In enterprise campus and data center networking, bouncing a switch port is one of the most fundamental operational tasks performed by network administrators. Whether restoring connectivity to a frozen IP camera, clearing stale MAC address entries after a patch cable swap, or re-initiating 802.1X port authentication for a stubborn workstation, cycling an interface resets the physical and data link states to resolve elusive Layer 1 and Layer 2 glitches.
While the concept of cycling an interface is simple, doing so safely and efficiently requires understanding Cisco IOS CLI mechanics, avoiding remote lockouts when working over an uplink, isolating PoE power cycling, and troubleshooting interfaces that refuse to come back up. This comprehensive guide covers everything from the core command reference to automated recovery scripts and diagnostic workflows.
Table of Contents
Port Bouncing Fundamentals: Purpose, Mechanics, and Side Effects
Before issuing commands on production hardware, it is critical to understand how an administrative port bounce interacts with network state tables and attached endpoints.
Frequently Asked Questions
Q: What does “bouncing a port” mean in Cisco networking?
A: Bouncing a port is the process of administratively disabling an interface and immediately re-enabling it. In Cisco IOS and Cisco IOS-XE, this is executed by issuing the shutdown command followed by no shutdown. This operation forces the physical link (PHY) down and brings it back up, simulating a physical cable disconnect and reconnect without requiring manual intervention in the wiring closet.
Q: What technical issues does bouncing a port solve?
A: Cycling an interface addresses several transient network states:
* DHCP Lease Re-negotiation: Forces the client endpoint to detect a media disconnect, causing its operating system to release its previous IP and broadcast a new DHCPDISCOVER.
* 802.1X and MAB Re-Authentication: Re-triggers Extensible Authentication Protocol over LAN (EAPoL) exchanges or MAC Authentication Bypass (MAB) with network access control servers like Cisco ISE.
* Flushing Dynamic CAM/MAC Tables: Instantly purges stale hardware address associations tied to that specific port on the switch.
* Auto-Negotiation Reset: Re-initializes Ethernet speed and duplex negotiation if an interface becomes locked in a mismatch or half-duplex mode.
Q: What side effects occur during a port bounce?
A: Dropping an interface introduces immediate operational side effects:
* Session Termination: All active TCP/UDP connections, voice calls, and secure tunnels traversing the interface terminate immediately.
* Spanning Tree Protocol (STP) State Transitions: The port re-evaluates topology state. Unless configured with spanning-tree portfast, the interface cycles through the Listening and Learning states before reaching Forwarding, introducing a 30-second delay.
* Upstream Routing Adjacencies: If the port is configured as a routed interface (no switchport), neighbor adjacencies for protocols like OSPF, EIGRP, or BGP drop, initiating route recalculation.
With the operational impact clear, let us examine the formal command structure defined by Cisco IOS.
Cisco IOS Interface Command Reference: shutdown and no shutdown
The administrative state of any physical, logical, or virtual interface within Cisco IOS and Cisco IOS-XE is controlled using the shutdown and no shutdown interface configuration commands.
Command Definition and Syntax
| Command | Command Mode | Description |
|---|---|---|
shutdown | Interface configuration (config-if) or Interface range (config-if-range) | Administratively disables the interface, turning off signal transmission and dropping link carrier state. |
no shutdown | Interface configuration (config-if) or Interface range (config-if-range) | Administratively enables the interface, permitting physical carrier detection and data forwarding. |
Operational Parameters and Defaults
- Default State: On Layer 2 Cisco Catalyst switches (such as the Catalyst 2960, 3850, 9200, and 9300 series), physical 10/100/1000/10G Ethernet ports default to the enabled state (
no shutdown). Conversely, routed interfaces, Switch Virtual Interfaces (SVIs/VLAN interfaces), and loopback interfaces on certain platform images default toshutdownuntil explicitly activated. - Hardware Signaling: Issuing
shutdowndeactivates the physical Layer 1 transceiver or PHY circuitry on copper interfaces and disables laser transmission on optical SFP/SFP+ transceivers. - Status Interpretation:
- When
shutdownis applied:Line statusreports administratively down;Line protocolreports down. - When
no shutdownis applied without a physical cable:Line statusreports down;Line protocolreports down. - When
no shutdownis applied with a valid physical link:Line statusreports up;Line protocolreports up.
Having established the baseline command reference, the following section covers the routine procedure for cycling client-facing access ports.
Step-by-Step Guide: Bouncing a Switchport via CLI
The standard administrative bounce is executed from the Cisco IOS command-line interface via console, SSH, or Telnet.
Step 1: Access Privileged EXEC Mode
Log in to the switch and elevate your privilege level:
Switch> enable
Password: *********
Switch#
Step 2: Enter Global Configuration Mode
Open the configuration staging buffer:
Switch# configure terminal
Enter configuration commands, one per line. End with CNTL/Z.
