The Gilbane Report: Volume 9, Number 3Privilege Management & Rights Management for Corporate Portals
April 2001
Download a PDF version of this article Read the news for this issue.
Privilege Management & Rights Management for Corporate Portals
It may not seem so when
you are busy implementing, testing, and collecting the initial reactions from
your users, but there is no doubt that it is significantly easier to build applications
to deliver customized content from multiple content repositories and data sources
than ever before. Whether you buy a 60% solution from an EIP (Enterprise Information
Portal) vendor, extend an existing content management or database application,
or build your own solution from the ground up, the application and information
integration technologies now available allow you to deliver a wide variety of
information to a broad audience. Unfortunately, this also makes the challenge
of information security more acute than ever before.
Whether the information
distributed is owned by you or syndicated, you need to make sure that employees,
partners, customers, and the curious have access to what they should, and can
reuse content in appropriate ways. Traditional tools and strategies need to
be extended and enhanced. This month Bill and David are joined by consultant
Larry Gussin to provide an extremely useful overview of the issues and approaches
that will help get you jump-started on a strategy.
Privilege
Management & Rights Management for Corporate Portals
With the quickly growing
demand for intranet-based enterprise information systems, as well as for extranet
extensions, the enterprise information portal (EIP) is becoming the primary
emerging solution to the problem of intelligent user access.
Enterprise information portals
extend Web content management (CM) solutions by delivering both enterprise and
commercial content and core enterprise and industry information through a single,
unified, and usually browser-based interface. An EIP may present Web sites,
documents, databases, email, and other information types from multiple servers,
and allow users to access this information through its portal server. The key
EIP goal is to provide more efficient access to business-critical information
for employees, customers, suppliers, and business partners.
With content management
and portal technologies emerging as a new, robust framework for enterprise and
extranet information, the traditional enterprise security solutions, which are
predicated on online network sessions and on providing document level access,
may no longer be adequate or efficiently manageable. IT managers should wonder,
for example, how these firewall-based solutions will be able support the potentially
huge emerging requirements for extranet, offline, and more granular access to
information.
Equally important is the
question of how information access security can be managed. If the rise of EIPs
reflects the need to address the growing number of information resources found
within enterprises, these information resources still require security decisions
from their business line managers. With the numbers and types of users of these
information resources also growing in number, as well as being potentially tied
to multiple locations and access relationships, the information access management
challenges become even more daunting.
With all this complexity,
enterprises must address important infrastructure requirements before they can
enjoy the benefits of extending enterprise information internally among their
business units and departments, and externally among their business participants.
Two of these requirements address questions of how enterprise managers can ensure
that:
- Users effectively access
the information they need.
- Business rules govern
how and by whom information is used.
Two distinct solution categories
exist that can address some part of the extended enterprise's need for information
and content security control: privilege management and digital rights management.
The solutions available today are still caught up in their cultures or origin,
but the real-world needs of enterprises may be answered by the right combination
of these solutions. Such a combination of approaches would effectively manage
both online and offline access to content, and provide a persistent protection
and control of information throughout its lifecycle.
Enterprise Information:
Content & Data
Viewed comprehensively,
enterprise information includes "content," which draws upon a publishing
model and includes such things as documentation, marketing and sales collateral,
and research reports, and "data" from applications involving e-commerce
and business processes. The types of information an enterprise must manage usually
include unstructured documents such as Word and PDF files, email, and syndicated
content feeds, as well as a complex array of media types, such as animation,
audio, and video. Structured content-in the forms, typically, of SGML or XML
documents and relational databases-also can form a large part of an enterprise's
information pool, whether as parts catalogs, indexes, or database reports.
An enterprise's portal-accessed
information also includes application data. APIs and middleware let enterprises
manage a wide range of information sources such as enterprise and third-party
data sources and applications, including legacy applications and ERP systems.
Meta-data, or data about information, also plays growing and crucial roles in
enterprise portals, and increasingly, XML is forming the basis for meta-data
applications.
