|
Broadband Exception Reporting
Simplifying network monitoring by only addressing exceptions
By: Bruce Bahlmann - Contributing Author (your
feedback
is important to us!)
Created: October 12, 2003
This paper discusses the benefits of
exception reporting available from Birds-Eye Network Services.
The primary goal of every broadband service provider is to acquire
and keep subscribers. Keeping subscribers requires providing a service
that delivers consistent quality and reliability. Broadband service
providers often equate quality and reliability to enterprise network
management, configuration management, and other conventional means of
monitoring equipment. The problem with these approaches is that they
make a mountain out of a molehill and require you to spend significantly
more money than what is required to consistently deliver service quality
and reliability.
I tend to approach monitoring as I would any problem: a gradual
evolution of need. Going out and purchasing a full fledged network
management solution can be one of the silliest purchases a broadband
service provider can make because of the follow on “expectations” that
in having such a solution the service provider will attain unprecedented
service quality and reliability. Service quality and reliability is nine
tenths sweat and one-tenth software. If you look at it that way, you
will realize that what you first need is not a management solution but
rather just simple reporting. I call this the aliens approach to network
management. In the movie Aliens, trooper Vasquez says, "I only need to
know one thing, where they are!" I would argue knowing where they are in
essence represents 90% of what network management is supposed to do. The
additional 10% represents 90% of the cost of more extravagant monitoring
solutions yet only 10% of the actual functionality needed.
Importance of Delivering a Highly Available Data
Service
Quality/reliability is number one reason why anyone would change data
service providers so long as the price between available options remains
reasonably close (give or take $5/month). However, as the price
difference between options exceeds $10/month for the same advertised
level of service, quality gives way to economy. So, if your looking for
a way to deliver service quality and reliability on a budget, going out
and buying an expensive network management solution will not generate
increased revenue – thus the importance of not spending a lot of money
on a network management solution.
Delivering quality broadband services can be a very inexpensive
venture or a very costly one depending on the sheer amount of
information you require to make decisions about service calls and to a
lesser extent, equipment upgrades. The more information and
functionality required the higher the price tag – there is virtually no
end to the pricing and options.
I am a very big advocate of something called exception reporting. In
“pure” exception reporting, the goal is to boil down your entire system
into a short list of high risk or troubled devices that need some
attention. Broadband service providers armed with such information
become exceptional problem resolution detectives because of the
decreased complexity overhead required to produce exactly what is
needed. Ideally, this exception list is very, very simple and provides
only what is needed to find the device and fix the issue. You can
definitely get carried away with this approach and begin collecting
additional information (all of which becomes increasingly less useful).
One recommended approach to pure exception reporting is using
something called a pre-engineered report. In a pre-engineered report,
the result is fixed in terms of what it looks like, how it collects
information, and the target business sector it addresses. The only
required inputs in such a report are what sources and thresholds should
be used and how frequently should the report be generated.
Pre-fabricated reports provide consistent information regardless of
where they are run – so you can easily have a number of systems attain
the same level of exception visibility without going through extensive
configuration comparisons. Interestingly, such a report is significantly
less expensive to build and support compared to more traditional network
management solutions that must cater to a wider array of service metric
needs, configurations, and target businesses.
Sledgehammer vs. Nearest Stone
Essentially, network management solutions have grown so large,
diverse, feature rich, and expensive that it is increasingly difficult
to manipulate such software to attain even the most basic need – e.g.
where are my high risk or troubled devices? Whether you are a broadband
service provider just starting out or an experienced provider
maintaining hundreds, thousands, or even millions of subscribers, your
network management needs are still fairly basic. However, the current
mentality is that you need a huge sledgehammer to solve this problem
when using the nearest solid object will do. I call this mentality the
carpenter’s approach to network problem resolution: “If all else fails,
force prevails”. Force, in this case, being a full-fledged network
management solution.
Figure 1.0 Snap-Shot of
Marginal Modems Exception Report
An example of using a stone (as opposed to a sledgehammer) is a
Marginal Modems exception report (see Figure 1.0). A Marginal Modems
exception report provides cable operators with only a listing of high
risk or troubled subscriber cable modems – those operating outside their
normal range. Armed with such a report, cable operators can easily find
and repair subscribers with borderline service before they experience
sustained service interruptions. When such reports are sufficiently
organized, cable operators can further deduce individual cases from more
wide spread plant issues. Deriving such functionality from a
full-fledged network management solution (if it is even possible)
requires significant training, configuration time, perhaps even
consultants and can easily cost many more times the price of an
exception report.
One of the keys to building a successful monitoring strategy is to
approach it in a “just in time” fashion. In other words, add only the
level of monitoring your organization needs when you need it and then
gradually evolve (build upon) this extent of monitoring from that point
carefully and in tune with your needs. Remember that more network
management is never going to generate more revenue so excessive
monitoring or collecting too much device data only translates into an
unnecessary expense. Be penny and operationally wise when it comes to
network management rather than pound-foolish and reap the rewards of a
cheap but tightly run business.
This paper discusses the benefits of
exception reporting available from Birds-Eye Network Services.
Can Birds-Eye.Net help you or your Company?
Receive your Birds-Eye.Net articles and white
papers hot off
the presses by adding our RSS feed to your reader.
|