The definition of Network Monitoring no longer confines to just setting up a solution that is capable of garnering information from the network. With the corporate world looking at IT as that critical function having a say on day to day business, the choice of a solution to Monitor Network must factor in the business aspects such as increased employee productivity and saving on infrastructure costs, besides the obvious need of the solution having to aide the business with all the ‘must–have’ functionality either builtin or by way of facilitating useful plug–ins. No two networks are same and so it is only fair to expect a network monitoring solution to work on a ‘one size fits all’ principle in the not–so–perfect IT world.
When we talk about IT directly impacting the business, it could be anything from a seemingly simple problem like a web page taking eternity to load or a poor LAN connectivity, to more serious ones like an important email from a prospect not making it to your inbox, a CRM database crash, or even dealing with a mischief maker within. While these issues can be addressed by a variety of vertical solutions in the market, there is nothing like the convenience of a single point access to visualize the entire network to manage the fault, performance, configuration, and security or the other resources within. With a wide range of IT functionality fitting into the ‘network monitoring’ umbrella, it is important to look for the ‘must–haves’ to help align your IT with the business goals. It serves to keep in mind that to an administrator, network performance, security, fault management, and reliability are not mutually exclusive. A solution that serves all of these on one platter, keeps the administrator and his network happy! Anything more is a welcome bonus!
Let us take a closer look at the main concerns of an administrator and what data he will mine for in the minutiae of information gathered by the solution :
With the corporate networks getting more complex due to huge and distributed infrastructure, automatic discovery leaves little room for manual errors. Constant upgrades and additions to the network is nothing new and this calls for a provision to initiate a discovery on demand too. So, a solution to Monitor Network must be capable of automating the discovery and it must also accommodate a forced discovery.
Smart Classification and Mapping
The infrastructure to be managed include network devices, servers, applications, and other resources on the network. Different parameters determine the health of performance of these resources on the network. Classification of the infrastructure based on the type, and provision to map or logically group devices like clustered environments or geographically distributed resources, empowers the administrator and helps him visualize his network and manage.
In-depth Network Performance Monitoring
Performance degradation is a nightmare of an administrator. Any network resource can pull down the performance of a network, and the factors affecting the performance could be internal or external. Faults such as a hardware resource outage, a WAN link failure, a database application crash etc., have a cascading effect and the impact is larger than we perceive. The key areas an administrator must keep an eye on to assure a network that is 100% available include, availability and uptime monitoring, system resources monitoring and network bandwidth monitoring.
A secured network is a good, healthy network. The challenges here are huge as the administrator cannot make the slightest compromise. Like any other aspect of network monitoring, ‘prevention is better than cure’ is the motto here too. While intrusion detection, intrusion prevention etc may not usually be a part of a Network Monitoring solution, support for plugging in even a third party utility augurs well with the network security administrators. The areas an administrator focuses on to secure his network starts at keeping a close tab on the system log messages including Windows Event Logs, Syslogs on Unix–based devices, Firewall logs etc.
Any good network monitoring solution must have an intelligent alerting mechanism using which the IT team can productively collaborate and work efficiently. As we discussed earlier, any component or a resource in a network can play truant and pull down the network. A delay in preventing a fault from occurring, or repairing a damage in a lesser turn–around time requires a fool proof alerting mechanism where the concerned engineer gets to know the source of the problem by way of a meaningful alert.
Scalability of the Solution
All said and done, a network monitoring solution must not take a beating and crash or it must do so with a warning at the least! A server on which you host the monitoring solution, or the monitoring application itself is as susceptible as the other resources on the network. Having a redundant server take over and provide uninterrupted service to Monitor Network is the dream of an administrator.