19
WWW.7X24EXCHANGE.ORGcommercial products such as hp
openview and ibM tivoli.
additionally, available solutions differ
based on the nature of the business
and the required usage such as web
application monitoring solutions
designed for monitoring transactions
by simulating user transactions. even
solutions in the same category such as
it infrastructure monitoring differ in
their licensing strategy and for high
growth environments solutions that
charge by device or monitor can
quickly increase the cost.
in this paper, we will share saudi
aramco’s eXpeC Computer Center
experience in choosing the proper
monitoring solution for data centers
that mainly use hpC systems and
what considerations need to be taken.
A. IT SERVICE MANAGEMENT
an effective monitoring solution of it
resources is a main part of it service
management (itsM), an approach
necessary for organizations to
manage large-scale it systems. itsM
is a process based approach to
manage an organization’s it systems
with a focus on it services rather than
it systems [2]. it aligns an
organization’s it services with
business processes to ensure
customer needs are effectively met. it
introduces management processes to
cover the entire life cycle of an it
service including; planning and
designing of a service, delivering the
service, support and finally continual
service improvement [3]. several of
these processes are dependent on it
systems monitoring such as change
management, problem management,
incident management and
configurations management [4].
having effective it service
management should be a strategic
goal for any organization, especially
organizations with medium to large it
systems such as an hpC in our case.
today, various best practices exist that
if implemented will ensure effective
itsM. available frameworks include,
but are not limited to information
technology infrastructure library (itil)
[3], Capability Maturity Model
integration (CMMi) [4], Microsoft
operations framework (Mof) [5] and
Control objectives for information
and related technology (Cobit) [6],
where standards include iso 20000
and iso 9000.
II. BUSINESS SERVICE
MANAGEMENT
a common element in an information
and Communications technology
environment is the business service
Management (bsM), which is a
standard approach to model, monitor
and manage it infrastructure in a
customer-centric and business-
oriented configuration. it is normally
positioned on top of itsM to ensure
that business and customer objectives
provide an input to define the itsM
model, and the business services
offered by the it organization. the
bsM process has driven the
development of the traditional
network monitoring software industry
to use complex layers of differing
technologies to close the gap
between knowing the problem from a
technical point-of-view and the
business impact. bsM has been
identified by the itil as a best
practice for it infrastructure
management and operations.
bsM consists of both a structured
process and enabling software,
allowing it organizations to operate
by service rather than by individual
configuration items (Cis) or
technology. this has led organizations
to prioritize their efforts, and
ultimately improve the services
delivered to upstream users. that
said, bsM is a way to bring together
many disjointed processes and
applications, and develop quantifiable
improvement in efficiency and ability
to view it infrastructure relevant to
business processes.
also, bsM, as a process, is not
software by itself. however, there are
tools and software solutions that are
considered critical enablers to
implement bsM in an organization [7]
[3]. thus, bsM software and tools
have been popular over the years.
Figure 1: Key Elements to Build a BSM.
unlike legacy network management
systems, bsM software [7] [8] provides
a unified view for data center
administrators to view and manage
devices, applications, networks and
events usually from a common
dashboard. traditional network
management systems focus on
measuring and monitoring the
technical availability and performance
metrics of it infrastructure, such as
Cpu and memory utilizations.
however, these tools do not provide
the necessary information to
determine the business impact of a
specific problem. for example, if a
server and a storage filer fail at the
same time, legacy tools cannot
determine which of these two
elements is more critical to the
business or which business processes
have been impacted by the failure.
III. FACTORS TO
CONSIDER
in recent years, there has been a rapid
development of software solutions to
help organizations implement the
bsM process. as a leading scientific
data center, eCC has the experience
BSM
Business Processes
Applications
Middle-Tier
IT Components
Data Center