Solution: Server Cabled to Wrong Switch Port¶
Triage¶
- Identify where the server is actually connected:
lldpctlorlldptool -t -n -i eno1on the server- Note the reported switch chassis name and port ID
- Compare with the cabling plan:
- Expected: sw-tor-c12 Eth1/33 (VLAN 320, production)
- Actual (from LLDP): sw-tor-c12 Eth1/9 (VLAN 50, development)
- Verify the target port (Eth1/33) is ready:
- On the switch:
show running-config interface Eth1/33 - Confirm it is configured for VLAN 320 and not in use
- Check the current IP and DHCP state:
ip addr show eno1-- note the 172.16.50.x address- This address is from the dev VLAN DHCP scope
Root Cause¶
The contractor plugged the server's network cable into switch port Eth1/9 instead of Eth1/33. The cable label reads "C12-P33" (correct), but the physical cable follows the wrong path in the cable management tray and terminates at Eth1/9. The ports are in different patch panel sections, and the contractor likely miscounted ports in the dense patch panel.
Port Eth1/9 is an active port configured for VLAN 50 (development). The server received a DHCP address from the dev VLAN scope and came up on the wrong network segment.
Fix¶
- Verify the target port:
- On the switch:
show interface Eth1/33 brief-- confirm status is "notconnect" (empty port) show running-config interface Eth1/33-- confirm VLAN 320 configuration- Move the cable:
- Physically trace the cable from the server's NIC to the switch
- Unplug from Eth1/9 and plug into Eth1/33
- This is a brief network interruption (seconds)
- Renew the IP address:
dhclient -r eno1 && dhclient eno1-- release old, request new- Or if using NetworkManager:
nmcli con down eno1 && nmcli con up eno1 - Verify connectivity:
ip addr show eno1-- should now show 10.30.20.xping 10.30.20.1-- gateway reachablelldpctl-- should now report Eth1/33- Update cable management:
- Re-label the cable if the label is misleading
- Physically re-route the cable through the correct cable management tray
- Take a photo of the final cabling for documentation
- Audit the batch:
- Check all other servers installed during the same contractor shift
- Run
lldpctlon each and compare with the cabling plan - Fix any other mismatch before those servers go into production
Rollback / Safety¶
- Moving a network cable is non-destructive; the worst case is a few seconds of packet loss.
- If Eth1/33 turns out to be misconfigured, the server will get no link/IP; just move the cable back and fix the switch config.
- The server was briefly on the dev VLAN; check if any dev traffic reached the server or if the server's DHCP lease created conflicts on the dev subnet.
Common Traps¶
- Reconfiguring the switch instead of moving the cable: Changing Eth1/9 from VLAN 50 to VLAN 320 "fixes" this server but breaks whatever was supposed to be on Eth1/9.
- Not checking if Eth1/9 is assigned to another server: If another server should be on Eth1/9, you may have displaced it. Verify.
- Trusting cable labels blindly: Labels can be wrong. Always verify with LLDP or physical trace.
- Forgetting DHCP lease cleanup: The old DHCP lease on the dev VLAN will persist on the DHCP server until it expires. Not harmful but messy.
- Not auditing the rest of the batch: If one cable is wrong, others from the same install are suspect.