IT service management standards such as the IT Infrastructure Library ® (ITIL®) are very mature and practised to a greater or lesser extent by most IT operations teams.

These models aim to provide a rigorous structured framework for managing problems, incidents and changes. Some of you may be familiar with the mantra “Don’t blow your nose without submitting a change request”. The decisions of IT departments is made based on data, it therefore follows that the effectiveness of IT departments are dependent on the quality of data available to the engineers, designers, architects and support staff.

In a complex network, with multiple teams with specific areas of expertise, time may be wasted determining the most efficient escalation path to achieve resolution. Or to think of it another way, time can be wasted by skilled engineers proving that their area of responsibility is not the cause of the problem. Often an issue is passed from one team to another, with each team proving that it is innocent, and passing it on to the next.  This is known as LTTI

What is needed is a simple unified view of your network and applications. One that allows everyone to see the same information, presented in such a way that allows any one of the observers to drill down to the required level of granularity. This is where NPM and APM comes in. A comprehensive APM and NPM solution gives insights into the performance of your network from the operation of the network to the efficiency of the code developed by your software engineers.

This “single pane of glass” means that all are working from the same information and allows the correct escalation path to be selected early on. As a result, the team best placed to deal with the issue is quickly identified. This ensures that engineers waste less time proving that their team is not responsible. When measured statistically this is often called MTTI

Visibility in the dark
  • cPacket’s cVu and SPIFEE appliances will, for example examine packets in real-time allowing you to perform Google-like searches for any packet anywhere in your network, while looking deep into the packets giving insight into applications performance and behaviour at full link speed, all over the network, all of the time.  This allows data capture even for transient events anywhere on the network when they occur.
  • NetScout provide a unified view of network and application performance through its nGeniusONE product. Integrating network performance, voice performance, the ability to decode and analyse financial trading applications as well as giving insights into the performance of applications through deep packet inspection and application dependency discovery. Speedy fault diagnosis is its aim with flexible service configuration.
  • Riverbed’s AppInternals suite allows for comprehensive application performance management. AppInternals has the ability to report accurate timings for events giving a complete end to end transaction trace. This includes things like database process times, network travel time, code level performance such as time to execute a method. With systems such as this you can correlate network events with server/platform events and the application events into a single view which can be examined by engineers or analysed by software.

Many performance management systems also perform base-lining of elements in your network, platforms or applications, which not only gives proactive indicators as soon as there is a change in the performance of your network, well before the services are severely impacted, but also provides invaluable information which allows engineers to identify areas of improvement.

The key to performance management is visibility. You can’t fix what you can’t see, and ensuring everyone can see the same rich data increases efficiency of your network and the productivity of your teams.

Related articles