o Local networks for linking Office Systems and Data
Processing Systems will be of increasing importance.
CONCLUSIONS
The Working Group was able to discuss and evaluate dissemination
facilities in some depth. It was less successful with distribution
and transmission facilities, perhaps because these facilities are more
of a second order user consideration: other system requirements more
directly responsive to the users needs were more easily dealt with.
The following planning goals and objectives are highlighted as
results of the Working Group proceedings:
o Goal 1 - Define system architecture for information dissemination
network.
As discussed above, there are several systems, engaged in
dissemination which are to be integrated into a network of
processes which need to be supportive of and consistent
with one another.
o Goal 2 - Define needed standards for network information
exchange.
The proliferation of cable formats and its impact on the cost
of information handling systems exemplifies the need for
standards. As new modes of information exchange are introduced,
the need for standards will be even greater.
o Goal 3 - Define requirements for local area networking needs
and develop a plan for a uniform system of local networks.
As indicated throughout this report, there will be increasing
needs to link office systems and data processing systems in
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
local areas. Even today the solutions to this problem are
starting to proliferate, and because of the lack of.uniformity,
will eventually cause others as networks become more expansive
and interwoven.
o Goal 4 - Develop tools and methods for specifying performance
objectives and validating the design of information system
networks, against those objectives.
In a complex network such as the Agency's Information Handling
Systems it is extremely difficult to achieve an adequate level of
assurance that performance objectives of the individual systems,
and the network as a whole, will be met. It is even difficult to
specify the objectives. But, we must begin to apply methods which
will give more assurance that systems will support the
operational needs when deployed. Information services must be
characterized as to the performance requirements they will impose
on underlying facilities. Facilities must be.analyzed and modeled
.to validate that they will accommodate the requirements.
SECRET
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
 Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3
Iq
Next 4 Page(s) In Document Denied
Approved For Release 2007/11/14: CIA-RDP86B00689R000300040010-3