Your network exists to deliver the applications and services that matter to your business because you are dependent upon these applications and services in order for your business to accomplish the tasks at hand. It is the network, and these applications and services provide the means by which you get the work of your business done. QuickBooks, E-mail, Print-sharing, Data-storage, VoIP, and every imaginable application provide the tools you rely on for the conduct of day-to-day business. Network users expect the services and applications on the network to not only work, but work fast.
Most small to medium businesses operate their network without a full-time network professional, they may have an in-house ‘computer guy’ (maybe even the business owner or general manager), or rely solely on the expertise of a consulting Information Technology company. In either case, it is doubtful that anyone is continually monitoring the performance of the network to determine how it is really being used.
Probably nobody has any idea as to traffic, packet capture and protocol analysis that can give you a view of each user’s network availability and performance. Even if you are fortunate to have your own full-time network engineer, they probably spend more time installing new-equipment, setting-up applications, performing operating system and application updates, and resolving unique issues then they do really monitoring the network’s specific performance parameters.
We all know the ‘typical complaint’, “the network is running slow today.” Somebody goes around asking, “does the network seem slow to you?” Some users will say that it is just “QuickBooks running slow across the network”, but other uses will say that they are having to “wait and wait” on internet applications when they attempt to access a particular website. You pick-up the phone and it seems to take twice as long as normal for you to get a dial tone over your VoIP based system.
And so, perhaps, the complaints either go ignored everyone telling themselves, it is just the way it is, or the slow and ineffective exploration begins. Some tech from the local IT-company attempts to use a Simple Network Management Protocol solution to troubleshoot the network and perhaps put out the network fire at hand. But this type of after-the-fact approach is doing nothing, more than likely the problem has resolved itself, at least temporarily, by the time the diagnosis even beings. What you really need is a good proactive network performance management system, but before we look at what such a system looks like, we need to understand about the level of monitoring this type of analysis demands.
Troubleshooting Network Activity
When it comes to troubleshooting complex, and intermediate network problems, you really are talking about packet-level measures of activity. Packets give you the ‘what’s going on’ description of real time activity of your network. It is possible to capture this information in two ways.
The first monitoring method involves connecting to whatever you suspect is the problem and then recording the activity. This type of approach is extremely limited, by the time you get it hooked up, the problem has self-resolved. In intermittent problems, you have missed the period of diminished performance.
The second involves establishing permanent key network spots to continuously capture data for analysis. This type of monitoring continuously records all the information needed for analysis on a historical, real-time and prospective basis
Permanent Performance Monitoring
This type of system should be completely automatic and dynamic using advanced analytics to give you a true sense of how things are running on a moment-by-moment basis including the who, what, where and how of your network. While we discussed the need for applications that monitor the ‘who is doing what’, and if network users are using or abusing your network, and your employment trust, in our last article, the application we are discussing here is far more about ‘metering’ every aspect of the health of your network.
When a patient is in the ‘intensive care unit’ of the hospital they are hooked up to all kinds of monitoring that measures their EKG, their blood pressure, body temperature, pulse rate, and blood oxygen levels. Input and output measures are made and recorded so that a graphic representation of the patient’s overall health can be assessed.
Well if you have network problems, it isn’t so easy to get ‘intensive care unit’ measuring in place before the network recovers or dies, so what I am proposing is that your network gets monitored in such an ‘intensive’ fashion at all times, so that the information is not only available at a moment’s notice, but that it is continually being recorded. When this monitoring is coupled with analytics developed from historical values, as well as maximized performance testing, the monitoring application can alert you when anything changes the networks status from ‘normal’. Measures such as response time, throughput, and connection rates should be measured and tracked as part of the base-line for your network standards. Monitoring of this type should include performance measures for latency, congestion, low bandwidth, consumptive applications or contention with other applications, in order to determine viable network reconfiguration options.
The monitoring application should include both dashboard displays of immediate and ‘trended’ performance parameters as well as available reporting that targets different audiences, not only the technical gurus but executives who what to know the ‘how and why’ of their networks. Of course this type of reporting makes it much easier to justify the expense associated with such monitoring applications.
With a best-in-breed network monitoring and troubleshooting application you should expect to see improvements in the availability of applications and services on your network, better response times by your network, and improved application performance. If you have network issues that just keep popping up, and all your IT resources can’t seem to identify and resolve them on a long-term basis, then it really is time that you find out ‘how your network is being used’, via a comprehensive network monitoring and troubleshooting application.
Note: This article neither recommends nor endorses any specific product, there are several excellent products on the market. Contact your local IT-company or consultant for assistance in identification, evaluation and acquisition of such a solution, or feel free to contact the author for specific recommendations.