Most office network problems stem not from a lack of expensive equipment but from basics skipped during installation. Seven common mistakes and their fixes, from port capacity to the IP plan and handover documents.
Most of the decisions made when an office network is set up stay invisible in day-to-day work, until the printer disappears, a call drops in the middle of a meeting or a new desk needs a cable run. A large share of these problems come not from a lack of expensive equipment but from a few basics skipped during installation. The seven mistakes below are common in small and medium-sized offices, and most can be avoided at little cost if they are considered from the start.
Seven common mistakes
1. Building the network for today's desk count
If there are 20 people in the office on installation day, a 24-port switch (the device that connects the devices on your network to each other) looks sufficient. But each employee connects more than a computer: IP phones, printers, wireless access points, cameras, meeting room displays and door access systems all need ports too. Within a few months the ports run out and small extra switches start appearing under desks.
When counting ports, plan more than one socket per workstation, count devices in shared areas separately and add room for growth. If devices such as access points and cameras draw power over the network cable (PoE), also calculate the switch's total PoE power budget; even when there are enough ports, there may not be enough power. Leave space in the network cabinet for another switch and patch panel as well.
2. Settling for home-grade and unmanaged equipment
Modem-routers designed for homes and unmanaged switches do the job at first, but their limits show as the network grows. Features such as splitting the network into segments (VLANs), per-port traffic monitoring, loop protection and detailed logging are either missing or very limited on these devices. When something goes wrong, it becomes hard to see what is happening on which port and which device.
Not every device on the network needs to be business-grade. But a managed core switch in the cabinet gives you real flexibility, both for troubleshooting and for separating the network later. When choosing equipment, also ask how long the manufacturer will keep releasing software updates for that model.
3. Taking no precautions against cable loops
Plugging both ends of a patch cable into the same switch by mistake, or into two extra switches under desks, creates a loop in the network. In a loop, broadcast traffic (messages sent to every device on the network) keeps multiplying and circulating, and can bring the whole office network to a halt within moments. The symptom is usually "everything stopped at once", and finding the cause can take hours.
Loop prevention protocols (such as Spanning Tree) and loop detection features on managed switches should be switched on. Limit uncontrolled extra switches under desks; if more ports are needed, prefer a new cable run from the cabinet. Not leaving the cabinet ends of unused wall sockets connected to the switch reduces both the loop risk and the risk of unauthorised connections.
4. Having no IP address plan
Every device on the network is identified by an IP address. Without a plan, an address set manually on a printer clashes with the pool of the service that hands out addresses automatically (DHCP); one day the printer "disappears" or the server can't be reached. Write down which range is handed out automatically and which fixed addresses the servers, printers and network devices will use. Using DHCP reservations instead of setting addresses by hand on each device also reduces conflicts.
The choice of address range matters too. Most home routers use the 192.168.1.x or 192.168.0.x range. If the office network uses the same range, employees connecting from home over VPN may have trouble reaching office devices. Choosing a less common private address range avoids this clash from the start. Also, the widely used /24 network offers at most 254 device addresses; as phones and wireless devices are added, that limit can fill up sooner than expected.
5. Putting everything on one flat network
Having staff computers, the server, cameras, guest phones and printers on the same network makes setup easy, but it creates a structure in which every device can reach every other device. A camera that hasn't been updated in years, or a guest device infected with malware, then sits on the same network as your server. Moving guests, cameras and servers into separate segments reduces this risk. Since separating them later takes more effort, designing your address and port plan with separation in mind makes life easier, even if you don't implement it on day one.
Wireless is not a substitute for cabled infrastructure. Connecting devices that don't move, such as desktop computers, servers, printers and access points, by cable wherever possible, and leaving the wireless network to people on the move, improves both speed and stability.
6. Treating the network cabinet and power as an afterthought
Network equipment often ends up in a leftover corner, an unventilated cupboard or next to the cleaning supplies. Devices that overheat can restart unexpectedly and wear out sooner. A ventilated, locked cabinet that only authorised people can open is a basic requirement for the network's physical security.
Plan for power cuts as well. If the modem, firewall and switches are not on an uninterruptible power supply (UPS), even a short outage drops every connection, and phones, cameras and access points powered over PoE go down with the switch. Size the UPS to the real consumption of these devices, including the PoE load, and schedule regular battery checks.
7. Accepting the installation without tests and documentation
When the installation is finished, everything may look fine; the real test comes with the first fault. At handover, ask for a port map in which every socket and patch panel port carries the same number, the cable test results, configuration backups of the devices, the IP address plan, and the administrator accounts handed over to the business. Also check that default administrator passwords have been changed, device firmware has been updated, and unused remote management features and UPnP (a feature that lets devices open ports in the firewall by themselves) have been switched off.
Without this information, even a simple faulty socket turns into tracing cables through the cabinet. And when management of the network passes to someone else or another company in the future, whoever takes over will need the same information.
Where to start on your existing network
Some of these mistakes may already be present on your network, and you don't need to fix them all at once. Start with low-cost, high-impact steps: change default passwords, connect the cabinet equipment to a UPS, list the devices with fixed addresses and the DHCP range, and label the ports to produce a port map. Timing larger jobs, such as extra switches, new cabling or network separation, to coincide with an office move, refurbishment or equipment replacement reduces both disruption and cost.
Make changes outside working hours where possible, and back up the device configuration before each step. Straight after changes such as a new address plan or network separation in particular, test that printers, servers and critical applications are working.
Checklist
- Have the port count and PoE power budget been calculated with room for growth?
- Is the core switch in the cabinet managed, and does it still receive updates?
- Is loop protection enabled, and are extra switches under desks under control?
- Are the DHCP range, fixed addresses and reservations recorded in a written plan?
- Have guest, camera and server devices been separated from the staff network?
- Is the network cabinet ventilated, locked and protected by a UPS?
- Have the port map, cable test results and configuration backups been handed over?
- Have default passwords been changed, and unnecessary remote management and UPnP switched off?