Many EIP solutions build
on CM tools, or integrate with them, because much of the information users require
is at the document or web content level, and because enterprises increasingly
must manage more electronic content, from an expanding variety of sources and
for different types of users. What constitutes "content" in content
management systems is changing, largely driven by the benefits of e-commerce
that draw enterprises to work beyond their own intranets to exchange content
and data with customers, partners, and suppliers. Because of e-commerce requirements,
content is increasingly being tied into other applications.
Enterprise Information
Portals at the Application Level
But content is only one
part of many EIP solutions. The concept of the integrated portal to crucial
company information can include access to core enterprise applications themselves.
An integrated portal interface can provide applications access, largely through
Application Server development environments that support such tools as Enterprise
Java Beans, Java 2 Enterprise Edition (J2EE), CORBA, and COM. At the application
level, EIPs are being designed to solve information problems in areas as diverse
as commerce, support, partner relations, workflow, collaboration, human resources,
supply chains, business intelligence, data mining, ERP, and application or system
integration.
The Web has ushered in a
model of multi-tier applications, for supporting many of the applications listed
above, and for managing such things as product data, inventory control, and
customer relationships. Yet these new applications are hard to architect, as
they often involve legacy systems, varying data sources, and complex communication
needs. Environments such as J2EE mitigate these challenges by providing comprehensive
services for deploying different applications as component or modular developments.
But as much as modularizing
applications can help in integrating those applications within a portal, application
server technology is not the clear answer for every enterprise. Some enterprises
are more "content-centric" and so are better served with the focus
on content management technologies within their EIP. Other enterprises may not
be prepared for the complexity and investment required to re-engineer a broad
scope of business process-related applications at the component development
level.
Enterprise Information
Portals at the Access Level
For all the wide-ranging
attributes that can describe an EIP-from the simple single access point for
a company's content, to application-integrating database platforms-and despite
the diverse nature of the target users, there are three basic information access
business requirements that are widely met by all the solutions:
- Access requires personalization.
Personalization, when coupled with effective search and retrieval technologies
and other information categorization techniques, helps to ensure that the
right information is presented to the people who need it.
- Access requires security.
Making information accessible is a double-edged sword, in that information
accessible to the wrong users can do an enterprise great harm. This issue
of information security can also involve compliance, data integrity, and authorization
for actions taken on information-whether in terms of financial transactions,
strategic planning, or other forms of commercial collaboration between enterprises.
- Access requires network
architecture flexibility. The EIP presents many types of information to many
types of users in many types of use conditions. Such complexities raise issues
in terms of integrating disparate information into one user interface, but
they also raise issues in terms of the network environment through which information
is presented. Information should be accessible across client-server and peer-to-peer
architectures, so that both online and offline access to and use of valuable
enterprise content is available as users' needs require.
Privilege Management &
Rights Management: Two Models of Information Access Security
Two distinct but possibly
compatible solutions are emerging to manage information access security in the
extended enterprise. Each builds on a set of core security technologies, including
but not limited to key management, encryption, meta-data, and hardware enforcement.
One approach builds on client-server architecture and the other on peer-to-peer,
as follows:
Privileges management,
which enforces centralized information security by maintaining valuable information-meta-data,
transactional or application data, or other content-on secure central servers.
Users, based on their hierarchically defined roles, gain privileges to view
such information, but not to download from the server except with central
authorization. Trust is enforced through control of access to this central
information service, as well as through client-side user authentication. E-mail
communications and attachments, as well as user-created information, are not
typically protected within this kind of trust environment.
Rights management,
which distributes valuable information and provides tools for locally enforcing
the secure, role-based management of the information. Based on their hierarchically
defined roles, users gain privileges to store, view, edit, and share the information.
Internal and external owners of information possess and can enforce ownership
rights governing the information. Users may become authors by locally creating,
enforcing security for, and sharing new valuable information. Each network
peer is empowered, depending on roles and circumstances, to act as a server,
a client, or both. Trust is enforced at each peer, which works either online
or offline in conjunction with centralized authorization management services.
E-mail communications and attachments, as well as other types of user-created
information, are protected within the trust environment.
Both of these solutions
raise questions of security, scalability, efficiency, and adequacy in addressing
the information access needs of the extended enterprise. To some degree, however,
these solutions grow from different traditions and solve different problems.
