Enabling
Enterprise
Security Management
For every
codified thought there is a purpose, a requirement that meets a need.
The structuring and aggregation of logic and data is the essence of
small programs that as building blocks can become large
applications. Regardless of the size there still remains a
purpose with a requirement that fills a need sustaining its life
expectancy.
For example, in
years past the long distance telecommunications companies bread and
butter application was their billing applications that serviced the
call data record (CDR)s based on Bellcore Labs standards. With
the advent of more sophisticated switches, embedded applications for
processing CDRs was fulfilling a newer advanced
requirement. While not reaching it's absolute end of life, it
evolved or was reinvented so that in the reincarnation it could support
more advanced hardware capabilities or even replace hardware as
something known as a soft switch.
People speak
about killer applications. As far as the telecom firm is concerned the
billing application is it. To the rest of the world an application like
Lotus 1-2-3 was the killer application of our times with the coming of
the IBM Personal Computer because millions of licenses were sold. It's
utility verses that of absolute automation of the process is what makes
it a publicly accepted killer application. And, the world waits
for the birth of the next killer application. The question is,
will it ever come to pass?
It seems
disingenious to believe that disseparate software work products could
be put together in such a way as to build a killer application.
After all each, and I am referring to open source software, was
constructed from totally different requirements without a shred of
thought documented with respect to cross purposes, dependencies or
supportive requirements. However, that is exactly what has
happened with the open source movement. For example, a
unparalleled development of the web server and the application
server. What does seem common that often a derivative work is
born from a work product such as the need for the servlet container
whose purpose was not originally inspired as a requirement in Apache
Server or any of the early web servers for that matter.
The open source
inventory of work products has reached a critical mass in terms of
variety, quality, distribution, support and all attributes of purpose,
usefulness of utility driven by requirements within the framework of
open standards. And yet it seems as though the IT industry is
still grasping for meaning and understanding as to where open source
fits in the scheme of things. Even firms that publish proprietary
applications cannot figure it out.
The next killer
application1 is going to the defining moment for the open source
movement. The next killer application will be a magnitude greater
than the standalone office application and surpass modern day
enterprise applications. In all of the IT industry we hear about
the standard mega-applications such as Enterprise Resource Planning
(ERP), Customer Relationship Management (CRM), Supply Chain Management
(SCM) and Sales Force Automation (SFA) solutions that address
functional business areas build on proprietary platforms. Every
computer software company web site one visits whether it is IBM,
Microsoft, Hewlett-Packard, Computer Associates or Oracle etc., there
is one major functional area missing. Rather, we believe it is
missing because the past nature of it's existence. The function was not
a crucial management function within corporate America or within the
Business Schools and Universities curriculum's. Safety and Security is
no where found on anyone's management radar screen. However, this
situation is about to change. With a convergence of IT security
and traditional corporate security, with a pressing need of traditional
corporate security since 9/11, the need for corporate governance and
compliance, business entities have literally invented a new corporate
management function. This has been witness by creation of new job
titles and roles such as Chief Security Officer or Chief Information
Officer. Unfortunately, there has not been any dedicated
applications equivalent to those such as CRM or ERP to support this new
convergence. The traditional corporate security department was
run by security directors/officers/managers that have been typically
concerned with access control, guard services or monitoring and
surveillance. The security department has been viewed as a cost
center primarily because there have never been any long term studies
that show return on investment on these expenditures.
Quantification is difficult at best because safety and security has
never been commoditized into its various aspects that could earn it a
reputation front and center in the boardroom. From mitigation of
risk and liability, workplace violence prevention, zero tolerance for
shrinkage, to intellectual property protection, hard data is simply not
available that is a result of direct management over all aspects of
safety and security.
The next killer
application is the Securitydirector, LLC Enterprise Security Portal
(ESP). It may become recognized at some point in the future as a
killer because it is differentiated by several key attributes:
- It is not
constructed using proprietary platforms and resources.
- It is built entirely
with open source software.
- It addresses and
supports this new crucial management function in the enterprise called
Safety and Security Management (SSM) or better titled Enterprise
Security Management (ESM).
Securitydirector,
LLC has been carefully crafting the Securitydirector Enterprise
Security Portal (ESP) on a business model premise of offering Fortune
500 professional level security management services to small and medium
size firms that would not be able to afford the same quality and level
of services from a dedicated department as found in the larger
corporations. The ESP is a management tool that facilitates
messaging and reporting for a precise level of knowledge diffusion
throughout the organization.
