22 / 34
Policy&Practice
December 2016
22
technology
speaks
By Susie Gager
The Best Route to Take
C
olorado was one of
the first states to
have a federally certi-
fied child support system
nearly 30 years ago. The
Colorado Department
of Human Services
(CDHS) implemented the
Automated Child Support
Enforcement System
(ACSES) in April 1986. At
the time it was a cutting
edge Natural/ADABAS
system operating on the
mainframe. Because of its
sustainability, along with
the work of dedicated
Natural developers, more
than $500,000,000 in
child support payments
was eventually distrib-
uted through the ACSES
each year.
Despite the reliability
of ACSES, the Governor’s
Office of Information
Technology (OIT) recognized that
the infrastructure was hindering the
organization’s future strategic goals.
Each year the problems and operating
costs of the mainframe were growing,
while the number of skilled Natural
developers and compatible software
services was shrinking. By 2008, it cost
CDHS $1.5 million per year to operate
on a hardware platform that was
headed toward extinction. CDHS and
OIT decided it was time to move ACSES
from the mainframe to a modern,
state-of-the-art operating platform.
ACSES consisted of 1.5 million
lines of non-comment Natural code
and more than 500 million records of
ADABAS data. Creating the require-
ments for a new system would
have been a three-year endeavor
by itself. The state was determined
not to spend that amount of time
or taxpayer dollars on a high-risk,
“rip and replace” solution. They had
concerns because ACSES is a mission-
critical system processing an average
of 5,500 payments daily, supporting
more than 200,000 families. They
knew determining the right approach
was paramount to the success of the
project, so they assembled a top-notch
team to design a request for proposal
for a low-risk solution.
The team knew that ACSES con-
tained 30 years of dependable
business logic, so any solution that
could not prove to be 100 percent
functionally equivalent would not
be considered. This led them to the
conclusion that system migration was
the best course of action. The question
that remained: Who could help them
transform ACSES from Natural to Java
and ADABAS to a relational database
management system?
The team agreed they needed to
find an experienced partner, and
not a vendor selling a “big bang”
solution. They agreed that in technical
projects of this size, if a vendor falls
short of promises, the existing staff
absorbs the work. In addition, the
team and the state recognized that
the value of ACSES came from the
developers who built, grew, and main-
tained it. The migrated system had to
be maintainable by the existing team
of developers.
The team formed a selection com-
mittee that narrowed down 17 vendors
See Route on page 31
Photo illustration by Chris Campbell