Cables get blamed for dead ports because cables are visible and ports are not. This elimination method convicts or clears a port with four quick tests — and stops the re-crimping loop.
Every uplink fault has two ends, and human nature blames the end you can hold. A cable can be cut, re-crimped and replaced while the real culprit — a switch port that died silently — watches from the rack. Ports do die: surges ride in on long runs, PoE faults cook single ports, sockets get torn by sideways yanks. This guide gives you the short, strict test sequence that proves a port dead (or clears it) before any cable gets cut, plus the software states that pretend to be hardware death.
Switch ports fail more often than switches do. The honest causes: a power surge down a long run (inter-floor and inter-building links are lightning antennas), a PoE fault cooking one port's electronics, a physically damaged socket from a cable yanked sideways, or simple age. The failure is usually silent and total for that one port, which is exactly why it gets misdiagnosed as a cable problem — the cable is the thing you can see, so the cable gets re-crimped three times while the port sits there dead.
Accuse in the order of testability. Take a factory patch cord and a laptop — two items you trust — and plug them into the suspect port:
On a managed switch, a port can be dead by instruction rather than damage. Before writing it off, check: administratively shut down; port security tripped and latched after an earlier violation; an error-disable state from a loop or a flapping link; or the port assigned to a VLAN the test laptop cannot use (this one shows a healthy link light with no traffic — a different signature from a truly dead port, which shows nothing at all). Clearing an error-disable without finding why it tripped just schedules the sequel; loops and flapping cables trip it again within the hour.
Managed switches keep per-port statistics, and a dying port often confesses there: a counter frozen at zero while neighbours count millions, or an error counter that climbed for weeks before the port gave up. If the port carries PoE devices, also check the PoE budget — a switch that has exhausted its power budget refuses new powered devices on healthy ports, which mimics port death for exactly the devices that need power. Non-PoE test gear separates this instantly.
A single dead port on an otherwise healthy switch is not an emergency. Move the uplink to a free port, label the dead port (tape over it and write "dead" — future technicians will thank you), and note it in whatever passes for site documentation. Plan around the pattern, though: ports dying one by one on the same switch, especially after storms, points at surges arriving on the long runs — that is an argument for surge protection or a fibre uplink, not for another identical switch to sacrifice. If the dead port is under warranty, the test log from this page is exactly the evidence the vendor will ask for.
See the live troubleshooting in the video — the same cable tested with a laptop, then between two switches, and the wiring fault found on camera.
▶️ Watch Full VideoUse a trusted factory patch cord and a trusted laptop. If that combination gets no link in the suspect port but links instantly in the neighbouring port — and a second device also fails in the suspect port — the port is dead or disabled. If the laptop links fine in the suspect port, the port is innocent and the fault travels with the original cable or the far end.
Probably not dead — a lit link means the electronics negotiated successfully. No traffic with a healthy link points at configuration: the port in a wrong VLAN, port security blocking the device, or an IP problem. Truly dead ports show no link at all with any cable or device. Managed switches add one more mimic — an error-disabled port — which the interface status reveals.
Moving to a free port restores service, and do label the dead one. But treat the cause seriously: surges arrive on long copper runs, especially between floors and buildings, and a switch that loses a port per storm is telling you the run needs surge protection — or replacing with fibre, which carries no electrical surge path at all.