[Editor's note: Alex Pinchev is the President and
CEO of MainControl, Inc., a privately held company which develops and sells IT configuration
management and tracking software. MainControl has over 100 enterprise companies
world-wide, managing close to 1.5 million PCs, including Computer Sciences Corporation,
Andersen Consulting and Canadian
Occidental Petroleum. You can find their web site at http://www.MainControl.com]
from Executive Software Team Jump Executive Software>@@@ How widespread a problem is the Year 2000 (Y2K) bug? While The Y2K
problem applies across industries, affecting any organization that
relies on date-based calculations to conduct business, especially
vulnerable are organizations that have developed and deployed
mission-critical applications to be run in the distributed computing
environment. Examples include industries such as finance, banking and
insurance, which utilize complex spreadsheets and databases to conduct
business.
The range of computing areas affected includes the PC BIOS (basic
input/output system), the real-time clock, operating systems, commercial
applications, homegrown programs and data. Applications most impacted
by the problem are spreadsheets and databases with date-based
calculations - examples include loan payment schedules, insurance
policies, retirement benefit formulas, etc.
But Y2K problems and solutions are significantly more magnified in a
distributed computing environment (such as those running Windows NT or
NT/95) than in a mainframe environment, due to its more complex nature.
In an enterprise running thousands or tens of thousands of PCs, the
number of applications that hide potential Y2K problems can be
staggering! Examples include a small spreadsheet application developed
by the accountant that feeds the company's general ledger, the sales
application used to manage and report sales and revenues, and the dozens
of client/server deployment projects where parts of corporate developed
or purchased applications were deployed on thousands of PCs. In an
average size enterprise, there could be thousands or tens of thousands
of applications, utilities, routines, etc., which could affect the
corporate applications and spoil an entire Y2K effort.
The key to even beginning to get a handle on a Y2K problem for any
company is to collect comprehensive inventory information about the IT
assets, including hardware, commercial applications, homegrown solutions
and files. This must be a very complete inventory. There are many
complex enterprises, with tens of thousands of PCs scattered around the
globe, that do not have information on how many PCs they own, where
they're located, or who's servicing them. They are running a variety of
operating systems, using different protocols, with thousands of software
applications in use. It is virtually impossible to address the Y2K
problem in these enterprises without information on what
hardware/software is being used.
Which brings us to the corrective steps to the Y2K problem within any
enterprise. The process for correcting the Y2K problem is: find and
identify the software, identify the suspect locations in the code that
deal with dates, correct or replace the code or the product, and test
and run the corrections in production.
That is, of course, easier said than done. How does one find and
identify all the software on thousands of distributed computers? And
once found, how can one make sure that the user's PC can support the
corrected application or the new software package? What if the solution
requires another four or six bytes per field - is there enough disk
space to support the larger database?
And, after correcting the problem, there is a need to re-deploy the
application to the end-user's PC. How does one accomplish this task for
thousands of applications and products for thousands and thousands of
desktops without errors? It could be a logistical nightmare. And there
is more to consider: how does the enterprise monitor that no new
unauthorized software is loaded onto the end-user's PC - from the moment
the problem was solved?
It can be easily seen that seeking professional help on this matter is
essential, unless your company has a dedicated group of experts to
accomplish the above steps. Do yourself the service of hiring a
qualified consultant to oversee the operation - remember, if you miss
computers or software in your correction or implementation, you could
nullify the whole effort. Additionally, a consultant may have ways to
automate portions of the process that save thousands of hours and are
for more thorough. For example, the first issue to address is how to
find and identify the end user hardware and software configurations. My
company has a product called MC/EMpower which builds an automated
enterprise-wide IT inventory that will present an updated view of the
current installed hardware and software configuration running on
thousands of users' desks. In this way, all software applications,
packages, and operating systems will be identified and ready for
evaluation from a Y2K perspective.
The second step is correcting or replacing code or software, once these
areas have been identified. This can involve some tough decisions, as
the preliminary step must be taken to determine the time and money
required to fix the identified problem areas. In order to determine
the best solution the company must determine: Is it cheaper to fix the
problem or buy new systems? How can the problem be solved by the
deadline? How much effort is required for either alternative? What are
the benefits/risks associated with each option? These are questions,
which must be answered before proceeding.
As part of this step, you must make sure that the desktop configuration
will be able to support the new Y2K compliant applications or vendor
software. If the inventory information collected in the first step was
done well, it will be quite easy to determine whether the end user
configuration has enough memory, hard disk space, and any other
prerequisite hardware or software in order to support the Y2K compliant
software.
The third issue is deployment of the corrected and Y2K-compliant
software, which can also be a very long and painful step. If at all
possible, you should avoid a machine-by-machine deployment by using
electronic software distribution. Our product MC/Empower uses such a
method, so that the fixes or new software can be electronically
distributed and correctly installed on all conforming end user PCs. In
this way, the enterprise ensures that all users get the same version of
software and all the fixes to address the Y2K problems.
This is a rough guideline of how to implement a Y2K solution in a
distributed PC environment. It is complex and involved, but can be
greatly simplified with the right solutions and professional help. @Macarlo, Inc. @Macarlo's Shareware & Web OS/2 Java Lobby Member
Java Site Accredited