Switch(config)#
Step 3: Select the Target Interface or Interface Range
Specify the exact physical interface you intend to bounce. You can select an individual port or a range of contiguous/non-contiguous ports:
! For a single interface:
Switch(config)# interface GigabitEthernet1/0/15
! For multiple interfaces simultaneously:
Switch(config)# interface range GigabitEthernet1/0/1 - 8, GigabitEthernet1/0/12
Switch(config-if-range)#
Step 4: Administratively Disable the Port
Execute the shutdown command:
Switch(config-if)# shutdown
System confirmation logs will confirm the state transition:
%LINK-5-CHANGED: Interface GigabitEthernet1/0/15, changed state to administratively down
%LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet1/0/15, changed state to down
Step 5: Allow Link-Down Propagation
Wait 3 to 5 seconds. This brief pause ensures that the connected device’s Network Interface Card (NIC) registers the loss of carrier signal, clears local socket state, and tears down active network sessions cleanly.
Step 6: Administratively Re-enable the Port
Restore the interface to operational service:
Switch(config-if)# no shutdown
Switch(config-if)# end
The switch logs will reflect the physical carrier link restoration:
%LINK-3-UPDOWN: Interface GigabitEthernet1/0/15, changed state to up
%LINEPROTO-5-UPDOWN: Line protocol on Interface GigabitEthernet1/0/15, changed state to up
Step 7: Verify Interface Operational Status
Confirm that the interface has negotiated properly and entered the forwarding state:
Switch# show interfaces GigabitEthernet1/0/15 status
Port Name Status Vlan Duplex Speed Type
Gi1/0/15 Workstation-A connected 10 a-full a-1000 10/100/1000BaseTX
While this step-by-step workflow works seamlessly for edge ports, issuing a manual shutdown on a remote uplink or management port will immediately disconnect your SSH session, stranding the switch in an unreachable state.
Template: Safely Bouncing Remote Uplinks Without Lockout
When managing a switch remotely over an IP network, bouncing the trunk port or management uplink over which your SSH session travels requires an atomic, automated process. Because an interactive shutdown kills the management session before no shutdown can be entered, you must execute both commands programmatically on the switch CPU.
Use the following Cisco IOS Embedded Event Manager (EEM) template to safely bounce remote interfaces without the risk of an unrecoverable lockout.
Production EEM Applet Template
! =========================================================================
! TEMPLATE: Automated Remote Port Bounce via Cisco IOS EEM
! Description: Disables and re-enables an interface locally on the switch.
! Usage: Replace <INTERFACE_NAME> with the target port (e.g., Gi1/0/48).
! =========================================================================
configure terminal
event manager applet BOUNCE_UPLINK authorization bypass
event none
action 1.0 syslog msg "EEM: Commencing safe bounce of interface <INTERFACE_NAME>"
action 2.0 cli command "enable"
action 3.0 cli command "configure terminal"
action 4.0 cli command "interface <INTERFACE_NAME>"
action 5.0 cli command "shutdown"
action 6.0 syslog msg "EEM: Interface <INTERFACE_NAME> administratively DOWN. Pausing..."
action 7.0 wait 5
action 8.0 cli command "interface <INTERFACE_NAME>"
action 9.0 cli command "no shutdown"
action 9.1 cli command "end"
action 9.2 syslog msg "EEM: Interface <INTERFACE_NAME> administratively UP. Bounce complete."
exit
Execution Command
Trigger the applet directly from privileged EXEC mode:
Switch# event manager run BOUNCE_UPLINK
(Note: Your SSH session will freeze momentarily during the 5-second wait window, after which the link will recover and connectivity will resume automatically.)
Optional Safety Net: Reload Timer Safeguard
If executing configuration changes directly on an uplink without EEM, configure a timed automatic rollback to protect against misconfiguration:
Switch# reload in 10
System configuration has been modified. Save? [yes/no]: no
Proceed with reload? [confirm] <Enter>
! Perform your changes. If connection is permanently lost, the switch reloads in 10 min.
! Once connectivity is verified safe, cancel the timer:
Switch# reload cancel
While administrative link bouncing resets data signaling, it does not always cut power to endpoints receiving Power over Ethernet.
Step-by-Step Guide: Power-Cycling PoE Devices Without Data Link Shutdown
Many edge devices—including Cisco IP phones, wireless access points, and PTZ IP surveillance cameras—run embedded operating systems that can experience kernel freezes or application crashes while the physical Ethernet link remains electrically up. In these cases, bouncing the data link via shutdown / no shutdown may leave the PoE power rails energized, failing to reboot the hardware.
To perform a true hard power reboot without modifying VLANs, STP settings, or port security parameters, toggle the inline power delivery system directly.
Step 1: Verify Connected PoE Draw and Power Mode
Check the inline power delivery status and device consumption on the target port:
Switch# show power inline GigabitEthernet1/0/5
Interface Admin Oper Power Device Class Max
(Watts)
--------- ------ ---------- ------- ------------------- ----- ----
Gi1/0/5 auto on 15.4 IP Phone 8845 4 30.0
Step 2: Enter Interface Configuration Mode
Access the specific switchport:
Switch# configure terminal
Switch(config)# interface GigabitEthernet1/0/5
Step 3: Disable Inline Power Delivery
Cut the DC voltage supplied across the twisted pairs by setting the power mode to never:
Switch(config-if)# power inline never
The switch will cut power immediately and log the status change:
%ILPOWER-5-POWER_DISCONNECT: Interface Gi1/0/5: inline power disconnected
%LINK-3-UPDOWN: Interface GigabitEthernet1/0/5, changed state to down
Step 4: Allow Capacitor Discharge
Wait 10 seconds. PoE endpoint power supply capacitors require several seconds to drain stored energy completely and ensure a true hardware cold restart.
