Manageability: The Underlying Resource Framework
of the enterprise is dependent upon the underlying resource framework
which supports the other disciplines. It consists of:
Level Agreements Organizational
Systems Management Infrastructure
The following section defines each of the key areas required to
ensure successful management of the enterprise.
ensure reliability, availability and serviceability objectives are
met, overall service levels must be defined, implemented and maintained.
Service level agreements are generally instituted between a department
providing a service and the user who is bein
provided the service.
levels not only provide the user with a definition of what to expect
regarding such things as system response time, mean time to repair,
etc. Service levels also support those providing the service by
defining measurable objectives which can be referenced in the event
of a service disruption.
service levels to be meaningful they must be documented. The
documenting of service levels provides a reference point for
all parties involved. Service levels should not only be documented,
the documentation should be distributed, agreed to and maintained.
of service levels should occur on a regular basis. If an area
is not meeting a particular service level the following should
caused the service level
can the service level miss be prevented from occurring again?
the service level valid (i.e., is it an attainable service
the first two points are obvious, the third item may not be.
For a service level to be effective it must be attainable.
If the service level cannot be met under ordinary circumstances,
changes to that service level may be required. An unachievable
service level, however, should not be confused with a regularly
missed service level which is the result of poor performance.
satisfaction is the ultimate goal of service levels. For this
reason, levels of reliability, serviceability and availability
must meet the needs of the customer. However, service levels
must also be realistic from the provider's standpoint. While
a user may believe that a system should be available 100% of
the time or, if down, should be fixed in a minimal amount of
time, this level of service from the provider may not be possible.
An effective service level is not only regularly attainable,
but when met, satisfies the customer.
systems management requires an organizational structure to be defined
which supports the needs of the enterprise. Ideally, the organization
should be created along functional, rather than organizational,
boundaries. This structure minimizes redundancy and provides users
with a more clearly defined organization.
ensure individuals within the organization and users outside the
organization understand who performs each system management function,
the following must be defined:
the enterprise with organizational charts is critical to ensure
users know who is responsible for each area of the systems
environment. The organizational chart should list each individual
within the organization and define what they do and who they
descriptions are used internally by the systems management
organization to define what each individual is responsible
for. The job description at a minimum should define overall
responsibilities, objectives and performance measures.
is important for the organization and its components to have
a clearly defined charter. The charter must contain the vision
of the organization, the mission and strategies. The vision
defines what each area is striving to achieve, the mission
defines the objective of the organization, and the strategy
indicates how the organization will achieve its mission and
vision, mission, and strategies must be understood by those
inside and outside the organization. Each individual within
the organization should understand their role in contributing
to vision and mission. Each department should have a clear
and concise mission which should support the overall organizational
plays a critical role in successful system management by allocating
appropriate individuals to specific tasks and ensuring that those
individuals have the skills necessary to perform their roles within
Within a changing environment, staffing becomes a tenuous process
at best. Management is required to support projects utilizing new
technology where skills may or may not be available. To compound
the issue many projects require a specific skill set for a short
period of time after which the skill may not be required.
an environment where new technology is introduced on a daily
basis, proper training and education are critical to successfully
managing the enterprise.
must be provided in a manner which supports an individual's
ability to do their job. As new technology is unveiled, training
must be provided as proximate to the implementation date as
possible. If training is provided too soon, newly acquired
skills may not be utilized on a day-to-day basis and may be
forgotten before the new technology is implemented.
and education need not be formal; brown-bag discussions and
one to one training are oftentimes more beneficial than formal
plays an important role in developing an effective systems
management infrastructure by providing a mechanism for individuals
to learn from the experiences of others. With mentoring, individuals
with certain skill sets are used as "role models"
to promote those skills. Within changing environments, mentoring
is key since it provides a mechanism for individuals to quickly
learn the skills of another.
pathing provides longer-term benefits by supporting an individuals
need for personal fulfillment. A career path provides an employee
with the road map necessary to achieve success. A career path
need not lead only to a management position, many individuals
may seek a more technical path. The goal is to provide individuals
with real expectations regarding personal growth within the
organization and a method for realizing that growth.
change may motivate certain individuals, others are intimidated
by change. Consideration must be given to defining what is required
to motivate individuals to excel within the changing environment.