In this paper, I
am proposing to write a text that is both technical in documenting the
construction of a major portal application using open source software
and its practical application driven by a new framework of requirements
that encompasses a convergent industry evolving into an entirely new
business segment that will have in the future corporate board level
visibility in managing better profitability through managed safety and
security.
I believe this
proposal is supported by a number of factors. Given the fact that
the ESP's evolution was a slow process between the concept owner and
the IT architect, the text will make direct correlations between each
of the open source packages used and the requirements it
fulfills. It's magnitude is greater than a standalone office
application yet the number of users will surpass even site licensing of
previous killer application. And, the ESP has the potential of bridging
the gap in safety and security between the public and private sectors.
But, what makes
the ESP a defining moment for the open source software movement?
To answer this question I think there are 2 key economic factors that
will set open source apart from the old traditional ERPs and CRMs of
the world. The defining moment has everything to do with
economics. Building the ESP required only the services of a
highly skilled IT architect. Almost no outsourcing to offshore
development companies and only 10% programming labor was required to
build a product in record time. Using all open source code and
the skilled integration techniques of an architect, the chief
difference between branded portal server products and the ESP is the
later had a lower cost of labor and shortened time to delivery.
Development costs are considerably lower because a company does not
need to train programmers to support and write code to a specific brand
of portal server products and development environments.
Utilization of more generic skills is more cost effective. Base line
systems administration and support may be the only unchanged factor.
This defining
mega-application will relieve the fear of open source utilization
because the most fearful constraint is the unknown question of package
integration. What will be the results? For example, when
posing the question to news groups or forums as to whether ZOPE would
work on top of Jetspeed, and oh yes, it has to be operational within a
totally secure session, no one really could give an answer because the
2 packages were never integrated before in such a way. The killer
application will show the ingenious ability of putting together
entirely disseparate open source software products in spite of their
different requirements and unparalleled development. The major
labor of the architect involves the research and planning that must be
considered when proposing the integration of open source
packages. The object of course is to not reinvent the wheel and
yet still meet the requirements developed from concept owners.
Demonstrating the effectiveness of building products from the open
source inventory takes the risk out of a possible failure of
integration. In the final analysis, the results will support the
premise of the defining moment.
Another aspect
and demonstration of flexibility in support of the defining moment has
to do with scalability both at the application server level and the
portals ability to service requirements from distinct vertical markets
or industries. The requirements once again drives the technical
side because the ESP must support many companies simultaneously in
their own Intranet like vacuum. In other words, portal utilization and
customization appears to be each client's own unique instance even
though 100 other firms are using it. Profiling and roles along with a
finer granularity of security within system management of such things
as content management for knowledge aggregation and delivery is what
elevates the ESP as a Super Star Application. Once again the
ability to easily integrate diverse open source work products is far
superior to similar efforts using the inventory of proprietary software
to do the same. Usually you are captive to the branded portal
server solutions and tools. This runs its own risk of
obsolescence, branded market consolidation or product sun setting.
Unlike current
day enterprise applications the ESP architecture is not mired down with
legacy deficiencies or self serving perpetual IT solutions do to the so
called monolithic bound applications. Many layers of unproven
attempts of web services and so-called enterprise services
architectures, data mining or any other acronyms one cares to use to
describe yet another way IT can restructure the world to improve
business performance and competitiveness, is probably costing more than
it is saving. Or, in the final analysis, the increase in
productivity is marginal. The proponents of web services seems to
be a sales pitch for more IT services that can produce more metrics,
more data and better information. The evolutionary path of the
ESP through the use of an open standards based approach is to take the
next logical step, taking information and transforming it into
knowledge. This end product is what management finds useful in
all types of decision support situations. In the security and
intelligence community the term "actionable intelligence" is derived
from data and information that has been transformed into knowledge.
Web services in
a component based facilitated layer is all about optimizing and
automating the transaction process. To support new relationships
a very clean level of abstraction is needed in order to make it work
and it would seem very clear that as more component integration is
required the complexity increases as well. Thus process modeling
is used to configure and control the behavior of an application.
Components are used to integrate the enterprise applications.
