BlogInfrastructure
Business networks: doing it right,
from cabling to switches
A badly built network does not fail outright — it just runs badly. What structured cabling means, why a managed switch matters, and the order things must be done in so you do not pay twice.
The sign that a network was built badly
It is not that the internet drops. It is that everything works, but poorly, and nobody knows why. Cameras stutter when someone copies a large file. Wi-Fi is fine in the office and non-existent in the hall. A new device gets plugged in and something breaks elsewhere. When the cause is hunted down, it leads to a cabinet with thirty unlabelled cables.
All of it has the same root: the network grew piece by piece, without a plan, and every piece was added by somebody else.
Structured cabling: what it really means
It means every network socket in the building runs, through its own uninterrupted cable, to one place: the patch panel in the rack. Every end is numbered at both ends and recorded in a diagram. Cables are not spliced, and small switches are not scattered around offices so someone can get two more ports.
The benefit shows two years later, when you move an office or add a camera: you change a cable in the rack instead of opening a wall. The extra cost at installation is small. The cost of not doing it is paid at every change, forever.
- Cable of the right category, run whole, with no splices
- Labelled at both ends, with the same numbering as the diagram
- A patch panel in the rack, not cables going straight into a switch
- Routes kept separate from power cabling, to avoid interference
- Every socket tested, with a report, at handover
Managed switches: not a luxury
A plain switch connects everything to everything. A managed switch lets you separate traffic into zones: cameras in one, computers in another, guest Wi-Fi fully isolated. That means a compromised device does not automatically reach the rest of the network, and a large transfer does not stall the cameras.
It is also where power over the network cable comes from, feeding cameras and access points without a separate socket beside each one. In a warehouse, that alone justifies the price difference.
The order things get done in
Measure the building, do not guess
Where the people are, where the equipment is, where cable can be run, which thick walls will kill the Wi-Fi.
Choose equipment against the plan
How many ports, how much power over cable, what coverage. Advice at purchase usually saves more than it costs.
Install
Cabling, rack, patch panel, switches, access points, UPS. With proper cable management, because in two years somebody will have to work in there.
Test and document
Every socket tested, the diagram handed over. Without a diagram, the next person to open the rack starts from zero.
The UPS everyone remembers too late
A ten-second power cut restarts the switches, the server and the cameras. If the server was writing to the database at that moment, sometimes it repairs itself and sometimes it does not. A UPS on the rack costs far less than a lost working day and far less than recovering a database.
For companies with cameras, a UPS also means surveillance keeps running during exactly the minutes when the power goes out.
What this is about, in full
Frequently asked questions
How long does a network take in a new hall or office?
Usually a few days of installation, after an on-site assessment. The long part is not the installation, it is choosing the right routes and equipment.
Can it be done in an already finished building?
Yes, but the routes get more complicated and surface trunking is sometimes used. That is why the on-site assessment happens before the quote, not after.
Do you work on houses as well as companies?
Yes. Houses add smart home and cameras, and the logic stays the same: cabling and network first, then the devices that sit on them.
Do you handle maintenance after installation?
Yes, with monitoring and callouts. An unsupervised network degrades slowly, and problems only get noticed once they hurt.