When the NHS required a complete IT solution, it turned to Family
Health Systems to help – and it turned to Network Instruments
As part of England's National Health System (NHS), Family Health
Systems (FHS) is a non-profit group committed to providing the
country's health organisations with the IT solutions they need to
improve patient care and maximise resources. With its headquarters
in Exeter, Devon, the FHS has offices located throughout England
and offers a complete IT solution for the NHS, wherever the
location. For the past 15 years, it has supported, maintained and
developed the Exeter System, software that provides fundamental
operational support to all of England's health authorities.
The ProblemFHS is committed to offering fast response times
to network problems. In order to continue to provide quick
response, comprehensive network monitoring became increasingly
important. All FHS support calls are received and managed by its
Helpdesk in Exeter and are then directed to the appropriate support
service in the field.Monitoring the network was fragmented at best,
so FHS had to expand its overall strategy of monitoring network
traffic. "We were using a number of products that highlighted bits
and pieces of what was happening," says Andy Dawson, network
communications engineer for FHS. "But we didn't have a coherent
package that would give us all of the information."
The
SolutionThe FHS network consists of several Ethernet LANs
located at its headquarters in Exeter and at each of its four
branch offices. These LANs are connected via an FDDI ring. It also
utilises about 40 NT servers, 12 NetWare servers and 20 Unix boxes.
With about 250 users throughout the country, there are a number of
problems that can and do occur on a regular basis. "We wanted
something that was integrated, that could do a good overall general
analysis and that would allow us to look at all segments equally,"
Dawson says.FHS ran across Distributed Observer from Network
Instruments on the Internet. Realising that it might be something
that could enhance the network, Dawson downloaded an evaluation
copy and was immediately impressed. "For starters, Observer had
packet capture and analysis information that we found useful," he
says. "It is really important to see what the actual traffic is
doing, and Observer allowed us to do that."After discovering
Observer, Dawson found Distributed Observer, a product that works
in a client/server environment and enables FHS to actually monitor
all of its remote branch locations throughout the country. With
Distributed Observer, companies stay informed with real-time LAN
traffic information and ongoing trend data that helps them quickly
pinpoint problems or anticipate potential problems before they lead
to disaster ( all from a single PC at one location using a
convenient Windows interface. By providing remote protocol analysis
and complete network monitoring from a central location,
Distributed Observer lets companies be in many places at one time,
monitoring multiple LANs and even multiple sites. They can view
each LAN more clearly and see all network traffic in real time.The
Distributed Observer console and the Probe both function in a
"client/server" relationship. The Distributed Observer console acts
as the server: it is installed on a PC running Microsoft Windows
95/98/NT. The Probes are installed on each segment of the network
being monitored ( one per segment - and act as clients to the
Distributed Observer console. The Distributed Observer console's
function is to collect and decode packets and/or real-time LAN
traffic data from the Probe clients. Any Windows 95/98/NT PC on the
network can host a Probe. The Probe runs in the background (or as a
service under NT), collecting network information and sending it to
the Distributed Observer console while the user continues to work
as usual. Probes communicate with the Distributed Observer console
using Microsoft Winsock and the TCP/IP protocol. The Distributed
Observer console can monitor up to 252 Probes at one time. Probes
can be located on any network ( local or remote, Ethernet, Token
Ring or FDDI ( and are connected to the Distributed Observer
console via IP. NT servers, IP routers or WAN connections can serve
as bridges between the Distributed Observer console and its Probes,
and these can be run unattended for collecting long term trending
information.Currently FHS has seven Probes installed and is able to
keep on top of its entire network. According to Dawson, Distributed
Observer gives them the remote features needed to monitor their
network at a reasonable price. "It is now our principal network
analysis tool and we use it on a regular basis," he says. "It is
helping us in a number of areas because of the many functions it
performs."One of the Distributed Observer features that FHS uses
most is Bandwidth Utilisation. "This allows us to see at anytime
exactly how much of the available bandwidth is being used on the
network," Dawson explains. "We can pinpoint a possible surge on the
network, what is causing it, and get the users to stop whatever
they are doing at that particular time and continue that task when
there isn't so much bandwidth being used." When bandwidth isn't the
problem, Distributed Observer discovers what the actual problem
might be. It may be a configuration error, several applications
running concurrently or a broadcast storm. "With Distributed
Observer, we have the ability to identify the type of problem we
are dealing with and correct it quickly."The product also helps
them in the reverse. For example, it may not be a network problem,
but the server may be down. "Distributed Observer points us in the
right direction, and we can immediately improve the
situation."Distributed Observer has made an already healthy network
even better. It also helps FHS plan for the future as new
technology appears on the horizon. "We use it to monitor the
network, and thus are keeping ahead of the game," Dawson says.
"Network problems can creep up rather quickly if you don't keep on
top of things. We know that our network is healthy because of
Distributed Observer."
Compiled by Craig Hinton(c) Network
Instruments 1999