Académique Documents
Professionnel Documents
Culture Documents
Oberst
Enterprise Systems
Management
E
arly administrative and enter- plications might be well-versed in the tems’ ESM framework and tools in 1998.
prise applications grew up in use of their particular tool, the informa- Few colleges or universities had experi-
mainframe environments, where tion gathered may not be understood or ence with these tools, largely because of
inte grated management and even accessible outside of the group. the high cost, which large businesses
monitoring tools provided status infor- For managers, tracking down the source could justify in terms of revenue growth
mation as well as performance tracking of problems involves repeated calls and potential and income liability, and be-
and tuning. Answers to questions such queries to each of the systems in an at- cause of the overall complexity of these
as “How come I can’t get to the Human tempt to pinpoint the problem, with no systems. Most institutions have moni-
Resources System?” or “Why is the Stu- easy way to consolidate and aggregate toring in place for their earlier main-
dent System slow?” could be ascertained the information from the underlying frame solutions and have developed or
b y l o o k i n g at t h e m a i n f ra m e, it s layers. acquired point-solution tools for man-
processes, or the network between the Enterprise Systems Management aging networks and UNIX systems.
user and the central computer. But in (ESM) vendors attempt to solve this Until the growth of new, complex sys-
today’s distributed systems, these kinds dilemma with a coordinated set of moni- tems, most have not felt the need for
of applications depend on multiple NT toring tools that work all the way up the broader ESM tools. With Tivoli, Prince-
and UNIX file servers, back-end data- protocol stack, from network connectiv- ton began a partnership that provided
base servers, front- and back-end Web ity and operating systems to complex an affordable path to determine the via-
servers, distributed output devices, and enterprise resource planning (ERP) bility of these products in the cam-
sometimes application suites, and that pro- pus environment. In addition to the
servers or transaction- vide a common reposi- systems-monitoring tools, Princeton also
processing monitors, tory for operation and acquired network monitoring (NetView)
along with all the net- performance monitor- as well as helpdesk (Tivoli Service
work path dependen- ing at all levels. At the Desk), job scheduling (Maestro), and
cies among these ele- h i gh e st l aye r, E S M output management tools (Destiny).
ments. Answering the tools attempt to model Our initial efforts focused on design-
same type of questions and monitor business ing and implementing the underlying
today is a much more systems and practices framework for Tivoli and establishing
difficult task because s o t hat t h e o v e ra l l independent implementation of the
of the complexity of health of a system or other products. The last three were es-
the systems and the va- application (e.g., pecially time-critical, since they were
riety of management Hu m a n Re s o u r c e s , replacing systems being phased out,
tools available. Student Records, Finance) can be deter- and efforts at integration took a back-
Each platform in a system (NT, mined. When problems occur, drill- seat. For the initial deployment, three
UNIX, Windows, network hubs and down tools permit recursive querying of staff positions were loaned to the roll-
routers, etc.) might have a different tool the supporting components to identify out effort. A year later a reorganization
for monitoring its operation and per- underlying causes. created a three-person ESM group to
formance. Above the operating system, run systems management, job schedul-
each of the services and applications Enterprise Systems Management ing, and output management. In many
might itself have a separate reporting at Princeton University of the large-industry Tivoli implemen-
and monitoring tool. Even though the As part of a broad implementation of tations, each of these efforts would have
groups responsible for the operation of new administrative systems, Princeton three to five people assigned to it, so our
each of these systems, services, and ap- University began to deploy Tivoli Sys- implementation to ok longer than