Nowadays, the IT network and its performance is an essential component of the performance of the overall organisation. If the organisation is providing network based services, network performance is critical to the success or otherwise of the organisation. As a result, the choice and application of network tools to manage network performance and to detect and signal network issues is an essential element in the engineer’s toolkit.
A network can be defined as structured cabling carrying network traffic. Just as in a highway network, traffic is moderated and controlled, and it can vary according to the time of day. Just as in a highway network it can suffer from bottlenecks, diversions and outright failures caused by faults or inadequate infrastructure on the route.
What is likely to be attached to a network made of structured cabling?
A network is typically made from a tree structure of up to three layers.
The user devices at the edge of the network are attached to switches. These switches are the access layer. In the case of a small network, the access layer switches are attached to the core. In the case of a larger network, a distribution layer switch consolidates several access layer switches into a single connection to the core.
At the centre is the core. The core is where all the big stuff sits. The Internet connection is here, the servers hosting the applications software are here, the ancillary servers with the anti-malware software, the email server are all here.
Network traffic to and from the user device therefore passes from the user device, via an access layer switch to the core, perhaps going through a distribution layer switch on the way.
Please note that a WiFi network can be installed instead of physical structured cabling. In this case Wireless Access Points (“APs”) are attached to the access layer switches, and the user devices use WiFi adapters to connect to the wireless environment created by the APs.
It is important for any size of network to manage the performance of the network. Users can be very quickly discouraged if the service they receive at their device is not satisfactory. In very large corporate networks the network management team have a Network Operations Centre where they use sophisticated network management tools to monitor traffic and link and device status.
As indicated above the network manager needs a range of tools to monitor network performance and raise alerts if network links fail or predefined service levels aren’t met.
There are several common problems encountered in network management:
If an internal or external link is broken, then devices attached to that link or using a service provided over that link will not be available.
If there is too much traffic for a link or network segment then all traffic slows and waits, or in some cases is rejected. Working out why is often a mixture of measurement and deduction. A network component may have failed elsewhere, causing additional network traffic on this link, there may be additional users on the link at peak times, it could even be the type of application operated by one user.
A malfunctioning device.
An attached device might be switched on but not configured properly or in the first stages of a hardware malfunction. In this case the device trying to send network traffic to this device is notified that the transmission has failed, and continually resends the failed traffic, causing the congestion. This may also be caused by a piece of faulty cabling or connector in the structured cabling itself.
A mis-configured device.
This occurs most often in an environment where a user can add their own devices. It is particularly common in the WiFi environment where any WiFi enabled device can potentially try to connect.
A more common problem nowadays is malware, either generated internally, or from an external attack on the network. Some malware is designed to flood a network with traffic and make it unusable for normal traffic.
This is where the network monitoring tools come into play.
By monitoring network traffic across the various LAN and WAN links the network manager can see how the traffic is moving and where any potential bottlenecks are forming. More advanced applications have a graphical interface showing the status of the network itself and the devices connected to it. Typically blue or green means good, orange is questionable and red means the link is down.
Network management must be proactive, not just responding
Many applications also include alarms that will alert the network manager to faulty equipment in the network.
Using a suitably configured network management tool can make the network manager’s life easier in being alerted to network problems and proactively monitoring network usage patterns.