Privileges management grows from an IT tradition of tightly governing information
access and collaboration within client-server environments. Rights management
grows from Internet and electronic publishing traditions that view information
as interactive and disseminated, information users as potential creators, and
communications as open beyond the client-server environment.
To a large degree, rights
management (RM) systems functionally overlap with privilege management (PM)
systems. Rights management systems could integrate with PM systems components
to maintain the primary PM user advantages of easy sign on and personalized
views of information. The ways RM systems extend or exceed PM systems in terms
of user advantages is defined by the differences between privileges and rights.
In privilege management,
once information is defined as valuable, all rights relating to it are executed
through secure central servers. Users gain access and viewing privileges, but
cannot store information locally or directly share it with other authorized
users. In this client-server model, information usage and sharing are restricted
in the service of greater control.
In rights management, central
control is still maintained through authentication and authorization, but the
execution of access and use is extended to a user's local device by means of
container-level and peer-level security. The distribution of management services
to peers means that rights can reside on user devices and that a user can be
hierarchically authorized to create, edit, distribute, and share in the managing
of information. Conversely, a party with rights bound to information-an employee,
a partner, a customer, a publisher-can, with hierarchical authorization, possess
business controls and opportunities in relation to that information that the
rights holder can extend directly to the individual user. The central control
of authentication and authorization found in both PM and RM systems means that
RM systems also can provide the network access and viewing privileges granted
by PM systems.
Privileges Management Information
Access Security Model
Privileges management systems
grow out of an IT tradition of centralized network security, yet they have evolved
as Internet communications are making application-specific enterprise information
and security models inadequate. Privilege management systems, which build upon
a number of core security tools culminating in policy servers, support authentication
and authorization access models across the extended enterprise. This centralization
of privilege policy management creates efficiency and scalability through several
means, including, most prominently, the decoupling of information applications
from security management. With little or no presence on the local client, however,
PM systems typically cannot extend centralized management to data and communications
on a loosely coupled client device.
Privileges management systems
began emerging in 1998, and Bear Sterns projects this to be a $1.4 billion market
in 2002, growing to $2.4 billion in 2004. While competitors include Baltimore
Technologies (www.baltimore.com), Entrust
(www.entrust.com), OpenNetwork Technologies
(www.opennetwork.com), and Oblix (www.oblix.com),
the Meta Group calls Netegrity (www.netgrity.com)
the first mover company in PM systems and says that it currently holds a 75%
market share.
The PM Security Solution:
Trust Bound to the Server
Privilege management systems
create a trusted environment by binding valuable information to IT-managed servers,
and by giving enterprise or extranet users role-based access to the information,
but no ability to store the information locally or send it to others.
In essence, PM vendors claim
that their systems provide an evolutionary step in the development of IT authorization
mechanisms, improving upon the decentralization of many applications imposing
their own security by providing centralized access management. By replacing
operating system-based authentication and authorization services, which are
built on decentralized access control lists (ACLs) that govern single applications
and manage resources on a single platform, PM systems can reduce significant
administrative resources.
The more efficient authorization
system, PM vendors claim, is based on providing centralized governance over
who gets access to what resources across multiple platforms. Using directories
and policy servers, the centralized authorization system lets administrators
set authorization policies based on user roles. There policies can govern access
privileges to multiple applications across multiple application servers. Authorization
governance can also be extended to servers that reside across the Internet.
Privilege management systems
use a number of security and information management technologies, most of which
are themselves in a state of evolution. These technologies include encryption,
keys, key management, certificates, and hardware security. Sitting above these
technologies, the basic components of a PM system include:
Directory servers.
Enterprises use collections of directories to hold data descriptive of users,
services, devices, applications, and relationships. These directories may
be embedded in email servers, firewalls, PBXs, applications servers, and file
servers. Directory services provide a mechanism for naming, describing, and
finding enterprise or external resources in these directories. Lightweight
Directory Access Protocol (LDAP) is a protocol that is emerging as the standard
way a directory service accesses a directory listing. It is the mechanism
that policy servers use to identify authorization and privileges policies
and the users and applications to which these policies apply.