This should not be confused with the idea of building the ESP with open
source packages or software work products running on top of a
supporting Java or object oriented implementation. As an example,
Silva is a Zope Product with its dependencies on Zope. The ESP
has its own processes defined where appropriate for example an incident
report that goes to the security director when filed. But the
development of new initiatives belies no dependencies on mounting a
portal technology or web services onto existing enterprise applications
and is not under the same handicaps as the aging enterprise
applications as we know them today. While we cannot deny that the CRM
and ERP enterprise applications have tried to raise the bar on
productivity, their failings often times were not necessarily from bad
design or logic, rather from weaknesses in human nature itself in the
use and appropriate applicability. There is very little literature
available on those colossal IT enterprise projects that failed whether
it happened during development, deployment or actual use with analytics
on operational results. It is sufficient to say that large firms like
Motorola have not been saved by their impressively expensive
development and deployment of enterprise applications to smooth out
their business fluctuations or even predicting that a major product
line is entering early into it's end of life due to competitive
pressures. They have the massive data but not the human means or
organizational behavior to figure out what the semantics implies. The
knowledge leading to sound actionable intelligence diffused to the
right people at the right time is simply not part of the equation.
Because the construction of the ESP is so unlike that of the enterprise
applications we know today, the portal will give firms strategic
advantages over fuzzy areas of decision support because the
customization to achieve such high level goals as knowledge, actionable
intelligence, and early warnings or alerts facilitates management with
preemptive powers and procedures in support of plans and policies.
Having said all
that based upon what we think the life span of the ESP will be, which
is of course an inexact prediction into the future, we think the ESP is
going to have a very unusual life cycle. Like most software products,
it will be deployed throughout the world, mature and go through the
usual revisions, but in a parallel dimension the ESP will start to
become a Knowledge Operating System. Initially, it will appear to
be like any other targeted enterprise application for use by management
in the daily mundane tasks of protecting operations and assets. But
like the child development process, the system will become smarter and
smarter as the database matures from the ability to aggregate data from
hundreds of client companies using the portal. Knowledge management
tools from text analysis, topic maps to summarization including early
warning alerts will accelerate the ESP into a second generation life
cycle. The chances of completing a second life cycle are very good
given the fact that it is built using the open source model. However,
the likelihood of advances in knowledge engineering will also propel a
2nd life cycle.
So, the question
has to be asked why is it important that this text be written? The ESP
will precipitate 4 general revenue streams during it's lifetime. First,
there is the revenue generated by it's subscription based Application
Service Provider model, second by the simple nature of firms engaging
the portal it will generate consulting revenue on the safety and
security incidents for directors, third it will generate IT consulting
development, support and administration and fourth as a knowledge
engine, Securitydirector will have the ability to sell it's knowledge
gold mine. Consequently, it is important to document both the
underlying technical infrastructure and the usability of the
application itself. The fourth revenue stream will come on-line
in the second life-cycle. It is also conceivable that licensing from
government or business entities that wish to have Total Content
Ownership or in-house deployments will be an additional source of
revenue. Revenue alone does not make this achievement important. Unless
the application delivers value to society in terms of a safer place to
live or work, it is irrelevant for purpose of fulfillment of
requirements. And neither can it be said that because it was built
through unique means of integration and package configuration, does it
make it important to write. What we do hope to show is that the portal
as host to the concepts of security thoughtfulness, planning, policy
formulation and procedures will become a referential work which
underlies the importance of its publishing.
The revenue
streams outlined above are internal to this final reorganization called
Enterprise Security Management (ESM). As a business object, ESM becomes
legitimate because now we can apply and execute on such principles as
Six Sigma. The pathways for statistical analysis and support empowers
owners, managers and executives to run this crucial sector of their
businesses at the strategic level.
ESM is a
conceptual way of viewing "the whole" while filtering out details about
its parts. This is sometimes referred to as "seeing the forest for the
trees." It recognizes the identity of a "higher level" aggregate
objects whose characteristics transcend the characteristics of its
component parts. Aggregate objects such as IT Security, Logistics
Security, Intellectual Property Protection, Workplace Safety, Executive
Protection, Incident Management, IT Information or Systems Security to
name a few all have underlying component parts. The component parts
become integral in computing the Process Capability Calculations. In
Six Sigma lingo this means measuring a sample of items or transactions
from a business process to make a projection of the number of defects
expected from the process in the long term. The defect rate, expressed
as DPMO (Defects Per Million Opportunities) is part of the essence of
Six Sigma. Expressing a business problem as a defect rate typically
provides a more direct way to communicate with stakeholders, members of
the project team or the strategic drivers who utilize ESM.
NOTE: The rest
of this document is a work-in-process with elementary ideas that will
be subject to change.