Step 5: Restore Automatic Inline Power Detection
Re-enable standard IEEE 802.3af/at/bt resistive discovery negotiation:
Switch(config-if)# power inline auto
Switch(config-if)# end
The switch will detect the signature resistance, allocate power budget, and energize the device:
%ILPOWER-7-DETECT: Interface Gi1/0/5: Power Device detected: IEEE PD
%ILPOWER-5-POWER_GRANTED: Interface Gi1/0/5: Power granted
%LINK-3-UPDOWN: Interface GigabitEthernet1/0/5, changed state to up
Step 6: Validate Operational Restoration
Confirm that inline power is active and the endpoint has re-registered:
Switch# show power inline GigabitEthernet1/0/5
Sometimes, whether after a PoE power cycle or a standard administrative bounce, an interface fails to transition to up/up and instead drops into an error state.
Troubleshooting Guide: Resolving Ports Down or err-disabled After a Bounce
When an interface remains down after issuing no shutdown, the switch has typically detected a Layer 2 policy violation or a Layer 1 signaling failure. Follow this systematic diagnostic sequence to identify and correct the underlying fault.
Step 1: Identify the Interface Operational State
Query the switch port summary to identify whether the port is disconnected or error-disabled:
Switch# show interfaces GigabitEthernet1/0/12 status
Port Name Status Vlan Duplex Speed Type
Gi1/0/12 Uplink-Distribution err-disabled 1 auto auto 10/100/1000BaseTX
If the status displays notconnect, the switch is administratively active but detects no electrical carrier. Verify physical cabling, patch panel punch-downs, SFP seating, and endpoint power. If the status displays err-disabled, proceed to Step 2.
Step 2: Identify the Root Cause of the Error
Execute the diagnostic command for error-disabled interfaces to inspect the triggering mechanism:
Switch# show interfaces status err-disabled
Port Name Status Reason Err-disabled Vlans
Gi1/0/12 Uplink-Distribution err-disabled bpduguard
Step 3: Remediate Common Root Causes
Before re-enabling the interface, correct the operational violation that triggered the protection mechanism:
- BPDU Guard (
bpduguard): A switch, rogue router, or virtualization host sending Spanning Tree Bridge Protocol Data Units was detected on a port configured withspanning-tree bpduguard enable(or PortFast). - Fix: Disconnect the unmanaged switch/device or disable BPDU Guard on trunk uplinks:
no spanning-tree bpduguard enable. - Port Security Violation (
psecure-violation): A newly connected device presented a MAC address that exceeded the configured threshold or violated sticky bindings. - Fix: Clear unauthorized MAC entries or update the security table:
clear port-security dynamic interface Gi1/0/12. - UniDirectional Link Detection (
udld): The switch detected that fiber transmit (TX) traffic is flowing, but receive (RX) continuity is broken. - Fix: Inspect optical patch cables, swap TX/RX pairs, or clean transceiver optics.
- Speed/Duplex Mismatch: Incompatible manual speed settings prevent physical framing sync.
- Fix: Align both ends to
speed autoandduplex auto.
Step 4: Clear the Err-Disabled State and Re-enable
Once the root cause is resolved, reset the port’s internal state machine by toggling the administrative state as outlined in our CLI guide above:
Switch# configure terminal
Switch(config)# interface GigabitEthernet1/0/12
Switch(config-if)# shutdown
Switch(config-if)# no shutdown
Switch(config-if)# end
Step 5: Configure Automatic Err-Disable Recovery (Best Practice)
To prevent network engineers from having to log in manually whenever a transient policy violation occurs, configure automated recovery timers in global configuration mode:
Switch(config)# errdisable recovery cause bpduguard
Switch(config)# errdisable recovery cause psecure-violation
Switch(config)# errdisable recovery cause udld
Switch(config)# errdisable recovery cause link-flap
Switch(config)# errdisable recovery interval 300
Switch(config)# end
This configuration directs the switch operating system to attempt an automatic recovery every 300 seconds (5 minutes), re-enabling the port automatically if the offending condition has cleared.
Conclusion
Bouncing a port on a Cisco switch is an indispensable troubleshooting technique that resets Layer 1 signaling and Layer 2 state machines without requiring physical access to equipment racks. For routine endpoint issues, a straightforward shutdown and no shutdown cycle cleanly clears dynamic MAC tables and forces client DHCP and 802.1X renegotiation. When working across remote connections, using Cisco IOS EEM scripts prevents accidental lockouts, while cycling inline PoE power provides a targeted cold reboot for frozen IP phones, cameras, and access points without affecting data port logic. Finally, combining diagnostic commands with automated errdisable recovery ensures that interfaces return to service quickly and stay resilient against unexpected network disruptions.











Comments