Policy servers.
Policy servers store role-based authorization data that provide access to
network resources. Policy servers can be used to serve authorizations stating
who, based on assigned user class in relation to a specific request, can access
what information, and in what way. Role-based policy management systems can
support personalized user interface, organizational hierarchies, multiple
roles for individual users, and organizational ownership or roles.
Meta-data. Meta-data
is used to enable machine-readable processing across heterogeneous systems.
XML-based markup languages are becoming the accepted mechanism for creating
meta-data standards. To scale their systems, PM systems will need XML-based
meta-data standards in a number of domains, and most PM vendors are very involved
in standards work. The meta-data standards efforts most directly important
to the success of privileges management include OSSML (securing authentication
and authorization data transmission), XACML (an access rules language), XKMS
(for key management), and XML Encryption.
Virtual Private Networks.
A virtual private network (VPN) is an encryption configuration that enables
the secure exchange of information across the Internet in the apparent form
of a WAN. It allows remote users to connect to enterprise servers, and also
supports exchanges of information between an enterprise and its customers
and partners. PM systems can use VPNs to provide access privileges across
the firewall.
While ensuring information
security, PM systems also provide administrative efficiencies and scalability,
by removing security management from applications and embedding it in a centralized,
shared services infrastructure. PM systems build on existing and emerging IT
infrastructures, allow for the delegation of management tasks across administrative
networks, and support auditing and accounting.
The User Experience: Role-Based
Server Access Privileges
The user environment within
a privileges management system is one of easy authentication and easy, personalized
access privileges, but no real support for offline security, authoring, or collaboration.
Characteristics of the user experience within PM systems include:
Single Sign On authentication.
PM systems allow user authentication techniques that go beyond password protection
for single applications and other weak and inefficient enforcement methods.
Using their centralized infrastructure, PM systems enable Single Sign On,
in which users log on once to obtain authentication and authorization to their
role-based privileges across application servers and VPNs. Single Sign On
improves user capabilities, reduces administrative load, and supports the
use of scaleable, weak-to-strong underlying authentication security mechanisms,
which are tied to the value of the information that must be secured.
Personalized views.
PM systems support personalized portals. A user may have a single view that
opens onto all of the valuable information that the user is authorized to
see, possibly distributed across multiple enterprise applications servers
and VPNs.
No store or forward
privileges. With valuable information centrally protected on servers and
no way to protect information stored on the local device-client-side trust
only governs authentication-users are provided minimal interactive access
to valuable information. They may not store it locally or forward it to other
users. This may be a significant liability to productivity in the increasingly
collaborative environments of the cross-departmental enterprise and the extended
enterprise.
Distributed Rights Management
Information Security Model
Rights management (RM) systems
grow from two impulses: one is to support the Internet and electronic publishing
tradition of interactive, open access to information; the other is to protect
digital forms of information that can allow easy editing, copying, and disseminating.
Rights management research began in the 1980s and the first systems appeared
in 1997. The 1999 advent of Napster introduced publishers, creators and users
to consumer-markets RM issues, but RM has also become an enterprise topic, especially
following on peer-to-peer networking.
The earliest RM systems,
and still the apparent market leaders, are Intertrust (www.intertrust.com),
ContentGuard (www.contentguard.com),
and Microsoft (www.microsoft.com) which
through an investment has access to ContentGuard's patents, though many new
competitors have appeared. Authentica (www.authentica.com)
and RightsMarket (www.rightsmarket.com),
for example, offer enterprise solutions that provide limited client-side extensions
to PM systems, while TrustData (www.aspsecure.com),
an Intertrust-based solution, builds ASP-like RM solutions. While RM market
sizing is quite preliminary and largely based on consumer sectors, IDC projects
an overall $2.02 billion business in 2004.
