Network Drift. It WILL Happen 100% of the Time. Guaranteed. | BAYCOM

No matter how well a wireless network is planned or how meticulously it is deployed, something will eventually happen over time to cause it to go adrift. From a need for increased capacity to power sources becoming undependable, changes to a wired network will occur. And unfortunately, there is no way around it.

The “normal” progression of events occurring on a well-functioning network not receiving ample administration or attention will occur to any network and may include:

• Mobile infrastructure placed in “non-optimal” locations
• Misconfigured radios inserted into the network
• Multiple applications enabled after network acceptance
• Wired network conditions changing (i.e. new VLANs added or altered, switch port configurations changed)
• Accidents causing power outages, fiber cuts, and antenna breaks

Over time, the issues listed above will happen slowly and surely, causing havoc on a network’s performance to the point of deterioration. This unwanted phenomenon, otherwise known as “Network Drift,” occurs 100% of the time to all networks. Guaranteed.

Fortunately, there is a solution. Rajant’s InstaMesh^® protocol does a superb job of hiding several small issues with respect to a Kinetic Mesh Network. The patented InstaMesh protocol dynamically routes around network outages, RF interference, detects network loops, and ensures that the application data gets through no matter what. But if the network remains unattended, issues will eventually accumulate to the point that InstaMesh can longer work around the multitude of issues that have occurred, and it will be the next “small” issue that will cause applications to suffer.

At this point, one experiences the proverbial, “the last straw that broke the camel’s back.” One can undo the last change that was made if the administrator is even aware of the change that occurred, but that may have already started the cavalcade of application issues presented to the users.

So how does one prevent Network Drift? Here are a few recommendations.

1. Obtain exact metrics on your network when it’s “healthy” and document those metrics. It’s always beneficial to know what “good” looks like. There is an excellent chance the next time you give your network a hard look is when someone complains. This is when you will see what “less than good” looks like. If you can’t recall what “good” is, it’s harder to find where to start, and more importantly, where you need to go.
2. If your network infrastructure moves (i.e., solar trailers or skids) quarterly, a NetCrumbler Run should be performed and analyzed for potential weaknesses in coverage.
3. Run Rajant’s BC|Enterprise and log historical metrics on your network. This will give you a point in time when things changed as well as give you metric histories of declining performance on a very granular level, down to a node and corresponding interfaces.

In summary, it is possible to stay in front of your Network Drift. With a protocol as robust as InstaMesh, it’s easy to get complacent and just let your network run. But a little pre-emptive maintenance and monitoring can make a world of difference.