Morale within the organization must also be considered since
high morale levels are necessary to ensure continued success.
days past, facilities were centered around the needs of the mainframe
and supporting peripherals. In today's world, decentralization of
the processing environment requires that consideration be given
to key components of the facilities infrastructure, including:
following section provides definition for each area and the special
circumstances which may exist as a result of the distributed environment.
relates to the placement of systems within the enterprise.
When mainframe systems were the only concern, location considerations
were based upon finding space within the data center. In today's
environment, while space is still an issue, midrange systems
and servers may reside outside the data center.
to the mainframe environment where dedicated communications
channels ensure adequate response even across large distances,
today's smaller systems must contend with inadequate networks
to support ever-increasing bandwidth demands. Often, placing
a server in the data center may not be feasible since response
time may be adversely affected.
power and proper cooling is paramount to minimizing system
downtime. As systems are no longer restricted to the data
center floor. The result is the need for procedures to define
acceptable system locations and methods for providing adequate
facilities to support them.
in today's environment is also complicated by the fact that
systems are distributed. Security in this instance does not
involve data security, but rather system security. Mission
critical systems must be protected with regard to access regardless
of being midrange or mainframe.
must address the need for growth within the data center. As
systems come on line, space must be available to support these
systems. Concerns regarding proximity and other facility components
(e.g., power, security) must be addressed and planned for.
cornerstone of today's processing environment revolves around
an adequate network infrastructure. Facilities must be created
in such a manner to support both internal network considerations
as well as external network issues.
a central point of control for today's complicated environment
has many benefits including:
points of contact
command center for the mainframe environment was easily constructed
since network and system issues were minimal. In today's environment,
complicated systems are managed with an ever-growing number
of tools. Although the need for a command center remains,
creating such a facility is a large undertaking.
operations are required to stay competitive. As with most other
areas of systems management, sustained operations has been made
more difficult with the implementation of distributed systems. The
key components critical to uninterrupted operations are:
standards are essential to a successful Operations environment
and are based on the ability to reproduce consistent actions
in every area of the environment.
procedures provide the framework for successful Operations.
Procedures are documented and when changed, are updated. The
procedures are consistent in their presentation and are distributed
for understanding throughout the organization.
achieve centralized, exception-based management of the Operations
environment, a continuous evolution towards automated operations
is necessary. There are two basic forms of automated operations,
lights dimmed and lights out.
Lights dimmed is the implementation of an Operations environment
where minimal staff on-site is utilized to run the systems as
required. Lights out operations requires no staff on site and
can be remotely managed.
support business changes dictated by the more competitive environment
of a global marketplace where low cost and superior performance
are key to business success, business processes need to drive product
development and deployment. Business process re-engineering is defined
as a rethinking of the role technology plays in support of business
functions. With the price/performance benefits of client/server
computing, many systems applications are being architected and deployed
on alternative computing platforms to the traditional central host-based
MVS architectures of the past.
The primary obstacle in the deployment of these alternative platforms
is the lack of a unified management system with which to support
them. In the MVS host-based world, system management has evolved
over the past two decades resulting in a cost breakdown of 20% for
support and 80% for hardware and software. Within the client/server
based computing environment, this relationship is reversed with
20% comprising the costs of hardware and software and 80% for management
of the environment.
Besides the standard system development practices of full user-based
acceptance testing to ensure quality systems meet functional requirements,
standardization of the operating environment becomes even more critical.
The elements integral to this standardization include:
Production acceptance to ensure operational compliance
Configuration management to ensure supportability
Version control to ensure functional integrity
Software distribution to ensure standardization.
Technology organizations can support an operating environment
in a cost effective manner only by limiting complexity and
reducing the number of nonstandard actions needed to provide
service. With business drivers requiring multiple technology
platforms, operational acceptance criteria must be defined
for all components of this extended enterprise. This means
that technology decisions must include supportability considerations
and that standards be defined and enforced as part of the
process of accepting systems into the production environment.
Use of standardized middleware and system management tools
should also be part of the acceptance criteria.
meeting the challenges of distributed systems management,
users face a multitude of requirements that fall into two
major categories: Event Management and Configuration Management.