The Security Solution:
Trust Distributed to the Peer
Rights management systems
largely build on the same security tools and meta-data tools as do privileges
management systems, and in enterprise deployments they may also use some of
the same policy server and directory-based authentication and authorization
techniques as do PM systems. Yet RM offers a very different control environment
from PM: it lets authorized users view, edit, create, and share content, and
achieves this capability by persistently protecting any content (or content
meta-data) that resides on peer devices or moves over a network. Rights management
vendors say that their distributed protection systems provide the security,
efficiency, and scalability support that extended enterprises require, including
peer-to-peer and online or offline environments, as follows:
Information delivery
is by secure communications. Different from PM systems, which provide
secure server access within the firewall and use VPNs to extend server access
beyond the firewall, RM systems use cryptographically secured containers to
protect information as it travels over a network. A container can hold, in
any combination, single or multiple content elements, as well as meta-data,
such as audit data, or role-based rules, possibly derived from policy servers,
that govern how the content can be used. Users can share information via the
same protocols.
Information can be
persistently protected locally. RM systems persistently protect information
on the local device, extending management control to that device, by providing
a secure local environment in which encrypted containers are processed and
information is used. The secure environment contains or interoperates with
tools for unpacking or packing containers, evaluating usage rules, and viewing
or editing content. It supports local and even offline management by means
of a database that contains information such as role-based viewing and authoring
privileges, audit data, and budget allowances. The database is modified through
the transmission of update information in a new secure container.
A distributed services
architecture extends to the peer. A network in which every peer device-from
PCs to phones to media production hardware-and every user application is secure
creates an architecture that supports a distributed services environment.
Rights management vendors claim that enterprises or contracted providers can
use this network to provide scalable, efficient online or offline services
for clearing and reporting financial, usage, rights, licensing, and compliance
information, as well as for other operations. Vendors also claim that, where
authorized by an enterprise, publishers and other commercial agents can use
the scalable and efficient distributed trust architecture to market and provide
products and services directly to individual users or collaborative groups.
Recent Developments with
PM & RM Vendors Suggest Some Initial Steps Toward Integration
As enterprises move beyond
the islands of information that have separated discrete data and content applications
and move to make business critical information available to the wide range of
users both in and outside the enterprise, the benefits of integration produce
significant challenges for the management and security of the information. Fortunately,
vendors will increasingly be working towards integrated solutions.
For example, Adobe (www.adobe.com)
has embedded InterTrust's DRM technology into just-released Acrobat 5.0, and
InterTrust and Artesia (www.artesia.com)
have announced a deal to integrate InterTrust's technology into Artesia's Digital
Asset Management product for media, publishing, entertainment as well as for
corporate enterprise applications.
In truth, information access
and security vendors are just starting to think through many of the enterprise-oriented
issues, as the complexity of the information sets presented through EIPs create
the market. Recently, for example, Netegrity (www.netegrity.com)
announced a bold though sketchy partnership program comprised, somewhat hopefully,
of over forty vendors who work collectively in the portal solution, application,
CRM, personalization, user provisioning, web services infrastructure, content
management, authentication, and core security spaces. Among these partners is
Authentica (www.authentica.com), one
of the few digital rights management-based enterprise content security companies
with product shipping.
There are a few other signs
of the convergence of PM and RM. Earlier this year, RightsMarket joined the
Entrust Alliance Developer Program, with the aim of making the company's RightsVault
work with the security infrastructure provided by Entrust's trust solutions.
The goal is to extend their security to file-level use, including text, audio,
email and HTML formats.
Closely watched P2P software
provider Groove Networks Inc. (www.groovenetworks.com),
has just announced the availability of Groove 1.0, which includes support for
a wide range of firewalls, reduced resource requirements, support for secure
roles and permissions, Microsoft Office and NetMeeting integration, and Enterprise
Network Services. Groove has partnered with Digital Goods (www.digitalgoods.com),
an early DRM vendor (as SoftLock) that now focuses on using DRM as part of its
digital content marketing services. The initial focus of the Digital Goods-Groove
application is subscription-based sales of digital content for the business
marketplace, such as financial and investment research, corporate intelligence,
and other time-sensitive products, as well as business training and education
materials. But Digital Goods sees other enterprise-oriented applications becoming
enabled through the Groove partnership, including protected one-to-one or group
sharing of sensitive content with real-time collaboration capabilities. With
the relatively easy integration of currently separate applications' information
access through LDAP management, this early-stage market and concept confusion
may soon begin to yield to reliable and useful control over the increasingly
complex information enterprises must share with wider and wider audiences. The
advent of enterprise-oriented peer-to-peer network services suggests the mixing
of client-server and peer-to-peer content access and use strategies. Once the
tools are really available business managers will learn if their increasingly
complex efforts to control access to and share content get a little simpler-and
more appropriate for different network situations.
