Reliability and performance testing is the only way to guarantee quality of service
- Posted:
- 16:46 26 May 2005
- Topics:
- IT Standards & Protocols
Any business planning to implement or improve an IP network
without testing all potential scenarios is sure to create a network
territory where users fear to tread.
The ultimate test lies in being able to set up and run an IP
network that delivers a delay-free, high-quality service to
users.
Although some enterprises expect to ensure a high quality of
service across the network by throwing bandwidth at the problem,
this is expensive and inefficient. They would be better off taking
a phased approach to equipment and network testing, service
validation, and ongoing performance, reliability and service
assurance.
Trust in the network hardware is crucial. It is vital that the
equipment that is going to be implemented has been tested
thoroughly before it comes anywhere near the organisation. That may
sound obvious, but all too often equipment is tested in the lab
with little regard to what might happen in the real world.
IT departments also need to be able to guarantee the requirements
of various applications and services and ensure that their networks
are ready to support them. This means that pre-deployment testing
needs to be tailored to individual business requirements.
Active tests to review how a network would behave under different
conditions should give a good idea of how the kit will react in the
working environment. For example, how will it cope with heavy data
loads or what will security performance be like under denial of
service attacks? By recreating real-world conditions in the lab,
you can see what the quality, reliability and throughput will be
like, thereby ensuring that the network will provide the level of
service required by the business.
However, lab-based trials are just the first step in preparing for
a new or improved IP network. Although this type of testing will
identify a large percentage of problems, live networks often bring
with them unexpected difficulties which cannot be identified in the
lab. Once a network goes live, it is in a constantly changing
environment. New users and access devices come online, technicians
reconfigure and misconfigure equipment and traffic patterns change
by the week, the day, the hour or even the minute. This is where we
need to start looking at performance and quality of service
issues.
Quality of service should embrace bandwidth allocation,
prioritisation and control over network latency for network
applications, but this can only be done when a network is operating
at full capacity to ensure that data traffic can cope with the
mixture of data and real-time applications. This is also the only
way to avoid bottlenecks and congestion, which can be caused by
sudden bursts of traffic and can hit various parts of the network,
including the internet gateway, switches or even heavy voice
traffic.
In some cases, problems pass users by. They may tolerate sluggish
e-mail transmission, slow file downloads and patchy internet
browsing, but they are unlikely to be forgiving if direct
communication links such as voice or video conference calls are
erratic or of poor quality. A half-second delay on a link
delivering voice will cause the user to hang up in frustration, and
rightly so.
Testing of network performance and reliability must be continuous.
And it is vital that enterprises can identify possible anomalies
and plan for worst-case scenarios so that disruption is kept to a
minimum. Test equipment experts can provide the right tools to do
the job and also the methodology to ensure that the correct
procedures are followed. The result will be a high performance,
reliable network.
So, you have tested before you implement, you have the tools in
place to assess performance now the network has gone live, but what
do you do if problems do arise? As more IP networks are deployed,
there is a growing pressure on enterprises to recruit and retain
the correct level of engineers to diagnose and resolve network
problems. However, with demand for skills high, maintaining the
correct staffing levels can be a problem.
Using test and diagnostic tools that are simple enough for less
skilled engineers is helping many businesses overcome staff
shortages. They can gain considerably by equipping less skilled
network engineers with automated test and diagnostic tools. This
gives them the ability to fix basic network problems without having
to escalate the problem to technicians, who may be in short
supply.
So the road to high performance IP networks is long and winding but
one thing is certain: sound testing and fast diagnosis is crucial
to rolling out and maintaining high quality, reliable networks.
Undertaking their own testing and monitoring can help large
enterprises looking to cut the operational costs associated with
implementing and maintaining corporate networks based on IP, as
well as reducing the risk of major failures.
Mark Orchart is director of European marketing at Spirent
Communications