Configuration Management disciplines will increasingly take
center stage in systems management as a means of controlling
environmental quality, enforcing policies and reducing complexity.
a focus on Configuration Management, Event Management will
be fighting a losing battle for availability, efficiency and
service quality if the sources of events and user calls are
not a key part of the management process. This requires configuration
management tools and processes and can be resolved with a
framework-based configuration management solution using pre-tested
components to enable source problem reduction.
in the host system environment where modules are checked in
and out of a source library and changes to production proceed
from testing through a change management system to ensure
system integrity, the client/server development and production
environment needs similar tools and procedures for version
the host-based environment, software distribution is a centralized
process controlled by change management procedures and staging
tools. However, in the far more complex world of client/server
computing, there are three key reasons why software distribution
will continue to grow as a strategic management issue:
The number of desktops deployed across enterprises continues
Major vendors are entering the market with enterprise offerings
The advancing storage and processing capabilities of desktop
factors enable and/or "force" more organizations
to load productivity applications on workstations, therefore
requiring updates, maintenance and general configuration management.
Software distribution tools which simplify the process of
updating software components and ensure enterprise compliance
with approved configurations are necessary to avoid service
outage and resultant problem resolution resource impacts.
software industry is both caught in and leading an architectural
change that is sweeping the entire technology scene. Corporate users
want to balance and optimize platforms so that application tasks
can reside where they are most cost and technology effective, and
where they can communicate and share information across centralized,
distributed and personal systems. The cost and style of computing
vary across platforms, and users are trying to achieve architectural
balance that exploits these differences. The critical missing pieces
in this networking puzzle are system software and middleware, which
provide the software "glue" that gives applications levels
of portability and connectivity.
systems and network management framework that makes use of middleware
objects simplifies this otherwise complex undertaking. A framework
sits between the operating system and distributed system management
functions, providing a common set of services to each. Without frameworks,
each distributed system management function would have to provide
this set of services (which is the problem with implementation of
point solutions) and any changes made to one would need to be proliferated
to the others, creating an update currency issue and significant
procedural complications. Without synchronized information available
in a framework, routine tasks such as software update distribution
becomes complicated and failure prone if the system is not alerted
that a network link is down or a node does not have the RAM, disk
or processing power to accommodate the update.
support is an iterative process which defines technology trends
and the support structure necessary to make them operable, then
implements through procedures for software and hardware installation
and maintenance and administration of the systems. While system
versions are in operation, the next wave of technology to support
upcoming business requirements is in the evaluation process, thus
creating a dynamic and market-leading technological environment.
Installation and Maintenance
systems become more complex, they risk becoming unmanageable.
This can happen by either increasing the user population or
increasing the number of managed elements. Environments that
are manageable when running a small number of users risk becoming
unmanageable as the user population grows or as the number
of components grows. In an environment where the number of
users must, to support business, increase significantly, the
only way to maintain a controllable number of elements is
to control the number of separate software components in use.
To support this goal, system software installation and maintenance
tools and procedures are necessary.
Installation and Maintenance
with software components, the complexity of the supporting
hardware environment determines its manageability. Standardization
of the hardware environment is necessary to maintain high
levels of system availability. To assist this effort, an up-to-date
asset inventory is essential. With a systems management framework
where automated component discovery is supported, the processes
of upgrading system hardware will evolve into standard support
Research and Development
into the next wave of deployable technology must be driven
by strategic business requirements with a constant focus on
supportability, with the understanding that services that
are too complex to be adequately supported will not provide
the high level of quality and low cost of service that the
global business market will reward. The support infrastructure
and tools must be part of the application research testing
process for cost effective technology options to be defined.
complex environments, system administration comprises the ongoing
monitoring of multiple interlinked components and, in order
to be handled in a cost effective manner, requires an automated
toolset deriving its policy information from a single source
and producing management information in a holistic fashion.
Component failure identification is necessary for recovery and
prevention, however from a system availability perspective,
a service outage no matter what the cause is a service level
miss for the customer and should be centrally tracked toward
the goal of customer satisfaction, continual service and product
JJ Kuhl 2002