What's happening here
You plug a laptop into port 1 and another into port 2 on the same switch. Both link lights are solid green. But the laptop on port 1 can't ping the one on port 2. No DHCP, no ARP, nothing. This happens a lot when you're setting up a small office network or testing in a lab. I've seen it on Cisco 2960s, Netgear GS108s, and Ubiquiti USW-Lite-16s.
The root cause is almost always a VLAN mismatch or a port mode misconfiguration. Each port belongs to a VLAN. If port 1 is in VLAN 10 and port 2 is in VLAN 20, they won't talk to each other without a router. If port 1 is an access port in VLAN 10 and port 2 is a trunk port tagged for VLAN 10, traffic also fails. The switch is doing its job — it's just that the two ports live in different broadcast domains.
Another common cause is port security locking the MAC address. When the switch sees a different MAC on a secured port, it drops traffic. The port stays up, but frames don't pass. Spanning-tree can also hold a port in blocking state if there's a loop, but that usually shows as no link light.
The real fix- Check the VLAN for each port. Log into the switch CLI or web GUI. Run
show vlan brief on Cisco, or look at the VLAN membership page on a managed switch. Write down the VLAN ID for both problematic ports. - Ensure both ports are in the same VLAN. If they aren't, change one to match. On Cisco, go to interface config:
interface GigabitEthernet0/1
switchport mode access
switchport access vlan 10
Do the same for port 2. - Check port mode. Both ports must be in the same mode — both access or both trunk. Mix them and traffic won't forward for the VLAN you expect. On Cisco,
show interfaces status shows the mode. If one says trunk and the other access, fix it: interface GigabitEthernet0/2
switchport mode access
(or trunk if you need trunking). - Clear port security locks. If port security is on, the switch might have learned a MAC from a different device earlier. Run
clear port-security sticky on the port or globally. Then reconnect the cable. - Verify MAC learning. After connecting both devices, check the MAC address table:
show mac address-table interface GigabitEthernet0/1
If you don't see the device MAC, something's wrong upstream — bad cable or the port is errdisabled.
What to check if it still fails
- Cable or link issue: Swap the cable. Even if the light is on, a bad pair can cause frames to get dropped. Use a cable tester or try a known good cable.
- Spanning-tree blocking: Run
show spanning-tree interface GigabitEthernet0/1. If it says BLOCKING, you've got a loop. Disconnect one cable from a redundant link. - Port errdisabled: Check
show interfaces status errdisabled. If the port is down due to a security violation, shut it down and reenable: interface GigabitEthernet0/1
shutdown
no shutdown
- Switch firmware bug: I've seen Netgear GS108Ev3 switches not forwarding traffic after a power cycle until you manually save the config. Update firmware or reboot and reapply config.
show vlan brief on Cisco, or look at the VLAN membership page on a managed switch. Write down the VLAN ID for both problematic ports.interface GigabitEthernet0/1
switchport mode access
switchport access vlan 10 Do the same for port 2.show interfaces status shows the mode. If one says trunk and the other access, fix it: interface GigabitEthernet0/2
switchport mode access (or trunk if you need trunking).clear port-security sticky on the port or globally. Then reconnect the cable.show mac address-table interface GigabitEthernet0/1 If you don't see the device MAC, something's wrong upstream — bad cable or the port is errdisabled.show spanning-tree interface GigabitEthernet0/1. If it says BLOCKING, you've got a loop. Disconnect one cable from a redundant link.show interfaces status errdisabled. If the port is down due to a security violation, shut it down and reenable: interface GigabitEthernet0/1
shutdown
no shutdownIf all else fails, factory reset the switch (hold the reset button for 10 seconds on most models) and reconfigure from scratch. It's rare, but corrupt configs do happen.