| Networking is getting tougher. Networks
| |
| | transient network conditions and
|
| must deliver a growing range of services,
| |
| | discovering the root causes of these
|
| from ERP, CRM and email to VoIP and web
| |
| | problems.
|
| services applications, each of which has
| |
| | 4) Accelerated time-to-benefit for new
|
| its own idiosyncrasies and requirements.
| |
| | and/or upgraded applications
|
| Each new service introduced onto the
| |
| | When C-level executives decide to make
|
| network contends for available resources
| |
| | investments in new applications and
|
| with every other service, impacting the
| |
| | services, they want to see those
|
| network's ability to support the
| |
| | investments pay off quickly. That's why
|
| business.
| |
| | the slow, staged production roll-outs of
|
| Meanwhile, the network itself is
| |
| | the past won't cut it anymore. Instead,
|
| constantly changing. New locations are
| |
| | networking teams need to be able to
|
| added - some of which may be in another
| |
| | quickly deploy new applications across
|
| country or on another continent.
| |
| | the enterprise. This can only happen if
|
| Equipment is upgraded and/or
| |
| | caution and uncertainty about the actual
|
| re-configured. New management and/or
| |
| | behavior of these applications in the
|
| security tools may themselves impact
| |
| | production environment is replaced by
|
| service performance. Decisions about data
| |
| | confidence and certainty in '05.
|
| center consolidation and business
| |
| | 5) More intelligent planning for and
|
| re-organization also affect the network
| |
| | support of business growth
|
| in different ways. All of this makes the
| |
| | Network managers constantly have to cope
|
| network a highly dynamic environment
| |
| | with change. They have to determine how
|
| where even subtle changes can have a
| |
| | increases in network utilization will
|
| major, unforeseen impact on application
| |
| | affect application performance. They have
|
| performance and availability.
| |
| | to decide how to best engineer the
|
| Yet business users expect this complex
| |
| | network to support business expansion,
|
| environment to be as reliable as
| |
| | re-organization or mergers and
|
| electricity - despite the fact that
| |
| | acquisitions. However, they can only do
|
| networking budgets are not being
| |
| | so if they have an effective means of
|
| increased in proportion to these growing
| |
| | performing capacity planning tasks and
|
| challenges. So network managers can't
| |
| | assessing a full range of "what-if"
|
| simply over-provision network
| |
| | scenarios. Such scenarios are also
|
| infrastructure to make sure every service
| |
| | critical for formulating realistic
|
| has all the bandwidth it needs. Moreover,
| |
| | contingency plans that can ensure
|
| over provisioning may not even solve the
| |
| | business continuity under a variety of
|
| problem and/or ensure the required level
| |
| | possible conditions.
|
| of performance.
| |
| | Looking at these challenges, it quickly
|
| That's why network managers are facing
| |
| | becomes evident that conventional
|
| many challenges, including:
| |
| | production network management tools alone
|
| 1) Pinpointing potential network
| |
| | are no longer sufficient for today's
|
| performance issues early in the
| |
| | networking teams. These tools are great
|
| development lifecycle
| |
| | for monitoring the production network and
|
| Ideally, the impact of the network on a
| |
| | discovering certain types of problems -
|
| new application or service should be
| |
| | but they don't enable network managers to
|
| dealt with from the very beginning of the
| |
| | validate new technologies and
|
| development process - when potential
| |
| | applications before they're deployed on
|
| problems are much easier and less
| |
| | the production network. They also force
|
| expensive to fix. Unfortunately, this is
| |
| | network managers to solve problems that
|
| rarely the case. Problems with an
| |
| | should have been addressed in application
|
| application's "networkability" are
| |
| | design.
|
| typically discovered only after its
| |
| | Conventional tools aren't very helpful
|
| roll-out into the production environment
| |
| | for troubleshooting intermittent and/or
|
| is initiated. At that point, it's usually
| |
| | transient network problems either, since
|
| too late to make any significant changes
| |
| | they don't provide a means of
|
| in the application's design. So the
| |
| | reconstructing and analyzing such
|
| problem gets pushed onto the shoulders of
| |
| | intermittent conditions. Nor do they help
|
| the networking team. That's why, in '05,
| |
| | accelerate production roll-outs,
|
| smart network managers will focus on
| |
| | facilitate experimentation with "what-if"
|
| nipping these problems in the bud.
| |
| | scenarios, or support formulation of
|
| 2) Validating new or modified
| |
| | network contingency plans.
|
| applications and infrastructure before
| |
| | So what's an overworked, under-resourced
|
| they are deployed in production
| |
| | network manager to do? The answer is to
|
| As the network becomes more complex and
| |
| | look at network modeling technologies.
|
| more critical to the day-to-day-operation
| |
| | These technologies provide an environment
|
| of the business, network performance
| |
| | in which new applications, technologies
|
| related risks associated with application
| |
| | and problem-solving strategies can be
|
| and infrastructure change are continuing
| |
| | safely and thoroughly evaluated. Because
|
| to rise. In fact, some of the worst
| |
| | they allow an application's network
|
| business interruptions that companies
| |
| | behavior to be fully validated before
|
| have historically experienced have not
| |
| | it's deployed in the production
|
| been the result of unexpected equipment
| |
| | environment, these technologies also
|
| failure. They've been the unexpected
| |
| | empower network managers to perform more
|
| consequence of a planned modification.
| |
| | rapid, glitch-free roll-outs. Plus,
|
| Networking teams must therefore implement
| |
| | modeling technologies are uniquely able
|
| change management best practices in '05
| |
| | to provide insight into any number of
|
| that prevent them from having to put out
| |
| | "what-if" scenarios - so network managers
|
| fires that they accidentally started
| |
| | can make plans for growth, corporate
|
| themselves.
| |
| | re-structuring and/or disaster recovery.
|
| 3) Improved troubleshooting of
| |
| | "Empirical" modeling solutions offer
|
| intermittent/transient network problems
| |
| | today's network management teams
|
| One of the most frustrating things for a
| |
| | particularly excellent business value,
|
| network manager is dealing with a problem
| |
| | because of their accuracy and relative
|
| that keeps disappearing before it can be
| |
| | ease of implementation. This accuracy and
|
| adequately understood and remedied.
| |
| | ease is achieved by running the actual
|
| However, as the business's tolerance for
| |
| | applications against a model that uses
|
| network interruptions continues to drop,
| |
| | captured conditions from the production
|
| these intermittent problems will become a
| |
| | environment. The result is a clear
|
| bigger management issue. So this year,
| |
| | understanding of the user experience well
|
| network management teams need to develop
| |
| | ahead of deployment.
|
| more effective methods for capturing
| |
| |
|