While I did not
research point blank whether or not a book about portal construction
would meet a market need, it is obvious that no portal built from CVS
or out of a proprietary box is useful out of the gate. Deployment is
trivial, tutorials are nice but taking it to the goal line is not.
Architects find the information they need to integrate applications and
portlets to be scattered all over the Internet. In fact, in most cases
even with proprietary products the vender's will not tell you were that
fine line is between their Wizards and where you have to get closer to
the metal and do some coding.
The following
outline is a proposed table of contents:
Securitydirector's
Enterprise Security Portal
Open Source
Portal Construction
From Humble
Beginnings to Advanced Knowledge Diffusion
Usage and case
management concepts.
Preface
Introduction
In general, we
introduce the Enterprise Security Portal (ESP) and remind the reader
that as an application, the product went through a formal design and
development process including the architecture of a database in support
of the business model and processes that came out of the requirements
definition phase of the project.
Chapter 1
Concept Owners
and a Business Model
This will be the
short story about the meeting of the Concept Owner and the Architect
describing the business model that predicates the technical
infrastructure contained in the following chapters. Brief historical
nature of traditional security consulting verses the Concept Owners
view of it being or becoming a function of management.
A discussion of
the realization that with the coming of portal technology, the vision
could be fulfilled. The business model description with the key players
or users of the system that begins to speak to portal customization and
profiling.
Key
characteristics as projected by the Concept Owner such as operational
security restrictions, incident reporting, messaging, content
management and publishing.
Chapter 2
Technical
Requirements Gathering and Open Source Harvesting
Working with
Concept Owners on a regular basis throughout the project, this chapter
begins with the research phase of the architect and ends with the
mapping of high level requirements to the low level packages of open
source software that will be integrated into the portal during the
construction process.
Harvesting from
the inventory of the open source community is a prelude to the
methodology and skills used to help developers understand the
relationships and the levels of integration necessary to bring the ESP
into production.
Even though
recognition for the need to add open source products came later in the
development and during construction, this chapter will fully disclose
everything in use to date. The reason for doing so is it is critical to
identify as quickly as possible to developers the points of
integration. This is were the real details that appeal to those looking
for the “how did they do that” and “what was the result” come into
play. The “why” will be obvious.
The following is
a general list of open source modules that will be the building blocks
for the coming of the enterprise portal for safety and security:
Linux
Development
Platform Tools & Resources
MySQL
Apache
Tomcat
Jetspeed
Velocity
Zope
Silva
The next 2
chapters will focus on the 2 primary areas that will illustrate the
construction process of coupling the various open source packages
through integration; system level security and profile management.
Chapter 3
Security
Integration Between OS Packages
Coverage of the
sequence of integration events that occurred to build the base without
SSL implementation. Second phase to address the development efforts and
system administration to configure, test and deploy SSL for an
end-to-end secure portal. Because so much of the content, reporting and
messaging must happen across the Internet, specific ssl encryption to
and from the client's agent browser must be secure.
We will look at
each point of integration to see what task had to be done in order to
facilitate ssl between the various packages. Reduce the complexity of
the throughput between Apache, Tomcat and Zope.
Chapter 4
Profile
Management Within The Context of the Jetspeed Servlet
Chapter 4 speaks
to the efforts of portal configuration and portlet deployment through
the ability to extend Jetspeed to support combining a user's base
profile with a custom profile. With marketing programs in mind for
different subscriber offerings, client company as well as unique user
profiles needed to be administered. Center of discussion does not focus
so much on allowing for users to customize their space as found in many
popular deployments, rather control of a subset and delivery of
portlets is emphasized.
This serves as
an introduction to portlet development efforts with respect to how this
becomes a tool in the hands of management.
Chapter 5
Portal User
Interface
The basic
interface for directors, client's Owner Manager and Employee
populations will be highlighted following technical discussion
illustrating what we have bought ourselves thus far. Portlet
descriptions with concerns of usability, weaknesses and strengths. A
portal without a content management system is no management tool.
Introduce the Zope Product and show how much it enhances the existing
Jetspeed framework.
Give some
background information on Zope, it's ZServer and object oriented
database. Document the content delivery methods of the Zope Management
Interface and the resulting server side code for targeted portlet
deployment.
While content
management is the foundation for controlling the paperless office and
diffusion of knowledge, publishing is the icing on the cake. Publishing
models through the use of Silva is the user friendly means to put the
power of content management in the hands of the authors, librarians and
consumers.