David R.
Guenette, Larry Gussin1, and Bill Trippe
Information Access
& Security Requirements
IT managers who are
building information access and security requirements for an EIP can use
the following, admittedly daunting sets of questions to guide their thinking:
Content value and
types. Is all content high-value or sensitive? Are there fiduciary,
regulatory, or compliance requirements that must be supported in regard
to the content? What are the types and amounts of information to be presented?
How varied are the sources of the information, among applications, servers,
and internally and externally to the enterprise? How are inter-enterprise
legacy systems integrated and presented? What is the role for emerging,
non-PC information platforms? Are rich media to be supported? How does
email, instant messaging, and other "ephemeral" communications
fit? What security and/or auditing is beneficial?
Content actions.
What level of content management and/or groupware collaboration needs
support? What is the role of workflow, versioning, updating, authority
in managing the creation and maintenance of the enterprise's content,
including intra-document or compound document management, and do these
actions need or benefit from being performed offline? What security issues
arise within dynamic page presentations? Can URL-level security resolve
most access challenges? What demands does contextual content delivery
make on information access security management? Will syndication and/or
aggregation of content into or by the EIP need to be supported? Can information
search and retrieval take place offline? Are the categories and taxonomies
that support retrieval, personalization, and contextual presentation of
information available offline?
Persistent protection
and securely associating business rules with content. What types and
in which situations might content demand offline access? Are mechanisms
for content value chain participation without participation in an enterprise's
client-server network useful, important, or valuable to some business
situations? Are super-distribution and controlled content dissemination
important and valuable tools in some business situations?
Support for mixing
of PM and RM models. Can privilege management and rights management
models co-exist within the same EIP? Are there efficiencies, expanded
business models, or other benefits to mixed models for information access
management in the extended enterprise?
Support for mixing
client-server and peer-to-peer models. How well does client-server
network architecture meet the requirements of the extended enterprise?
What are the advantages and disadvantages of peer-to-peer security models?
Can a peer-to-peer network-based model of rights management be applied
to authentication-based trust within a client-server network architecture?
What role might rights, payments, and/or usage clearinghouses play?
Centralized information
access management tied to directories, and especially LDAP. Does the
policy management or rights management system address LDAP, or provide
or support "directory smart APIs"? Depending on the type of
EIP and the type on enterprise platforms, other directory forms, such
as NDS or AD may need to be considered. Personalization engines will also
need to address LDAP, or links from LDAP to user profiles.
Appropriate privilege
or rights assignment authority. Who sets information access permissions
and/or rights policies? How flexible and granular is the user class or
role definitions? How are policies determined across various business
departments and areas of business responsibilities? Who defines users
classes and roles, granularity of information for access and privileges,
and who implements the management of these privileges and/or rights? Does
the solution permit ad hoc authentication or authorization by managers?
What is the PM and/or RM systems' capacity to address privacy, especially
in industries where regulation of privacy is becoming law?
Auditing requirements
must be supported. What levels of audit/usage tracking is required
now, and how might that grow a year or two down the road?
Other, required
technologies and services must be identified. Determine the integration
required with other networks or platforms within the enterprise. Is an
ASP a viable way to build an EIP capability? What is required to use extranet
technology, such as VPN? Can the EIP securely interact with an e-marketplace
exchange or vertical industry portal to support extranet e-commerce functions?
What is the management overhead, services, and technologies available
for PKI, digital certificate, or other required security foundations?
Interoperability
and standards are crucial areas to watch. What are the interoperability
issues of content management and application data? Does the EIP solution
work with XML-based data, content, security, and interchange standards?
What level of application server technology is needed to integrate legacy
information systems?
|
1 Larry Gussin (ldgussin@home.com) is
a consultant specializing in rights management B2B and B2C infrastructures.
|